Adedoyin Ahmed Hussain Ouns Bouachir Fadi Al-Turjman Moayad Aloqaily
Adedoyin Ahmed Hussain Ouns Bouachir Fadi Al-Turjman Moayad Aloqaily
Adedoyin Ahmed Hussain Ouns Bouachir Fadi Al-Turjman Moayad Aloqaily
ABSTRACT Artificial Intelligence (AI) intent is to facilitate human limits. It is getting a standpoint on
human administrations, filled by the growing availability of restorative clinical data and quick progression of
insightful strategies. Motivated by the need to highlight the need for employing AI in battling the COVID-19
Crisis, this survey summarizes the current state of AI applications in clinical administrations while battling
COVID-19. Furthermore, we highlight the application of Big Data while understanding this virus. We also
overview various intelligence techniques and methods that can be applied to various types of medical
information-based pandemic. We classify the existing AI techniques in clinical data analysis, including
neural systems, classical SVM, and edge significant learning. Also, an emphasis has been made on regions
that utilize AI-oriented cloud computing in combating various similar viruses to COVID-19. This survey
study is an attempt to benefit medical practitioners and medical researchers in overpowering their faced
difficulties while handling COVID-19 big data. The investigated techniques put forth advances in medical
data analysis with an exactness of up to 90%. We further end up with a detailed discussion about how AI
implementation can be a huge advantage in combating various similar viruses.
INDEX TERMS Big data, the IoT, artificial intelligence, cloud computing, deep learning.
This work is licensed under a Creative Commons Attribution 4.0 License. For more information, see https://creativecommons.org/licenses/by/4.0/
128776 VOLUME 8, 2020
A. A. Hussain et al.: AI Techniques for COVID-19
Meanwhile, Insilico Medicine situated in Hong Kong as of and morals contemplations [20] [21]. For instance, Google
late declared that it’s AI calculations had planned six new as of late distributed a rundown of moral standards identified
atoms that could end viral replication [6]. with the advancement of AI in June 2018 [22]. As of now,
Because medical information continues to improve rapidly, AI apparatuses and it is associated previously cited problems
medical data has grown quickly and huge numbers of medical are still inside the domain of researches and portrayal, how-
devices have created an incredible risk on current hospital ever, ordinarily, it acknowledged that the mentioned devices
information systems [7]. The field of healthcare generates will upset clinical methods and clinical network is starting to
a wide range of data on medical diagnosis, patient records, pay attention to this situation [23]. In any case, medication
treatment, medication, diagnosis, etc. The real concern is isn’t incorporating these instruments as fast as the innovation
that because of the limited data processing, the accuracy of is progressing. Also, lacking inclusion and also the participa-
certain reports has an impact on the organization of health- tion of wellbeing experts, AI can’t be coordinated into present
care [8]. From that issue, the data is useless without an practice. While a specific degree of overstatement appears to
accurate result which has been produced from the process- have assumed control over the conversation of AI in the clini-
ing of the huge data. Using the artificial intelligence can cal field, we additionally found that different contemplations
improve the result by employing systems that collect hun- and information have developed among every class of partner.
dreds of actual patient information and examine it by experts From one point of view specialists and scientists feature the
and artificial intelligence tools [9]. Big data in healthcare requirement for excellent clinical information; then again,
sectors are already applied in some hospitals, especially in doctors are as yet sitting tight for proof of the convenience
the field of radiology by using advanced algorithms such as of these instruments, and miracle if they will be considered
Deep Learning [10]. Big data visualization plays an impor- capable if there should arise an occurrence of a physical
tant role also not only in the management of the hospital issue because of an AI apparatus that they don’t completely
but also in reducing the medical waste [11]. Simultane- comprehend. In this way, in the coming years, society should
ously, various organizations in France have indicated their be careful in regard to the spot given to huge information
enthusiasm for the utilization of AI in medicinal services. and AI. Moreover, since AI keeps on pushing ahead, it is up
Therefore, as indicated by key supposition pioneers, a few to all on-screen characters required to characterize the basic
claims to fame could be completely replaced by AI gadgets, focuses for a reasonable type of clinical predictable with their
prompting an expert change for the influenced doctors, with qualities as we recognized right now.
regards to the adamant strengths which is pathology, and also
radiology [12]. Hence, some countries like France have a A. COMPARISON TO OTHER STUDIES
portion of these pros that are good and that they adapt to As indicated by the meaning given by Marvin Minsky, which
numerous fast changes like this, as in numerous different is known as the father of AI, it essentially implies that any
nations [13], [14]. Although the USA is topping the front line, machine can carry out a responsibility which is viewed as a
France, presently attempts to meet up with speed, going to the keen one by individuals. To be sure, request for AI is the one
massive measure of wellbeing information assembled by the in which the utilization falls in two classes: a) undertaking to
organization and open administrations at the National System replicate the capacities of a human cerebrum and b) arrange-
of Health Data situated in France [15]. However, measures ment of devices to do assignments that currently recommend
of information inside the framework continue expanding also human actions. Reproduced insight is detached into vari-
it as of now incorporates the information from repaid social ous sub-disciplines, concentrating on unmistakable issues,
insurance counting medical coverage information, the clin- for example, vision, critical thinking, language appreciation,
ical reason for death (CépiDC), handicap information got- learning, and so forth. Authors in [24] said a worldview of
ten at Independent-living support fund, with an example of research and a few parts of AI have become spots of multidis-
information by strengthening medical coverage associations. ciplinary trade where rationalists, therapists, PC researchers,
Moreover, after reports [16], the president of France, declared and other people who are keen on the different issues of AI
the formation of a ‘‘Wellbeing Data Hub’’ with a solid pur- can meet. Computer-based intelligence can likewise be com-
pose of France worldwide AI system. It is confided in out- prehended as an idea, for example, the common and also the
siders between wellbeing information makers, entertainers hypothetical view accepted the human cerebrum, and gives
wishing to utilize such information, and residents or common of a solid or extraordinary view of imagination that connects
help agents. The main strategic the center point will be to with it to relate the different discernments that it has of that
advance the social event of clinical information, which is view. During the Dartmouth Conference, the proposition of
the information gathered throughout care, in a mysterious a plan to design AI was put forth, as an order as well as an
way [17] [18]. Simultaneously, continuous problems come idea as stated by authors in [25]. Authors in [26] proposed
out of utilizing AI in human services, one of the principal a calculation called the Normalized Maximum Likelihood
one is the troubles with the utilization of wellbeing infor- (NML) technique, which consolidates substitution and mea-
mation, for example, information obligation, protection con- surable styles. In [27] authors proposed an improved AI
cerns [19], cybersecurity worries, is the subject of duty, also model of NML having capacities of parting the arrange-
the coordination of AI instruments into modern enforcement ment into fixed-size squares and encoding them utilizing
the historical backdrop of a subsequence and controlling trouble conditions and kicked the bucket of various organ
the equivalent with Hamming separation called GeNML. brokenness disorders respectively. In spite of the robustness
The author in [28] delivered another arrangement pressure of exploration in the area of COVID-19 advancement, appar-
calculation called Expert Model (XM) utilizing measurable ently, at the hour of this composition, there is no study that
strategies. In [29] the author created GenBit Compress, a cal- gives a thorough survey of the COVID-19 flare-up and its
culation that packs tedious and non-dreary successions utiliz- expected ramifications. Besides, no work in the current state
ing the thoughts of broadened twofold tree. LCA, a Lossless of the art endeavors to overview advances, specifically in
Compression Algorithm was given in [30] for dealing with AI, datasets, deep learning, and cloud while dealing with the
a few surmised rehashes and complimentary palindromes. pandemic. In Table 1, we highlight these main differences and
A calculation, DNASC, that joins factual and substitution limitations in comparison to this survey.
strategies, was proposed in [31]. The authors in [32] put
forth a technique using AI tools, were affected patients with TABLE 1. Comparison to other studies.
COVID-19 can be managed remotely based on self-isolation
and symptomatic management. With the fast improvement
of PC innovations, computerized picture handling tech have
been generally implemented in the clinical field, including
organ division and picture upgrade and fix, offering help
for ensuing clinical findings. In [33] [34], authors proposed
deep learning innovations, for example, convolutional neural
system (CNN) with the solid capacity of nonlinear displaying,
having broad applications in clinical picture preparing as
well. In [35] [36], relevant examinations were led on the con-
clusion of pneumonic modules, while in [37], [38] grouping
of benign and threatening tumors, [39] and aspiratory tuber-
culosis investigation and sickness expectation [40] around
the world. The authors in [41] have emphasized the impor-
tance of the COVID-19 computed tomography (CT) quali-
ties which can be accessible on several stages. Meanwhile,
the authors in [42] have looked into six distributed exam-
inations perceiving the clinical qualities of the COVID-19.
In their work, authors have summed up these examinations
and gave a concise review of clinical highlights and potential
medicines for the COVID-19. Authors in [43] and [44] give
a short outline of the COVID-19 episode in addition to its
clinical highlights, avoidance, finding, and treatment. The
essential issue with both of these works is that they audit a
subset of a lot more extensive subject. Despite the fact that
these studies shed some light on the current situation of the
COVID-19 flare-up, they give an extremely short and con-
strained thoughts regarding its specific circumstances from
the technical perspectives. Authors in [45] considered ninety- B. SCOPE OF THE SURVEY AND CONTRIBUTIONS
nine cases of the COVID-19, forty-nine of whom had an The models referenced above are regions where AI is applied
immediate connection to the Wuhan seafood market, known in checking infections like the COVID-19 which have made
to be the COVID-19 focal point. Their discoveries of the epi- extraordinary accomplishments. Corresponding to the studies
demiological, clinical, and radiological qualities of the illness referenced, for the effective utilization of new AI methods,
have been distributed. To decide the clinical qualities of the it is imperative to investigate and appreciate the troubles
COVID-19, authors in [46] have contemplated on about one related to the possibility of AI and infections. At present,
hundred and thirty-eight cases located in Wuhan. The authors we will show an overview of AI as a rule of general health and
have considered points of interest, in socioeconomics, signs put an accentuation on specific methodology and systems to
and indications, and clinical history of the considerable num- be kept into thought while overseeing difficulties. One of the
ber of patients to survey their cases cautiously. The creators major upsides about AI has been comprehensively examined
have likewise introduced the research facility and discoveries above in clinical organizations. Man-caused thinking can
of these patients to show the impacts of the SARS-CoV-2 utilize refined figuring to acquire abstracts from a tremendous
infection on various fundamental organs of the body. In their amount of human organization information, and a brief time-
discoveries, they report that among all the patients that were frame later utilizes the comprehension of data to help clinical
considered, 17% and 11% of cases showed intense respiratory implementation. Whereas, in like way be furnished in ideas
in the past years. They are being recognized with a picture, on their examination approach. One of the specialists in [53]
also hereditary and finally electrophysiological (EP) informa- employed in air transportation specialist, which talked on
tion, just because they consist of huge segments of unstruc- the human–PC communication (HCI). However, to him, even
tured paper writings, for example, medical jots that are not though HCI isn’t explicit to AI, scrutinizing the robotization
good for analyses. As an outcome, the comparing AI appli- of errands and cut-off points ought to be considered as one
cations center for first changing over the disorganized con- of the principal objectives of the mix of AI into social insur-
tent to machine-justifiable electronic medical record (EMR). ance apparatuses. He clarified that, right now, of the primary
However, for instance, authors in [51] utilized AI advances to thoughts is to designate it just when it’s fundamental. Without
remove phenotypic highlights from case reports to improve this, it is a danger of deskilling. Deskilling is when losing the
the finding exactness of the inborn irregularities. In [52] capability of the practitioner who doesn’t have the foggiest
features that Infervision’s of the application of AI limits idea how to do an assignment that he did before because he
the weight of the procedures by facilitating the judgments quit performing it to support the machine. The related AI
and observing of COVID-19. As an ever-increasing number classified learning techniques will be described below and
of sweeps is done then the calculation learns and improves more emphasis will be laid upon it in later sections. In table 3,
precision together with the infection. The estimation of AI we will show the comparable differences.
becomes an integral factor by decreasing the weight of clin-
icians in a situation, for example, the present COVID-19 TABLE 3. Differences between the AI learning techniques.
episode. Thinking about a quick development in the present
flare-up concerning the contamination of medicinal services
experts, the author in [52] calls attention to were the Infer-
vision AI implementation can help ensure practitioners. The
passing of clinical specialists, like the expert who noticed
about the infection, features the situation of clinicians on
the bleeding edge. In [52] the article states that man-to-man
medical clinic related transmission represented 41% of the
present cases in an investigation of patients at a hospital in
Wuhan. We likewise realize that over 1000 emergency clinic
practitioners are being affirmed contaminated. However, it is
the place Infervision’s AI implementation could assist. With
a CT lung check, the AI is intended to rapidly distinguish
injuries of conceivable COVID-19 pneumonia, in gauging
its capacity, form, and thickness, also analyze the change of
numerous lung sores on the pictures, which gives all quanti-
tative report that helps specialists to make a quick judgment.
While manually reading the CT output, this may take as long
as 15 minutes, whereas, AI will wrap up the picture in not
more than 10 seconds. Implementation of this innovation in
COVID-19 is not yet distributed into a companion inspected
diary.
A. AI LEARNING TECHNIQUES
A significant rebuke of clinical specialists is that the sort of
media perspective of AI was the discussion had nothing to do
regarding the sort of AI they were taking a shot at, which
has a considerably more explicit and thin definition. They
had all the earmarks of being sure about the advancement in
their examination, lamenting the too-moderate interpretation 1) SUPERVISED LEARNING
from research to rehearse, regardless of whether working with This is one of the most utilized procedures in the health
the social insurance industry was viewed as an approach to system and it is commonly entrenched. This learning method
quicken the interpretation. They additionally griped about the utilizes information in making precise desires and learns the
troubles prompted by enactment with regards to gathering mapping between the yield and its relating input while gath-
information for examines. As indicated by them, this was the ering analysis through the learning strategy in recognizing
best way to see their examination financing increment and things reliant on near characteristics. Various ways currently
along these lines permit successful improvement of AI and used in envisioning an outcome or the future outcome or
an assurance of its quality. They didn’t take an interest in the gathering to a great deal of needed states [54]. The ordinary
conversation about the concerns in general and concentrated procedures finished accurately currently is the SVM, NN,
and RC. In setting up these estimations, generally, the limit technique. K-means and also the autoencoder is the most
is being described in the medical system. It can induce the regularly known unsupervised technique. One of the most
best in association between the yield and data. By then, broadly perceived jobs of this learning technique in the
the cost work tells the practitioner how far it can be from the medical line is the quirk acknowledgment [58]. The data
accurate output, so this goes about as a data signal. The model made from the affiliation will start from equivalent scattering,
had the choice to arrange medical data with an exactness of if there is any sort of interference or any counterfeit read
up to 90%. All things considered with significant learning; from the affiliation data, this data will be hailed an exemption
it has become a jump forward framework right now in all and can without a lot of a stretch be hailed or seen. The
locales. Supervised learning can assemble an astute stage K-means is a common and generally utilized grouping mech-
for programmed observing and forecasting of COVID-19. anism in image processing [59]. This calculation requires the
A neural system can likewise be created to extricate the client to indicate the quantity of cluster k to be produced.
visual highlights of this ailment, and this would help in Like in [59] it was used to isolate clamor from PD flags
legitimate checking and treatment of the influenced people. paying little attention to conceivable few sub-classifications
It can give the most recent updates of the patients and answer inside the PD and commotion gatherings. Thus, with this
associated queries with COVID-19. The learning approach idea, it can be applied to CT scan images and other medical
can encourage the preparation of systems by considering applications as well including the COVID-19. In [60] the
the contribution layer as in [55]. An Xception, depthwise author put forth a novel structure for this learning model for
based CNN approach concerning convolution distinguishable opportunistic cameras that captures moving data from the
layers was proposed in [56]. It begins with two convolu- stream. Afterward, the neural system is utilized to foresee the
tion layers, trailed by depthwise divisible convolution layers, movement of the event streamed. This movement is utilized
four convolution layers, and a completely associated layer. to endeavor and to evacuate any movement obscured in the
Whereas in [57] it was used in classifying the bed posture streamed pictures. In figure 3, to elaborate more on this
utilizing various bed pressure sensors. With its capability and idea, we illustrate the applied learning method, where we
high-efficiency outcome, it can serve as a useful tool in com- denote the training attribute and the testing attributes obtained
bating the COVID-19. In figure 2, we illustrate the learning from the patient by the symbol X. In this particular scenario
method, where we denote the training attribute and the testing accuracy is not gained, the goal of the approach is mainly
attributes received from the patient with the symbol X, and the to reveal any fascinating samples that can be discovered in
outcome of interest by Y. The program is fed by a preparation the given information. And extra added/utilized information
dataset with a label (X) which relatively yields a value (Y), can assist in affirming or disconfirming these samples that it
and hence, the functionality will be determined according to finds.
this dataset. This functionality will be at that time utilized
for arranging new information to compare the yielded values,
with the suspicion that the new information complies with.
3) REINFORCEMENT LEARNING
The procedure stated is held to an utmost degree, it wears
2) UNSUPERVISED LEARNING an appearance as to how individuals investigate and learn
Rather than the above learning technique that uses checked through their ordinary calendar assignments. This learning
data, this learning technique uses no information signal and technique is neither of the two stated above; it is, consider-
no names. This technique is commonly used in finding cov- ably progressively, a kind of a crossbreed approach. In this
ered structures of given data and breaks it into near gather- technique, there is an alternate or single authority that makes
ings. It is mainly used for elucidating exhibiting and design sense of acceptable behavior in a circumstance with the end
distinguishing proof. This is a promising kind of estima- goal that they endeavor to increase their immense total prize
tion to achieve the general AI need, yet it is absent behind or score. At particular advances, where the current state is
not in the slightest degree like the above-discussed learning given, the customer picks the best action on the explanation
of a methodology, with this action the state of nature will pandemic, in addition to lives and financial harm, will be
change, and the practitioners assembles a prize sign in a brief awful if deep learning isn’t actualized sooner or later. Improv-
moment in the best case, or state later, in which this is an issue ing AI, one of the most encouraging information-analytical
referred to as delayed reward, which makes the practitioner to tools, to help diminish these vulnerabilities, is of utmost inter-
not choose the movement taken in the past that made way to est nowadays. Encouragingly, information researchers have
this prize [61]. The practitioner learns an optima approach, responded to the call. However, AI has not yet been effective
which is the mapping from the state to move, without going against COVID-19. The utilization of AI is hampered by the
before data on nature. The game plan is discovered only by absence of information, and by a lot of the uproarious and
performing clinical experimentation, or misuse and critical anomaly information. Defeating these imperatives requires a
examination. This model has the capacity to decide while uti- cautious harmony between information security and the gen-
lizing cooperative involvements and evaluative inputs. Unlike eral clinical concerns and increasingly thorough human-AI
the customary supervised learning strategies that generally communication. It is impossible that these issues are tended
depends on one-shot, thorough and supervised signals [62]. to be of much assistance during the current pandemic. Rather,
It handles successive dynamic issues with inspective, eval- deep learning may help with the pandemic. Gathering ana-
uative, and deferred inputs. Such particular highlights make lytic information on who is irresistible can be fundamental
the RL method a reasonable option for the growing ground- to spare lives and restricting the monetary ruin because of
breaking arrangements in an assortment of the human medi- regulations. Authors in [65] recommended utilizing deep
cal areas, where diagnosing choices or treatment regimens are learning to follow and to foresee how the COVID-19 ailment
generally described by a delayed and successive strategy [63]. can spread after some time and over space. For example,
This model empowers a specialist to learn viable procedures following a past pandemic, like the zika-infection, a powerful
in consecutive dynamic issues by experimental communi- neural system was created to anticipate its spread. Models,
cations with its conditions [64]. For example, the Markov for example, should be re-prepared utilizing information from
Decision Process (MDP) has been utilized in the exploration the COVID-19 pandemic. This is by all accounts happening
of hypothetical dynamic in stochastic settings and has been now. At different establishments, they proposed the utiliza-
utilized as a general structure to formalize the models’ issue. tion of calculations prepared to anticipate the occasional
MDP can be implemented in different structures relying upon flue, which are presently be re-prepared on new information
what data can be indicated from the earlier stage. In figure 4, from COVID-19. Authors in [66] say that the deep learning
we illustrate the assumed learning method, where we denote framework can rapidly and dependably name chest X-ray
the training attribute and the testing attributes obtained from from COVID-19 patients by classifying pneumonia as irreg-
the patient with the symbol X, and the outcome of interest by ular, which could help accelerate analysis amid a progressing
the symbol Y. Here, we need an accurate prediction. However, pandemic. Notwithstanding, these learning frameworks have
the obtained information is not matching with the ideal yields. classified the ailment accurately 87% of the time contrasted
Thus, the assumed mechanism needs to be boosted by picking with human services experts. Such frameworks additionally
the most accurate moves to make. are accurate 92% of the time, instead of 90% like the human
specialists. Authors in [66] propose a convenient analysis and
triaging for PUIs that can be critical for the control of rising
irresistible viruses such as the current COVID-19. Because
of the restriction of nucleic corrosive-based research, there is
a pressing need to search for quick elective techniques that
can be utilized by cutting edge human services personals for
rapidly and precisely diagnosing the infection. In the current
investigation, the authors were able to build up a deep learn-
ing program by dissecting agent CT pictures while utilizing a
FIGURE 4. Illustration of the reinforcement learning technique. deep learning-based strategy. In the following, we summarize
the application areas where deep learning can contribute to
combating COVID-19:
III. DEEP LEARNING IN COVID-19 • Early alerts and warning against COVID-19.
The utilization of deep learning can be valuable to perceive, • Early COVID-19 prediction and tracking.
analyse, anticipate, and clarify the COVID-19 infection, and • Early prognosis and diagnosis.
help in keeping up financial effects. Since the flare-up of the • Social control.
pandemic, there has been a scramble to utilize and investigate • Early cure and treatments.
AI, and other information-analytical instruments, for these In the medical image analysis, deep learning algorithms can
reasons [65]. It intends to draw speedy take-away from a categorize, identify, and list the characteristics of diseases
quick extending conversation and develop a collection of from medical images. It also allows the extension of theoret-
huge data repository for research, medical arrangements, ical targets and creates predictive treatment mechanisms for
and clinical examinations. The expense of the COVID-19 patients. It employs the deep algorithms of AI with big data.
In terms of clinical operations, the complete use of these Big 5) Another restriction of deep learning in its present struc-
Data technologies would significantly reduce the experienced ture is its reliance on human information. Human skill
medical expenses. Applying a deep learning algorithm such is principal to control the execution of the deep learning
as the Convolutional Neural Network (CNN) in diagnosis methods and have an effective influence on the fight
can help the specialists in setting the best treatment plan. against the COVID-19 pandemic [68].
Authors in [66] released their findings for the detection of Regardless of the above-mentioned difficulties, the deep
strabismus by using CNN. A CNN model has been trained learning frameworks commitment to the battle against the
based on a collection of data gathered by an eye motion COVID-19 pandemic can’t be ignored. As of late, deep learn-
tracker, including visual data from both normal and stra- ing innovation has made shocking advances in the medical
bismus visions while considering gazing deviation (GaDe). field. Such improvements serve to demonstrate the capability
After training with several GaDe images, the CNN model was of deep learning in helping the COVID-19 pandemic admin-
successful in identifying strabismus. The emerging Internet istration sector.
of Medical Things (IoMT) paradigm, which relies on intelli-
gent medical systems, can also track significant human health IV. COMMUTING AI TECHNIQUES IN COMBATING
issues continuously. In this paradigm, robots are providing COVID-19
help to elderly and disabled people via their smart homes Simulation preparation has developed exponentially around
and virtual assistants. Epidemic diseases can be also tracked the globe in recent years. This is in huge part connected
and avoided by integrating the data collected with the IoMT with the ordered residency preparing arrangement for doctors
sensors and the information obtained from the associated which started since the year 2015. In any case, the significant
health system [66]. focal point of those re-enactments was to meet the instructive
needs of various preparing programs or to enhance certain
A. DEEP LEARNING DIFFICULTIES educational programs [69]. In the wake of recognizing the
This model can assume a fundamental role in relieving the requirement for preparing, recreation has become an incredi-
effect of the COVID-19 pandemic. Moreover, at present, deep ble weapon battling against the virus, as it can guarantee quiet
learning is still in the prefatory stages. Difficulties and imped- security as well as giving a sheltered learning and preparing
iments towards the utilization of deep learning in fighting the condition for HCWs to manage the COVID-19 [70]. In [71]
COVID-19 can be listed as follows: Boltzmann function-based regression analysis was proposed
1) In yielding a dependable and precise outcome, the deep in simulating and forecasting the SARS-CoV-2 virus. This
learning model requires a generous measure of the method was applied to the cumulative confirmed cases in
training information. In any case, inferable from the the province of Wuhan. Boltzmann function gives a direct
remarkable idea of the pandemic, there is a shortage of estimate of potential confirmed cases. In addition to the
recorded information on which to train a deep learning Boltzmann function, it doesn’t require much more detailed
model, which has subsequently rendered the model data for analysis. In [72] [73], the cumulative cases are only
wasteful [67]. needed. Also, it allowed the forecasting of future courses
2) It isn’t just the nonattendance of open information that in other cities. However, the estimates of the results are not
has blocked the exhibition of the deep learning model, guaranteed due to uncertainty in the data being reported.
an excessive amount of irregular and anomaly informa- However, machine learning techniques can be applied to
tion has additionally introduced a test to the produc- provide an accurate predicting method with the provided data
tive utilization of the deep learning advancements. The set. Pattern recognition can also be suggested in recognizing
Trends’ of Google Flu bombed activity reveals insight similarities in other infected hosts. In [74] the virus is similar
into the way that a downpour of information hubris to the SARS virus due to an intermediary host from bat to
can immerse deep learning algorithms and restrain their humans. The civic cat was said to be an intermediary host
working [67]. between bat and man. While the SARS-CoV-2 virus can
3) Another restricting challenge, is the intrinsic suspicion highly be related to that of SARS, with a high percentage
that every single imaginable possibility in some ran- of the bat sequence, pattern matching can be carried out in
dom circumstances are equivalent to the ones showed studying the sequences of animals situated in the seafood
in the dataset they have been prepared on [67]. market of origin. This is a benchmark for us. Not long before
4) The utilization of deep learning for mass observation the New Year, we took a rollcall at what number of individuals
has been also seen by numerous individuals as a vio- left Wuhan longer than a day, and this data originates from
lation of protection. Although at present, individuals clinical information. The clinical research gives information
have caught the way that general medical concerns are on movement around Wuhan, these clarify the usage of AI
a higher priority than information security concerns, models which utilizes this information to anticipate the most
the protection traps related with the utilization of deep probable area of where novel coronavirus may show up
learning have ingrained a feeling of dread among the straightaway and this may illuminate where and how to run
public that administrations may keep observing them fringe checks. This additionally vouches for the quality of
significantly after the pandemic is over [68]. observing news reports and web-based life to help recreate
the movement of an episode and to give point by point quiet point, there will be a high possibility of miss-characterization.
level information with regards to the health crisis. In the sub- A significant property of this method is that it gives the
sections below we will discuss this individual implementable assurance of the used model parameters, and thus the arrange-
technique. In table IV we discuss the difference between the ment is constantly ideal. Besides, many existing enhancement
applications of the techniques. instruments are promptly appropriate for the execution of this
method. In that capacity, this method has been widely utilized
A. SUPPORT VECTOR MACHINE in clinical research. For example, in [76] it was applied to
This strategy is predominantly utilized for arranging the sub- distinguish imaging biomarkers of neurological and mental
ject into two states or groups, whereas the result Y is regarded sickness. In [77] it was evaluated to be utilized in the determi-
as a classifier: were Y = 1 or −1 speaks to if the nth patient nation of malignant growth. In [78] it was utilized to mix this
is in the first or second state, separately. This technique is method with other measurable devices to accomplish early
stretched out for situations with multiple states. The fun- discovery of Alzheimer’s sickness. In [79] it was utilized
damental supposition that will be, that the subjects can be in testing the intensity of a disconnected machine or man’s
isolated into two gatherings through a choice limit which is interface which controls the upper appendage prostheses.
characterized on the trait X, The preparation objective is to
locate the ideal weight so the subsequent arrangements that B. NEURAL NETWORK
concur with the results however much as could reasonably be With this technique, it very well as an augmentation of direct
expected, that is, with the littlest misclassification blunder, relapse to catch complicated non-linear connections admits
the mistake of ordering a patient into an inappropriate state. input factors and a result. In the neural system, the relation-
Instinctively, the best loads must permit a) The indication of ai ship admits the result, and also the info factors are delin-
to be equivalent to Y therefore the grouping is right, and (b) ai eated by various shrouded layers with a mix of pre-indicated
most be distant away from 0 so the vagueness characterization usefulness. The objective is to appraise the loads through
will be limited. Were ai is gotten by a choice standard which is information and result information with the goal that the
if ai > 0, the nth patient is arranged to state 1. These can be normal blunder between the result and their forecasts is lim-
accomplished by choosing loads that limit a quadratic mis- ited. We depict the technique in the accompanying model.
fortune work [75]. Besides, accepting that the new patients In [80] it was utilized in stroke finding. In their examination,
originate from a similar populace, the subsequent loads may the info factors Xi... Xp is given by p=16 with stroke-related
be implemented to arrange these latest patients dependent manifestations, which includes paranesthesia of either the
on the patient’s characteristics. Figure 5 illustrates the SVM leg or hand, intense disarray, sight, issues with portability,
method. Here, we present the hyper-plan A, B, and C, and all and so on. The output Y is twofold: Y=1/0 demonstrates
are isolating the classes well. Presently, that hyper-plan can the nth patient is not diagnosed with a stroke. However,
be distinguished effectively. By expanding the separations the yield criteria of intrigue are in the likelihood of the cause
between the closest information point and the hyper-plan, of stroke that is ai . Moreover, the preparation objective is
it will assist in choosing the privilege plan. The margin is to discover the load’s weight, which limits the expectation
referred to as the distance between them. Another lightning blunder. The minimization can be performed through stan-
explanation behind choosing the hyper-plan with a higher dard improvement calculations, for example, neighborhood
edge is the robustness (e.g., the hyper-plan C). On the off quadratic estimation or slope plummet streamlining that is
chance that we select a hyper-plan having a low edge, at that remembered for both MATLAB and R. If the new information
originates from a similar slope plunge advancement, that is
remembered for both MATLAB and R. If the new information
originates from a similar populace, the subsequent weight
can be utilized to anticipate the results dependent on their
particular characteristics.
Comparative strategies have been utilized to analyze
malignancy like in [81], where the sources of information
are the PCs assessed from 6567 qualities and the results are
the tumor classifications. In [82] it was utilized to anticipate
bosom malignant growth, with the sources of info being the
surface data from mammographic pictures and the results
being tumor markers. While in [83] it was utilized a pro-
gressively modern model of the neural system to analyze
Parkinson’s sickness dependent by the contributions of motor,
also non-motor problems and neuroimages. Figure 6 gives
an illustration of the neural network and its utilized hidden
layers, the X’s are inputs, and W’s are weights per neuron.
FIGURE 5. SVM illustration. The output Y is calculated based on utility function f as
TABLE 4. Differences between the above methods. model probabilities in distinguishing proof for melanoma and
seborrheic keratosis classes for skin class. Authors in [99] uti-
lized the AlexNet tool and google tools for convolutional neu-
ral systems and joint model yield probabilities for melanoma
order, this can also be proposed for fluorescence polymerase
chain reaction for the COVID-19. Four distinct strategies
lead to the last probabilistic polymerase chain forecast. The
combination systems are the maximal probabilities, product,
sum and the larger part casting a ballot. This model outflanks
single deep learning strategies like AlexNet, GoogLeNet, and
VGGNet. The authors in [100] utilized convolutional neural
systems for classifying skin sore with tools like DenseNet,
ResNet, and Inception V3. They anticipated skin injuries
in mind utilizing these procedures for other possible dis- utilizing every one of these neural models, and again assessed
eases. The deep learning model was trained by utilizing sore sorts by combining yield evaluations of every one of
about 129,450 clinical pictures by the authors to perceive the models. The examination presents two preparing systems.
noncancerous skin growth called seborrheic keratosis, ker- One of them is to prepare the models utilizing unique pictures.
atinocyte, and carcinomas basal cells, nevi tumors, and Another is to prepare the models utilizing edited skin pictures
life-threatening malignant tumor. Authors in [95] convey a around the locale of intrigue. However, in [101] the authors
different type of method for the classification of the proceed- produced deep learning models and combined these models
ings of the COVID-19. They put together their method con- for forecasting skin injury with tools like DenseNet, Incep-
cerning a deep learning fine-tuned ResNet-152 model. The tion V3, and ResNet. The creator’s combination approach
made method permits the classification of pneumonia and prompted higher exactness of forecast than single neural net-
other stated conditions as above. However, in [73] authors put work models. In [102], for chest computed tomography loca-
forth a deep learning method for detecting the 2019-ncov out- tion, a joint two ResNet models were utilized with the above
break by utilizing ResNet-101 and Inception-v4. The authors method in mind. The creators put together their methodol-
perceived various traits of pneumonia in their findings. They ogy concerning convolving pictures utilizing a 3-dimensional
utilized a few outer datasets for making deep learning models. model and afterward consolidating saliency maps of the two
Authors in [96] also utilized google tools for characterizing models. Finally, they assessed the chest images utilizing joint
skin sore by using convolutional neural method. To begin yield saliency maps. Just like in [103] were the authors
with, this strategy assembled harmful and non-harmful sorts perceive melanoma among seborrheic keratosis and nevus
of skin injuries. Secondly, the technique further partitioned they utilized tools like two ResNet models. In the first place,
threatening and non-threatening sorts into epithelial and the creators utilizing every one of the ResNet models by antic-
melanocytic states. In like manner, they further isolated each ipating melanoma. The presentation results have indicated
gathering into more subcategories. Subsequently, they could that the two ResNet models outflanks every one of the single
perceive skin injuries of up to twenty-one classes. While ResNet models.
in [97] with the help of the above-mentioned technique the
authors utilized deep convolutional neural systems for classi- 3) FEATURE-BASED DEEP LEARNING METHODS
fying CT scans with tools like Inception v3, DenseNet-210, In [104] the authors include the extraction and afterward
and ResNet-152. These models identify chest diseases, basal classifying these traits utilizing bolster vector machines with
cell carcinoma, and also like the one stated above. How- tools like AlexNet, and VGGNet. The creators in like manner
ever, the ResNet-152 models outflank different models for combined separated highlights for skin sore acknowledg-
melanoma and melanocytic nevi identification. The Inception ment. This examination shows that melded highlights for sore
v3 model beats different models for actinic keratosis and expectations beat the qualities of every one of the deep learn-
favourable CT scan arrangement. DenseNet-210 gave the ing models. In [105] it was utilized with various component
best outcomes to basal cell carcinoma and dermatofibroma pictures for recognizing CT imaging features of the novel
arrangement. The majority of these tools have been utilized coronavirus. The outcomes have indicated that intertwining
in combating the COVID-19 but it is still not effective because handcraft and deep highlights gave higher precision in the
of the lack of the large dataset. images. In [106], in figuring out how to perceive dangerous
melanoma and seborrheic keratosis skin injuries the authors
2) ENSEMBLE DEEP LEARNING METHODS proposed a deep learning based on Gabor wavelet. The
In [96] authors utilized a gathering of convolutional neural technique represented the skin picture into seven directional
system tools called ResNet for perceiving fever-like cough for Gabor band and afterward displayed every one of the portray-
the COVID-19. This methodology was put forth concerning als utilizing the deep learning tools like AlexNet, In [105] The
creating a few deep models for ResNet. This tool expands on examination indicated that COVID-19 commonly shows on
various expansion information. They melded every one of the CT with respective ground-glass and consolidative aspiratory
opacities. Nodular opacities, insane clearing design, and a whether social removing measures are working. Dissimilar to
fringe dissemination of illness might be an extra highlight reconnaissance quantifies that track the developments of spe-
in early conclusion. The analysts likewise noticed that lung cific people, these endeavors break down huge informational
cavitation, discrete pneumonic knobs, pleural emissions, and collections to reveal designs in individuals’ developments
lymphadenopathy are typically missing in instances of 2019- and conduct through the span of the pandemic. Specialists
nCoV. and practitioners are progressively utilizing man-made brain-
power, AI, and characteristic language preparing to follow
D. NATURAL LANGUAGE PROCESSES and contain coronavirus, just as addition an increasingly
Pictures, hereditary information, and EP are machine- exhaustive comprehension of the illness. In the months since
justifiable with the goal that the AI calculations can be legit- COVID-19 hit the world, analysts have been working dili-
imately performed after appropriate pre-handling or quality gently attempting to reveal the idea of the infection, like why
control forms. Notwithstanding, huge extents of clinical data it influences some more than others, what measures can help
are story content just like, physical assessment, reports from diminish the spread, and where the ailment will probably go
the clinical lab, useful notes, and also release outlines that straightaway as well as the likes of skin lesions. Scientists
are not structured and immense for PC programs. With this are utilizing Big Data and examination to all the more likely
unique situation, this method focuses on removing valuable grasp coronavirus from various edges. The foundation as
data from the story content which helps medical dynam- of late reported that it would offer government substances,
ics [107]. The methods pipeline consists of these principal look into associations, and industrial access to inventive AI
parts: (a) preparing content and (b) order. Through content apparatuses, just as specialists in information and general
preparation, the NLP recognizes a progression of malady wellbeing help to battle COVID-19 [112].
applicable catchphrases from the medical notes dependent on No nation knows the complete number of individuals con-
its verifiable databases. At that point, the subsets catchphrases taminated with COVID-19. The known fact is the contam-
are chosen by analyzing their consequences for the order of ination status of the individuals who have been tried. Each
the typical and unusual scenarios. The approved catchphrases one of the individuals who have a lab-affirmed disease is
at that point enter and advance the organized information considered affirmed cases. This implies the tallies of affirmed
to help clinical situations. This pipeline was created to help cases rely upon how much a nation test. Without testing,
the clinical situation on cautioning medical care courses of there is no information. Testing is our window onto the
action, checking antagonistic impacts, etc. For instance, [108] pandemic and how it is spreading. Without information on
indicated that the presentation of this method for utilizing the who is contaminated by the infection, we have no chance to
reports from a chest X-beam can assist the said anti-toxin get of understanding the pandemic. Without this information,
relative framework in alarming practitioners for conceivable we cannot know which nations are progressing nicely, and
needs for hostile to infective treatment. Authors in [109] which are simply underreporting cases and passing. To deci-
utilized this method in screening the research center based pher any information on affirmed cases we have to know
on unfavorable impacts. Besides, the pipelines can assist in how much testing for COVID-19 the nation does. Like the
sickness analysis. Just like, [110] distinguished 14 cerebral author in [113] was able to get a total of 618 transverse-
aneurysms sickness-related factors through executing this section CT samples which includes 219 from 110 patients
method based on medical notes. However, the subsequent with COVID-19 as the dataset of the infected hasn’t being
factors are effectively utilized for characterizing the cerebral released to the public. Table 5 shows the accessible datasets
sickening patient and the normal patient, given with 85% and in writing so far related to their references concerning other
96% precision values by the preparation and approval tests, viruses.
individually. Creators in [111] executed the NLP to remove TABLE 5. Relevant dataset.
the fringe blood vessel malady related watchwords from story
medical notes. Thus, the watchwords, afterward are utilized
to order the typical and also patients with fringe blood vessel
ailment, which accomplishes over 91% accuracy.
V. DATASETS
The COVID-19 has prodded enthusiasm for large information
to follow the spread of the quick-moving pathogen and to
design ailment anticipation endeavors. Yet, the critical need to
contain the episode shouldn’t cloud pondering huge informa- A. BATTLING COVID-19 WITH A COMPREHENSIVE
tion’s capability to accomplish more damage than anything DATASET
else. Big information has been demonstrated with the abil-
Organizations and governments overall are tapping the ity to help battling irresistible illnesses like COVID-19
area information of a large number of web and cell phone [116], [117]. Huge information conceivably gives various
clients for intimations about how the infection spreads and promising answers for helping the battle against COVID-19
pandemic. By consolidating with AI investigation, the dataset and 1s as 1s. Similarly, the higher the AUC, the better the
encourages us to comprehend the COVID-19 as far as flare-up model is at recognizing patients with illness and no infec-
following, infection structure, sickness treatment, and immu- tion. The framework proposed by [127] utilizing Riesz and
nization fabricating [118], [119]. For instance, enormous Gabor changes acquired a 98.73% accuracy in distinguishing
information related to advance AI-based devices can man- lunge illness. It expanded to 99.53% accuracy with the mix
ufacture complex reenactment models utilizing coronavirus of Riesz, Gabor, fractal measurement, dark level co-event
information streams for flare-up estimation. This would help grid, and dim level run-length framework-based highlights.
medical offices in observing the coronavirus spread and Authors in [128] demonstrated that a deep CNN could dis-
getting ready for better preventive estimations [119], [120]. tinguish lung knobs with an opposition execution metric of
Models from huge information likewise underpins future 0.7967. They also put forth a thickness based spatial bunching
expectation of COVID-19 pandemic by its information accu- with clamor DBSCAN calculation to improve the location
mulation capacity to use a lot of information for early iden- affectability of the system. Likewise, the CAD framework
tification. Also, the dataset examination from an assortment dependent on a 3D skeletonization was proposed by [129].
of certifiable sources including tainted patients can help exe- This can help radiologists to separate lung knobs from inter-
cute enormous scope for COVID-19 examinations to create ferential vessels. Changing a picture to the time recurrence
far-reaching treatment arrangements with high dependabil- area by the optimal fractional S-transform (OFrST) strat-
ity [122], [123]. This would likewise help social insurance egy can give important data and highlights, which can be
suppliers to comprehend the infection improvement for better utilized to analyze ailments. This technique was applied to
reaction to the different treatment and conclusions. lung CT pictures by [130] to identify and separate knobs
from the vessels. In [131] they utilized the Teager–Kaiser
B. DATASETS AND THE SPREAD OF COVID-19 vitality (TKE) in the time–recurrence area to get the vitality
Another task of the dataset is to follow the COVID-19 spread, circulation and describe lung knobs with a sensitivity of
which is of principal significance for human services associ- 97.87%. The author reasoned that CNN was better than a
ations and governments in controlling effectively the coro- deep belief network (DBN) and stacked denoising autoen-
navirus pandemic [124]. Various recent rising arrangements coder (SDAE) in diagnosing threatening lung knobs with an
utilizing large information have been proposed to help fol- AUC of 0.899. Moreover, CAD frameworks are a valuable
low up the COVID-19 spread. For example, the examination device to screen lung work after transplantation. With this,
in [125] recommended a major information-based system a framework dependent on deep learning on how to arrange
for following the COVID-19 spread. A numerous straight COVID-19 contamination versus another typical and viral
model is fabricated utilizing neighborhood populace and air pneumonia sickness can be proposed. An estimated hypoth-
travelers as evaluated factors that are helpful to measure the esized deep learning would help radiologists in diagnosing
fluctuation of revealed cases in China urban areas. to be contamination identified with COVID-19. A convolutional
more explicit, the creators utilized a Spearman relationship neural system (CNN) to analyze diseases identified with
examination for the day by day client traffic from Wuhan the COVID-19 end up being proficient. The zone under the
and the all-out client traffic in this period with the quantity collector working trademark (ROC) bend (AUC), exactness
of 49 affirmed cases. The explanatory outcomes show a (ACC), affectability (SE), and explicitness (SP) execution
high relationship between the positive disease cases and the merits are utilized to test the precision of the strategies [132].
populace size. The creators in [126] thought about utilizing
both enormous information for spatial investigation strategies A. ACCURACY
and Geographic Information Systems (GIS) innovation which The accuracy of a strategy decides how right the qualities are
would encourage information obtaining and the joining of anticipated. The score utilizes a blend of accuracy and review
heterogeneous information from medical information assets, to compute a fair normal outcome. Eq. (1) tells the best way
for example, governments, patients, clinical labs and the to compute this exactness, where TP, TN, FP, and FN are gen-
general population. uinely positive, genuine negative, bogus positive, and bogus
negative individually. In [132] the creators show the exactness
VI. PERFORMANCE EVALUATION level of up to 90% when contrasting CT examination with that
In deep learning, performance assessment is a basic task. of the COVID-19 patients.
With this, with regards to CNN models, we can depend on
TP + TN
an AUC - ROC Curve. At the point when we have to check or Accuracy = (1)
imagine the exhibition of the multiclass order issue, we uti- TP + TN + FP + FN
lize AUC and ROC bend. It is one of the most significant Among all systems, the best performance was accomplished
assessment measurements for checking any characterization by ResNet-101 and Xception. ResNet-101 could recognize
model’s presentation. The ROC is a likelihood bend and AUC COVID-19 from non-COVID-19 cases with an AUC of
represent the degree or proportion of detachability. It tells 0.994. Xception accomplished an AUC of 0.994. Be that
how much model is fit for recognizing classes. The higher as it may, the presentation of the radiologist was moderate
the AUC curve, the better the model is at anticipating 0s as 0s with an AUC of 0.873. ResNet-101 can be considered as a
high affectability model to portray and analyze COVID-19 variety of wealth as we endeavor to supplant our physical
diseases and can be utilized as an adjuvant device in radiology real factors with virtual devices that make a computerized
offices. change of our social associations and everyday lives. On the
off chance that one innovation had the option to spark a bril-
B. SENSITIVITY liance amid office termination, isolation, social separation,
Sensitivity shows what number of the right outcomes are and uneasiness, it’s distributed computing. Cloud adminis-
found. Eq. (2) tells the best way to figure this precision, trations have empowered us to proceed with our advanced
where TP and FN are genuinely positive, and bogus negative lives through different applications. Nonetheless, many cloud
separately. In any case, in [132] the affectability was at its suppliers are helping the battle against COVID-19 in a variety
top with a 90% on second examination, while contrasting CT of ways. As of late a business association propelled its first
filter with that of the COVID-19 patients. The deep learning Salesforce Care answer for Healthcare Systems, structured
models utilized have similarly demonstrated high location explicitly for human services suppliers encountering an inun-
results with ResNet-152 design being the best of them. dation of solicitations due to COVID-19 [133], [134]. They
have additionally extended deals with extra free in helping
TP
Sensetivity = (2) organizations in any industry remain associated with part-
TP + FN ners, in any event, when everybody is working remotely.
The best execution was accomplished by ResNet-101 and The consideration arrangements are accessible promptly and
Xception. ResNet-101 could recognize COVID-19 from can be set up rapidly. A significant association has offered
non-COVID-19 cases with a sensitivity of 100%. Xception AI-controlled distributed computing stage and supercomput-
accomplished a sensitivity of 98.04%. Be that as it may, ing bunch for nothing to control worldwide research founda-
the presentation of the radiologist was moderate with an tions to quicken viral quality sequencing, protein-screening,
affectability of 89.21%. ResNet-101 can be considered as a and other research in treating or forestalling the Coronavirus.
high sensitivity model to describe and analyse COVID-19 Elite Computing Consortium has been sorted out to unite
diseases and can be utilized as an adjuvant apparatus in offices around the globe and different offices and scholarly
radiology divisions. pioneers to give access to the world’s most remarkable supe-
rior figuring assets on the side of COVID-19 research [135].
C. SPECIFICITY The consortium will deliver an exceptional measure of pro-
Specificity decides the reproducibility of the estimation or cessing power for the research facility, which is the quickest
what number of the forecasts are right. In order words, supercomputer on the planet. Likewise, with this framework,
it gauges the extent of negatives effectively ordered. Eq. (3) it permits analysts to run huge quantities of computations
tells the best way to ascertain this precision, where TN and in the study of disease transmission, bioinformatics, and
FP are genuine negative and bogus negative individually. atomic displaying. This makes tests conceivable that would
Notwithstanding, as opposed to the affectability the explic- somehow, or another take a long time to finish whenever
itness was at a top with a 91% imprint by creators in [132], worked on more slow, customary registering stages. Cut-
while contrasting CT filter with that of the COVID-19 ting edge advancements are probably going to get the most
patients. approval – particularly by the media. In any case, new busi-
nesses have been faithful clients of cloud suppliers [136].
TN
Specificity = (3) They ought to be bolstered in their endeavors to remain above
TN + FP water. They may not all be as alluring as supercomputers,
It was understood that ResNet-101 and Xception were the yet they give employments and administration their nearby
best. ResNet-101 could recognize COVID-19 from non- economies. Their capacity to keep on supporting their client
COVID-19 cases with a specificity of 99.02%. Xception base using cloud administrations will be a crucial piece of
accomplished a specificity of 100%. Be that as it may, their recuperation. We’re in phenomenal occasions all around
the exhibition of the radiologist was moderate with an explic- and this is an open door for huge tech to step up and bolster
itness of 83.33%. ResNet-101 can be considered as a high their customers and lead the path in helping organizations
specificity model to portray and analyse COVID-19 con- endure. Some of the cloud services are discussed below [137].
taminations and can be utilized as an adjuvant apparatus in
radiology divisions. A. VIRTUAL NETWORK
As we probably are aware, a virtual system associates
VII. INCORPORATING CLOUD COMPUTING IN administrations and assets like virtual machines and database
COMBATING COVID-19 applications with one another and the remainder of the web
The previous months have been one of segregation, both through a safe, encoded, and private system. This mod-
socially and genuinely. In an offer to forestall COVID-19 ule gives a structure that gives your undertaking cloud
from spreading, our physical universes have become foundation substance. Virtual frameworks organization sup-
geofenced to envelop our home and short neighborhood ports the association of varying organizations and devices
strolls. Accordingly, our virtual lives have woken up in a on a single hardware called a virtual organization switch.
The centralization of control lessens the cost and unusualness administrators, or Internet authority affiliations (ISP) may
of working and keeping up gear and programming differenti- have obvious space names, IP addresses, email affiliation,
ated and controlling different separate devices in for the most record lists, logs, and assessment. Also, security frameworks
part confined land territories. Upkeep staff and directors can and passwords are kept up as though they were in a submitted
present device drivers, perform tests, and resolve issues on server condition. Also, to diminish Web empowering costs,
the remote machines from a singular zone. It might be basic server programming establishment provisioning is a signif-
to introduce virtual systems association programming on the icant part of the time accessible. A surge of virtual servers
remote PCs or servers to abuse this progression. A few dealers in a physical machine may impel the resource being covered
offer virtual systems association programming. Two or three and if a virtual server uses a more noteworthy number of focal
merchants offer wide virtual structures association associa- points than another, the execution it gives can take everything
tions, permitting businesses to sort out heads to redistribute into account. At present, the virtual part just infers that it’s not
work and assets for the vendor. With this, professionals can a submitted server, that is, the entire PC can’t run the server
in-coordinate their thoughts for all intents and purposes with programming. Virtual Web servers are an incredibly notable
the assistance of the virtual system in fighting the COVID-19 strategy for giving simplicity web encouraging organizations.
issues [138]. As opposed to requiring an alternate PC for each server,
The central restriction of virtual frameworks organization numerous virtual servers can co-live on a comparable PC.
is the way that particular issues can be settled unmistakably When in doubt, execution can’t and each site goes about just
by direct physical contact with the hardware being referred as it is being served by a committed server. Regardless, if an
to. Models join broken wires, set drives, and imperfect chips. unreasonable number of virtual servers harp on a comparative
Occasionally, virtual frameworks organization may make PC, or if one virtual server starts storing resources, Web pages
it possible to circumvent a discomfort by abusing elective will be passed on even more bit by bit [141].
resources until an expert who can fix it or substitute it. Virtual
systems administration depends on physical PC organizing C. VIRTUAL STORAGE
standards; however, its capacities are generally programming Virtual storage alludes to the virtualized type of a capacity
driven. The VMs impart by tending to the predefined IP medium, as such it exists as a build inside a virtual domain.
address of every goal virtual machine. Essentially, a virtual This goes about as a deliberation between the client and
neighbourhood is made through programming based virtual the genuine stockpiling equipment. Virtual storage is a sta-
switches that give a correspondence between all virtual and ple element of distributed computing and is accessible for
associated machines. Virtual systems administration addi- the most part as online stockpiling or reinforcement. Vir-
tionally might be executed on virtual machines that are intro- tual storage used to be synonymous with virtual memory,
duced or sent on Internet-empowered physical servers [139]. which was an augmentation of the fundamental memory gave
through optional stockpiling. In any case, with the appearance
B. VIRTUAL SERVER of distributed computing, the term has gotten progressively
This is most occasions alluded to as a virtual machine, this exacting, just importance stockpiling that has been made in
is a type of virtualization in the cloud that mimics a physical a virtual domain. A virtual stockpiling medium or gadget is
server, and which is managed by cloud suppliers in having the typically likewise connected with virtual machines. A virtual
equivalent server with various cloud buyers in giving cloud machine appears to the client as an undeniable PC with a
purchasers single mimicked virtual server. The quantity of working framework and all the fancy odds and ends, and
examples a given physical server can share is restricted by its obviously with its very own capacity drive [142], [143].
ability. As an aware system, the virtual server speaks to the Storage virtualization is a framework organization practice
most central structure square of cloud situations. Each vir- that permits decoupling the physical association of the equip-
tual server can have various IT assets, cloud-based arrange- ment from its intelligent portrayal. Utilizing this method,
ments, and different other distributed computing systems. clients don’t need to be stressed over the particular area
The launch of virtual servers from picture documents is an of their information, which can be distinguished in utiliz-
asset distribution process that can be finished quickly and ing a sensible way. Storage virtualization permits us to bri-
on-request. Cloud shoppers that introduce or pay by service dle a wide scope of storerooms and speak to them under
virtual servers may redo their environment autonomously a solitary legitimate record framework. There are various
from various cloud customers that may be making use of systems for storage virtualization, one of the most famous
the virtual servers facilitated by the equivalent fundamental being system-based virtualization by methods for storage ter-
physical server [140]. ritory systems. This model utilizes a system available gadget
However, a virtual server copies and gives server func- through an enormous transfer speed association with give
tionalities. In place of a completely distinctive submitted storerooms [144].
server, two or three virtual servers may be carried out on
a single server. An individual virtual server is consigned to D. VIRTUAL MIDDLEWARE AND APPLICATION
another OS, programming, and autonomous reboot provi- Middleware is a product stage that sits between an appli-
sioning. A virtual server that is for Web empowering, page cation/gadget and another application/gadget. It makes the
association between any two customers, servers, databases, clinical notes. In this way, a short time later they request the
or even applications conceivable. Cloud middleware is con- patients with periphery vein infirmity, which achieves over
stantly available to the client as a remote programming stage 91% precision. However, more future researches are to be
for correspondence or the board of information. Ordinar- carried out to guarantee that everybody imparts and works
ily, it is arranged between the working framework and an together in a manner that abstains from missing any basic
application, cloud middleware gives various functionalities focuses. Moral contemplations will assume a huge job by
to the client. It helps in the formation of business appli- assisting us to go around potential snags in the reception
cations, encourages simultaneousness, exchanges, stringing, of artificial intelligence instruments in helping the circum-
and informing and offers an assistance part engineering struc- stance. Along these lines, the rest of the inquiries that despite
ture for making administration situated design applications. everything destroys them, for example, the question of the
Web servers, application servers, and databases are instances sharing of duties, should be tended to. We should join a
of cloud middleware. Middleware programs give correspon- discussion among all partners worried, no matter what. In this
dence administrations and fill the need of a detachment with way, it appears to be even more imperative to concentrate
the goal that various applications can send and get messages. on quiet voices. Computer-based intelligence isn’t just useful
Various applications arranged at various physical areas can in the treatment of COVID-19, nut additionally, for their
be tied together to play out an undertaking through cloud appropriate medical check-ups. Future inquiries about this
middleware [145], [32]. can be considered to follow the emergency of COVID-19 at
various scales, for example, clinical, and epidemiological sit-
VIII. CONCLUSION uation. It will be useful to encourage the examination of this
Typically, the artificial intelligence tool is arriving at the infection by utilizing dissecting the accessible information.
clinical field in present times. It is presently a reality that Likewise, in creating appropriate treatment, protective pro-
we should face to encourage their appearance. In this sur- cedures, medication, and immunization advancement. More
vey, we discuss various AI techniques that help in speeding experiments and surveys are to be carried out concerning the
up researches and assisting in the current COVID-19 crisis. big data gathered from each patient, utilizing this data can
Also, various learning techniques were emphasized. Cloud reduce the spread or if worth saying finding a solution to this
computing plays a vital role in virtualization since everyone pandemic. Further work will along these lines need to address
is in isolation. We discuss various areas in which cloud com- their desires with the goal that the improvement of AI is to
puting can assist in concurring with this current pandemic. serve patients and not despite them. The techniques put forth
Consolidating enormous information and AI could prompt advances medical data with an exactness of up to 90%.
a significant achievement for the two patients and experts.
In any case, even though we distinguished huge numbers of REFERENCES
the main impetuses for the usage of AI in the clinical frame- [1] T. B. Murdoch and A. S. Detsky, ‘‘The inevitable application of big data
work, the previously mentioned hindrances could likewise to health care,’’ J. Amer. Med. Assoc. vol. 309, no. 13, pp. 1351–1352,
2013.
prevent it, particularly if the estimations of the partners are not [2] E. Kolker, V. Özdemir, and E. Kolker, ‘‘How healthcare can refocus on
regarded. Simulated intelligence and huge information must its super-customers (patients, n = 1) and customers (doctors and nurses)
be incorporated and utilized in a moral way on the off chance by leveraging lessons from Amazon, Uber, and Watson,’’ OMICS, J.
Integrative Biol., vol. 20, no. 6, pp. 33–329, 2016.
that we need to create AI apparatuses that will be palatable for [3] S. E. Dilsizian and E. L. Siegel, ‘‘Artificial intelligence in medicine and
handling this pandemic. AI has been a powerful tool to distin- cardiac imaging: Harnessing big data and advanced computing to provide
guish early diseases due to coronavirus and helps in checking personalized medical diagnosis and treatment,’’ Current Cardiol. Rep.,
vol. 16, no. 1, p. 441, Jan. 2014.
the state of the tainted patients. It can essentially improve [4] X. Xu, P. Chen, J. Wang, J. Feng, H. Zhou, X. Li, W. Zhong, and
treatment consistency and dynamic by creating valuable cal- P. Hao, ‘‘Evolution of the novel coronavirus from the ongoing Wuhan
culations. Computer-based intelligence isn’t just useful in the outbreak and modeling of its spike protein for risk of human transmis-
sion,’’ Sci. China Life Sci., vol. 63, no. 3, pp. 457–460, Mar. 2020, doi:
treatment of COVID-19 contaminated patients yet addition- 10.1007/s11427-020-1637-5.
ally for their appropriate medical checkups. It can follow the [5] C. Huang et al., ‘‘Clinical features of patients infected with 2019 novel
emergency of COVID-19 at various scales, for example, clin- coronavirus in Wuhan, China,’’ Lancet, vol. 395, pp. 497–506, Feb. 2020,
doi: 10.1016/S0140-6736(20)30183-5.
ical and epidemiological applications. It is additionally useful [6] J. F.-W. Chan et al., ‘‘A familial cluster of pneumonia associated with
to encourage the exploration of this infection by utilizing the 2019 novel coronavirus indicating person-to-person transmission:
examining the accessible information. Simulated intelligence A study of a family cluster,’’ Lancet, vol. 395, pp. 514–523, Feb. 2020,
doi: 10.1016/S0140-6736(20)30154-9.
can help in creating appropriate treatment regimens, coun- [7] J. Latif, C. Xiao, A. Imran, and S. Tu, ‘‘Medical imaging using machine
teraction methodologies, and medication and immunization learning and deep learning algorithms: A review,’’ in Proc. 2nd Int. Conf.
advancement. Be that as it may, the resulting variables of AI Comput., Math. Eng. Technol. (iCoMET), Sukkur, Pakistan, Jan. 2019,
pp. 1–5.
are viably used for describing the cerebral sickening patient [8] Y. Yang and T. Chen, ‘‘Analysis and visualization implementation of
and the typical patient, which were effectively grouped with medical big data resource sharing mechanism based on deep learning,’’
85% and 96% accuracy esteem by the planning and endorse- IEEE Access, vol. 7, pp. 156077–156088, 2019.
[9] H. Zhao, G. Li, and W. Feng, ‘‘Research on application of artificial
ment tests, independently. Likewise, the NLP was used to intelligence in medical education,’’ in Proc. Int. Conf. Eng. Simulation
evacuate the periphery vein illness-related watchwords from Intell. Control (ESAIC), Changsha, China, Aug. 2018, pp. 340–342.
[10] F. Al-Turjman, M. H. Nawaz, and U. D. Ulusar, ‘‘Intelligence in the [35] M. H. Hesamian, W. Jia, X. He, and P. Kennedy, ‘‘Deep learning
Internet of medical things era: A systematic review of current and techniques for medical image segmentation: Achievements
future trends,’’ Comput. Commun., vol. 150, pp. 644–660, Jan. 2020, doi: and challenges,’’ J. Digit. Imag., vol. 32, no. 4, pp. 582–596,
10.1016/j.comcom.2019.12.030. Aug. 2019.
[11] Y.-K. Chan, Y.-F. Chen, T. Pham, W. Chang, and M.-Y. Hsieh, ‘‘Artificial [36] M. Akagi, Y. Nakamura, T. Higaki, K. Narita, Y. Honda, J. Zhou, Z. Yu,
intelligence in medical applications,’’ J. Healthcare Eng., vol. 2018, N. Akino, and K. Awai, ‘‘Correction to: Deep learning reconstruction
Jul. 2018, Art. no. 4827875, doi: 10.1155/2018/4827875. improves image quality of abdominal ultra-high-resolution CT,’’ Eur.
[12] S. Jha and E. J. Topol, ‘‘Adapting to artificial intelligence: Radiolo- Radiol., vol. 29, no. 8, pp. 4526–4527, Aug. 2019.
gists and pathologists as information specialists,’’ J. Amer. Med. Assoc., [37] P. Nardelli, D. Jimenez-Carretero, D. Bermejo-Pelaez, G. R. Washko,
vol. 316, no. 22, pp. 2353–2354, 2016. F. N. Rahaghi, M. J. Ledesma-Carbayo, and R. S. J. Estepar, ‘‘Pulmonary
[13] SFR-IA Group, CERF, French Radiology Community, ‘‘Artificial intel- Artery–Vein classification in CT images using deep learning,’’ IEEE
ligence and medical imaging 2018: French radiology community white Trans. Med. Imag., vol. 37, no. 11, pp. 2428–2440, Nov. 2018.
paper,’’ Diagnostic Intervent. Imag., vol. 99, no. 11, pp. 727–742, 2018. [38] W. Zhu, Y. Huang, L. Zeng, X. Chen, Y. Liu, Z. Qian, N. Du, W. Fan, and
[14] K. J. Dreyer and J. R. Geis, ‘‘When machines think: Radiology’s next X. Xie, ‘‘AnatomyNet: Deep learning for fast and fully automated whole-
frontier,’’ Radiology, vol. 285, no. 3, pp. 713–718, 2017. volume segmentation of head and neck anatomy,’’ Med. Phys., vol. 46,
[15] Legifrance. (2016). Law of Modernization of Our Health no. 2, pp. 576–589, Feb. 2019.
System. [Online]. Available: https://www.legifrance.gouv.fr/eli/loi/ [39] P. Huang, S. Park, R. Yan, J. Lee, L. C. Chu, C. T. Lin, A. Hussien,
2016/1/26/AFSX1418355L/jo/texte J. Rathmell, B. Thomas, C. Chen, R. Hales, D. S. Ettinger, M. Brock,
[16] C. Villani. Giving Meaning to Artificial Intelligence: For a National P. Hu, E. K. Fishman, E. Gabrielson, and S. Lam, ‘‘Added value
and European Strategy. Accessed: Feb. 19, 2019. [Online]. Avail- of computer-aided CT image features for early lung cancer diagnosis
able: http://www.ladocumentation franc aise.fr/rapports-publi cs/18400 with small pulmonary nodules: A matched case-control study,’’ Radi-
0159/index.shtml ology, vol. 286, no. 1, pp. 286–295, Jan. 2018, doi: 10.1148/radiol.
[17] (2018). Ministry of Solidarity and Health. Health Data Hub Report, 2017162725.
Prefiguration Mission. [Online]. Available: https://solidarites- [40] D. Ardila, A. P. Kiraly, S. Bharadwaj, B. Choi, J. J. Reicher, L. Peng,
sante.gouv.fr/ministere/documentation-et-publications-officielles/ D. Tse, M. Etemadi, W. Ye, G. Corrado, D. P. Naidich, and S. Shetty,
rapports/sante/article/rapport-health-data-hub-mission-de-prefiguration ‘‘End-to-end lung cancer screening with three-dimensional deep learning
[18] Légifrance. (2019). Law on the Organization and Transformation on low-dose chest computed tomography,’’ Nature Med., vol. 25, no. 6,
of the Health System. [Online]. Available: https://www.legifrance. pp. 954–961, Jun. 2019, doi: 10.1038/s41591-019-0447-x.
gouv.fr/eli/loi/2019/7/24/SSAX1900401L/jo/texte [41] D. Salehi, A. Abedi, S. Balakrishnan, and A. Gholamrezanezhad, ‘‘Coro-
[19] W. N. Price, II, and I. G. Cohen, ‘‘Privacy in the age of medical big data,’’ navirus disease 2019 (COVID-19): A systematic review of imaging find-
Nature Med. vol. 25, no. 1, p. 37, 2019. ings in 919 patients,’’ Amer. J. Roentgenol., vol. 2015, pp. 1–7, Mar. 2020.
[20] C. Lamanna and L. Byrne, ‘‘Should artificial intelligence augment med- [42] F. Jiang, L. Deng, L. Zhang, Y. Cai, C. W. Cheung, and Z. Xia, ‘‘Review
ical decision making? The case for an autonomy algorithm,’’ AMA J. of the clinical characteristics of coronavirus disease 2019 (COVID-19),’’
Ethics, vol. 20, no. 9, pp. 902–910, 2018. J. Gen. Internal Med., vol. 35, no. 5, pp. 1545–1549, May 2020, doi:
[21] N. Martinez-Martin, T. R. Insel, P. Dagum, H. T. Greely, and M. K. Cho, 10.1007/s11606-020-05762-w.
‘‘Data mining for health: Staking out the ethical territory of digital
[43] T. Singhal, ‘‘A review of coronavirus disease-2019 (COVID-19),’’ Indian
phenotyping,’’ npj Digit. Med., vol. 1, no. 1, p. 68, Dec. 2018.
J. Pediatrics, vol. 87, no. 4, pp. 281–286, Apr. 2020.
[22] Google. (Feb. 19, 2019). Our Principles. [Online]. Available:
[44] C. Sohrabi, Z. Alsafi, N. O’Neill, M. Khan, A. Kerwan, A. Al-Jabir,
https://ai.google/principles/
C. Iosifidis, and R. Agha, ‘‘World health organization declares global
[23] A. Esteva, A. Robicquet, B. Ramsundar, V. Kuleshov, M. DePristo,
emergency: A review of the 2019 novel coronavirus (COVID-19),’’ Int.
K. Chou, C. Cui, G. Corrado, S. Thrun, and J. Dean, ‘‘A guide to deep
J. Surg., vol. 76, pp. 71–76, Apr. 2020.
learning in healthcare,’’ Nature Med., vol. 25, no. 1, p. 24, 2019.
[24] J. C. Pomerol, ‘‘Artificial intelligence and human decision making,’’ Eur. [45] N. Chen, M. Zhou, X. Dong, J. Qu, F. Gong, Y. Han, Y. Qiu, J. Wang,
J. Oper. Res., vol. 99, no. 1, pp. 3–25, 1997. Y. Liu, Y. Wei, J. Xia, T. Yu, X. Zhang, and L. Zhang, ‘‘Epidemiolog-
ical and clinical characteristics of 99 cases of 2019 novel coronavirus
[25] OPECST. (2017). For a Controlled, Useful and Demystified Arti-
pneumonia in Wuhan, China: A descriptive study,’’ Lancet, vol. 395,
ficial Intelligence. Accessed: Feb. 19, 2019. [Online]. Available:
no. 10223, pp. 507–513, Feb. 2020.
https://www.senat.fr/rap/r16-464-1/r16-464-1.html
[26] I. Tabus, G. Korodi, and J. Rissanen, ‘‘DNA sequence compression using [46] D. Wang, B. Hu, C. Hu, F. Zhu, X. Liu, J. Zhang, B. Wang, H. Xiang,
the normalized maximum likelihood model for discrete regression,’’ in Z. Cheng, Y. Xiong, Y. Zhao, Y. Li, X. Wang, and Z. Peng, ‘‘Clinical
Proc. Data Compress. Conf. (DCC), 2003, p. 253. characteristics of 138 hospitalized patients with 2019 novel coronavirus–
[27] G. Koradi and I. Tabus, ‘‘An efficient normalized maximum likelihood infected pneumonia in Wuhan, China,’’ J. Amer. Med. Assoc., vol. 323,
algorithm for DNA sequence compression,’’ ACM Trans. Inf. Syst., no. 11, p. 1061, Mar. 2020.
vol. 23, no. 1, pp. 3–34, 2005. [47] (2017). Administration UFaD Guidance for Industry: Electronic
[28] M. D. Cao, T. I. Dix, L. Allison, and C. Mears, ‘‘A simple statistical algo- Source Data in Clinical Investigations. [Online]. Available:
rithm for biological sequence compression,’’ in Proc. Data Compress. https://www.fda.gov/downloads/drugs/guidances/ucm328691
Conf. (DCC), Snowbird, UT, USA, Mar. 2007, pp. 43–52. [48] R. J. Gillies, P. E. Kinahan, and H. Hrica, ‘‘Radiomics: Images are more
[29] R. Rajeswari and A. Apparao, ‘‘GenBit compress-algorithm for repetitive than pictures, they are data,’’ Radiology, vol. 278, no. 2, pp. 563–577,
and nonrepetitive DNA sequences,’’ J. Theor. Appl. Inf. Technol., vol. 11, 2016.
pp. 25–29, Jan. 2010. [49] C. Li, G.-Y. Liang, W.-Z. Yao, J. Sui, X. Shen, Y.-Q. Zhang, H. Peng,
[30] T. H. A. Soliman, T. F. Gharib, A. A. Alian, and M. A. E. Sharkawy, W.-W. Hong, Y.-C. Ye, Z.-Y. Zhang, W.-H. Zhang, L.-H. Yin, and
‘‘A lossless compression algorithm for DNA sequences,’’ Int. J. Bioinf. Y.-P. Pu, ‘‘Integrated analysis of long non-coding RNA competing inter-
Res. Appl., vol. 5, no. 6, p. 593, 2009. actions reveals the potential role in progression of human gastric cancer,’’
[31] K. N. Mishra, D. A. Aaggarwal, D. E. Abdelhadi, and D. P. C. Srivastava, Int. J. Oncol. vol. 48, no. 5, pp. 1965–1976, 2016.
‘‘An efficient horizontal and vertical method for online DNA sequence [50] H. Shin, K. H. Kim, and C. Song, ‘‘Electrodiagnosis support system for
compression,’’ Int. J. Comput. Appl., vol. 3, no. 1, pp. 39–46, Jun. 2010. localizing neural injury in an upper limb,’’ J. Amer. Med. Inform. Assoc.,
[32] T. Greenhalgh, G. C. H. Koh, and J. Car, ‘‘COVID-19: A remote assess- vol. 17, no. 3, pp. 345–347, 2010.
ment in primary care,’’ BMJ, vol. 368, p. m1182, Mar. 2020. [51] G. Karakülah, O. Dicle, O. Koşaner, A. S. Karakülah, C. C. Birant,
[33] X. Liu, S. Guo, B. Yang, S. Ma, H. Zhang, J. Li, C. Sun, L. Jin, T. Berber, and S. Canbek, ‘‘Computer based extraction of phenoptypic
X. Li, Q. Yang, and Y. Fu, ‘‘Automatic organ segmentation for CT scans features of human congenital anomalies from the digital literature with
based on super-pixel and convolutional neural networks,’’ J. Digit. Imag., natural language processing techniques,’’ Stud. Health Technol. Inform.,
vol. 31, no. 6, pp. 748–760, 2018. vol. 205, pp. 570–574, Aug. 2014.
[34] M. Gharbi, J. Chen, J. T. Barron, S. W. Hasinoff, and F. Durand, ‘‘Deep [52] B. McCall, ‘‘COVID-19 and artificial intelligence: Protecting health-care
bilateral learning for real-time image enhancement,’’ ACM Trans. Graph., workers and curbing the spread,’’ Lancet Digit. Health, vol. 2, no. 4,
vol. 36, no. 4, p. 118, 2017. pp. e166–e167, 2020, doi: 10.1016/S2589-7500(20)30054-6.
[53] FDA. (2019). Commissioner O of the Statement From FDA Com- [72] Y. Yang, ‘‘Epidemiological and clinical features of the 2019
missioner Scott Gottlieb, M.D. on Steps Toward a New, Tailored novel coronavirus outbreak in China,’’ medRxiv, Feb. 2020, doi:
Review Framework for Artificial Intelligence-Based Medical Devices. 10.1101/2020.02.10.20021675.
[Online] Available: https://news-events/press-announcements/statement- [73] J. T. Wu, K. Leung, and G. M. Leung, ‘‘Nowcasting and forecasting the
fda-commissioner-scott-gottlieb-mdsteps potential domestic and international spread of the 2019-nCoV outbreak
[54] F. Al-Turjman, ‘‘Intelligence and security in big 5G-oriented IoNT: originating in Wuhan, China: A modelling study,’’ Lancet, vol. 396,
An overview,’’ Future Gener. Comput. Syst., vol. 102, pp. 357–368, pp. 689–697, Feb./Mar. 2020, doi: 10.1016/S0140-6736(20)30260-9.
Jan. 2020. [74] W. Ji, W. Wang, X. Zhao, J. Zai, and X. Li, ‘‘Homologous recombination
[55] J. Bedford, D. Enria, J. Giesecke, D. L. Heymann, C. Ihekweazu, within the spike glycoprotein of the newly identified coronavirus may
V. Kobinger, H. C. Lane, V. Memish, M.-D. Oh, A. A. Sall, A. Schuchat, boost cross-species transmission from snake to human,’’ J. Med. Virol.,
K. Ungchusak, and L. H. Wieler, ‘‘COVID-19: Towards controlling of a vol. 92, Jan. 2020, doi: 10.1002/jmv.25682.
pandemic,’’ Lancet, vol. 395, no. 10229, pp. 1015–1018, 2020. [75] C. M. Bishop, Ed., Pattern Recognition and Machine Learning (Informa-
[56] F. Chollet, ‘‘Xception: Deep learning with depthwise separable convo- tion Science and Statistics). 2007.
lutions,’’ in Proc. IEEE Conf. Comput. Vis. Pattern Recognit. (CVPR), [76] G. Orrù, W. Pettersson-Yeo, A. F. Marquand, G. Sartori, and A. Mechelli,
Jul. 2017, pp. 1251–1258. ‘‘Using support vector machine to identify imaging biomarkers of neu-
[57] G. Matar, J.-M. Lina, and G. Kaddoum, ‘‘Artificial neural network for rological and psychiatric disease: A critical review,’’ Neurosci. Biobehav.
in-bed posture classification using bed-sheet pressure sensors,’’ IEEE J. Rev., vol. 36, pp. 1140–1152, Apr. 2012.
Biomed. Health Informat., vol. 24, no. 1, pp. 101–110, Jan. 2020. [77] N. H. Sweilam, A. A. Tharwat, and N. K. A. Moniem, ‘‘Support vector
[58] O. Gozes, M. Frid-Adar, H. Greenspan, P. D. Browning, H. Zhang, W. Ji, machine for diagnosis cancer disease: A comparative study,’’ Egyptian
A. Bernheim, and E. Siegel, ‘‘Rapid AI development cycle for the coro- Informat. J., vol. 11, no. 2, pp. 81–92, Dec. 2010.
navirus (COVID-19) pandemic: Initial results for automated detection [78] L. Khedher, J. Ramírez, J. M. Górriz, A. Brahim, and F. Segovia ‘‘Early
& patient monitoring using deep learning CT image analysis,’’ 2020, diagnosis of Alzheimer’s disease based on partial least squares, principal
arXiv:2003.05037. [Online]. Available: http://arxiv.org/abs/2003.05037 component analysis and support vector machine using segmented MRI
[59] A. Ranjan, V. Jampani, L. Balles, K. Kim, D. Sun, J. Wulff, and images,’’ Neurocomputing, vol. 151, pp. 139–150, Mar. 2015.
M. J. Black, ‘‘Competitive collaboration: Joint unsupervised learning of [79] D. Farina, I. Vujaklija, M. Sartori, T. Kapelner, F. Negro, N. Jiang,
depth, camera motion, optical flow and motion segmentation,’’ in Proc. K. Bergmeister, A. Andalib, J. Principe, and O. C. Aszmann,
IEEE/CVF Conf. Comput. Vis. Pattern Recognit. (CVPR), Jun. 2019, ‘‘Man/machine interface based on the discharge timings of spinal
pp. 12240–12249. motor neurons after targeted muscle reinnervation,’’ Nature Biomed.
[60] A. Z. Zhu, L. Yuan, K. Chaney, and K. Daniilidis, ‘‘Unsupervised event- Eng., vol. 1, no. 2, p. 0025, Feb. 2017.
based learning of optical flow, depth, and egomotion,’’ in Proc. IEEE/CVF [80] L. Mirtskhulava, J. Wong, S. Al-Majeed, and G. Pearce, ‘‘Artificial neural
Conf. Comput. Vis. Pattern Recognit. (CVPR), Jun. 2019, pp. 989–997. network model in stroke diagnosis,’’ in Proc. 17th UKSim-AMSS Int.
[61] F. Al-Turjman, M. Z. Hasan, and H. Al-Rizzo, ‘‘Task scheduling in cloud- Conf. Modelling Simulation (UKSim), Mar. 2015, pp. 50–53.
based survivability applications using swarm optimization in IoT,’’ Trans. [81] J. Khan, J. S. Wei, M. Ringnér, L. H. Saal, M. Ladanyi, F. Westermann,
Emerg. Telecommun. Technol., vol. 30, no. 8, p. e3539, 2018. F. Berthold, M. Schwab, C. R. Antonescu, C. Peterson, and P. S. Meltzer,
[62] K. H. Wan, S. S. Huang, A. L. Young, and D. S. C. Lam, ‘‘Precautionary ‘‘Classification and diagnostic prediction of cancers using gene expres-
measures needed for ophthalmologists during pandemic of the coron- sion profiling and artificial neural networks,’’ Nature Med., vol. 7,
avirus disease 2019 (COVID-19),’’ Acta Ophthalmologica, vol. 98, no. 3, pp. 673–679, Jun. 2001.
pp. 221–222, May 2020. [82] J. Dheeba, N. A. Singh, and S. T. Selvi, ‘‘Computer-aided detection of
[63] L. Li, L. Qin, Z. Xu, Y. Yin, X. Wang, B. Kong, J. Bai, Y. Lu, breast cancer on mammograms: A swarm intelligence optimized wavelet
Z. Fang, Q. Song, K. Cao, D. Liu, G. Wang, Q. Xu, X. Fang, S. Zhang, neural network approach,’’ J. Biomed. Informat., vol. 49, pp. 45–52,
J. Xia, and J. Xia, ‘‘Artificial intelligence distinguishes COVID-19 from Jun. 2014.
community acquired pneumonia on chest CT,’’ Radiology, Mar. 2020, [83] T. J. Hirschauer, H. Adeli, and J. A. Buford, ‘‘Computer-aided diagnosis
Art. no. 200905, doi: 10.1148/radiol.2020200905. of Parkinson’s disease using enhanced probabilistic neural network,’’
[64] E. M. Forman, S. G. Kerrigan, M. L. Butryn, A. S. Juarascio, J. Med. Syst., vol. 39, no. 11, p. 179, Nov. 2015.
S. M. Manasse, S. Ontañón, D. H. Dallal, R. J. Crochiere, and D. Moskow, [84] A. Esteva, B. Kuprel, R. A. Novoa, J. Ko, S. M. Swetter, H. M. Blau, and
‘‘Can the artificial intelligence technique of reinforcement learning use S. Thrun, ‘‘Dermatologist-level classification of skin cancer with deep
continuously-monitored digital data to optimize treatment for weight neural networks,’’ Nature, vol. 542, pp. 115–118, Jan. 2017.
loss?’’ J. Behav. Med., vol. 42, pp. 276–290, Aug. 2018. [85] S. S. Han, M. S. Kim, W. Lim, G. H. Park, I. Park, and S. E. Chang,
[65] W. Naudé. (2020). Artificial Intelligence Against COVID-19: An Early ‘‘Classification of the clinical images for benign and malignant cutaneous
Review. [Online]. Available: https://towardsdatascience.com/artificial- tumors using a deep learning algorithm,’’ J. Investigative Dermatol.,
intelligence-against-covid-19-an-early-review-92a8360edaba vol. 138 pp. 1529–1538, Jul. 2018.
[66] M. Smolaks. (2020). Deep Learning Systems Can Speed Up Diagnosis [86] A. Menegola, J. Tavares, M. Fornaciali, L. T. Li, S. E. F. De Avila, and E.
of COVID-19 Patients. [Online]. Available: https://aibusiness.com/deep- Valle, ‘‘RECOD titans at ISIC challenge 2017,’’ 2017, arXiv:1703.04819.
learning-systems-can-identify-covid-19-infections-on-x-rays/ [Online]. Available: https://arxiv.org/abs/1703.04819
[67] W. Naude. (2020). Artificial Intelligence Against COVID-19: An Early [87] Y. Fujisawa, Y. Otomo, Y. Ogata, Y. Nakamura, R. Fujita, Y. Ishitsuka,
Review, Medium. [Online]. Available: https://towardsdatascience. R. Watanabe, N. Okiyama, K. Ohara, and M. Fujimoto, ‘‘Deep-learning-
com/artificial-intelligence-against-covid-19-an-early-review- based, computer-aided classifier developed with a small dataset of clinical
92a8360edaba images surpasses board-certified dermatologists in skin tumour diagno-
[68] M. Hollister. (2020). COVID-19: AI can help, but the right sis,’’ Brit. J. Dermatol., vol. 180, no. 2, pp. 373–381, Feb. 2019.
human input is key. World Economic Forum. [Online]. Available: [88] A. Rezvantalab, H. Safigholi, and S. Karimijeshni, ‘‘Dermatologist level
https://www.weforum.org/agenda/2020/03/covid-19-crisis- dermoscopy skin cancer classification using different deep learning con-
artificialintelligence-creativity/ volutional neural networks algorithms,’’ CoRR, vol. abs/1810.10348,
[69] S. Wang, B. Kang, J. Ma, X. Zeng, M. Xiao, J. Guo, M. Cai, J. Yang, Y. Li, Mar. 2017.
X. Meng, and B. Xu, ‘‘A deep learning algorithm using CT images to [89] A. Toh. (2020). Big Data Could Undermine the COVID-19 Response.
screen for corona virus disease (COVID-19),’’ MedRxiv, 2020. [Online]. [Online]. Available: https://www.wired.com/story/big-data-could-
Available: https://doi.org/10.1101/2020.02.14.20023028. undermine-the-covid-19-response/
[70] A. Devereaux, M. D. Christian, J. R. Dichter, J. A. Geiling, and [90] M. Combalia, N. C. F. Codella, V. Rotemberg, B. Helba, V. Vilaplana,
L. Rubinson, ‘‘Summary of suggestions from the task force for mass O. Reiter, C. Carrera, A. Barreiro, A. C. Halpern, S. Puig, and
critical care summit, January 26–27, 2007,’’ Chest, vol. 133, no. 5, J. Malvehy, ‘‘BCN20000: Dermoscopic lesions in the wild,’’ 2019,
pp. 1S–7S, May 2008, doi: 10.1378/chest.08-0649. arXiv:1908.02288. [Online]. Available: http://arxiv.org/abs/1908.02288
[71] N. E. Huang and F. Qiao, ‘‘A data driven time-dependent transmission [91] P. Tschandl, C. Rosendahl, and H. Kittler, ‘‘The HAM10000 dataset, a
rate for tracking an epidemic: A case study of 2019-nCoV,’’ Sci. Bull., large collection of multi-source dermatoscopic images of common pig-
vol. 65, no. 6, pp. 425–427, Mar. 2020, doi: 10.1016/j.scib.2020.02.005. mented skin lesions,’’ Sci. Data, vol. 5, no. 1, Dec. 2018, Art. no. 180161.
[92] R. Vaishya, M. Javaid, I. H. Khan, and A. Haleem, ‘‘Artificial intelli- [112] X. Xu, X. Jiang, C. Ma, P. Du, X. Li, S. Lv, L. Yu, Y. Chen, J. Su, G. Lang,
gence (AI) applications for COVID-19 pandemic,’’ Diabetes Metabolic Y. Li, H. Zhao, K. Xu, L. Ruan, and W. Wu, ‘‘Deep learning system to
Syndrome, Clin. Res. Rev., vol. 14, no. 4, pp. 337–339, Jul. 2020. screen coronavirus disease 2019 pneumonia,’’ 2020, arXiv:2002.09334.
[93] A. A. Ardakani, A. R. Kanafi, U. R. Acharya, N. Khadem, and [Online]. Available: http://arxiv.org/abs/2002.09334
A. Mohammadi, ‘‘Application of deep learning technique to manage [113] M. Abadi et al., ‘‘TensorFlow: Large-scale machine learning on hetero-
COVID-19 in routine clinical practice using CT images: Results of 10 geneous distributed Systems,’’ 2016, arXiv:1603.04467. [Online]. Avail-
convolutional neural networks,’’ Comput. Biol. Med., vol. 121, Jun. 2020, able: https://arxiv.org/abs/1603.04467
Art. no. 103795. [114] E. Long, H. Lin, Z. Liu, X. Wu, L. Wang, J. Jiang, Y. An, Z. Lin, X. Li,
[94] N. C. F. Codella, D. Gutman, M. E. Celebi, B. Helba, M. A. Marchetti, J. Chen, J. Li, Q. Cao, D. Wang, X. Liu, W. Chen, and Y. Liu, ‘‘An arti-
S. W. Dusza, A. Kalloo, K. Liopyris, N. Mishra, H. Kittler, and ficial intelligence platform for the multihospital collaborative manage-
A. Halpern, ‘‘Skin lesion analysis toward melanoma detection: A chal- ment of congenital cataracts,’’ Nature Biomed. Eng., vol. 1, Jan. 2017,
lenge at the 2017 international symposium on biomedical imaging (ISBI), Art. no. 0024.
hosted by the international skin imaging collaboration (ISIC),’’ in Proc. [115] S. Chilamkurthy, R. Ghosh, S. Tanamala, M. Biviji, N. G. Campeau,
IEEE 15th Int. Symp. Biomed. Imag. (ISBI), Apr. 2018, p. 168. V. K. Venugopal, V. Mahajan, P. Rao, and P. Warier, ‘‘Deep learning algo-
[95] Q. Li et al., ‘‘Early transmission dynamics in Wuhan, China, of novel rithms for detection of critical findings in head CT scans: A retrospective
coronavirus–infected pneumonia,’’ New England J. Med., vol. 382, study,’’ Lancet, vol. 392, no. 10162, pp. 2388–2396, Dec. 2018.
pp. 1199–1207, Mar. 2020, doi: 10.1056/NEJMoa2001316. [116] S. Chae, S. Kwon, and D. Lee, ‘‘Predicting infectious disease using deep
[96] T. NyÍri and A. Kiss, ‘‘Novel ensembling methods for dermatological learning and big data,’’ Int. J. Environ. Res. Public Health, vol. 15, no. 8,
image classification,’’ in Proc. Int. Conf. Theory Pract. Natural Comput. p. 1596, Jul. 2018.
Cham, Switzerland: Springer, 2018, pp. 438–448. [117] S. Bansal, G. Chowell, L. Simonsen, A. Vespignani, and C. Viboud,
[97] W.-J. Guan et al., ‘‘Clinical characteristics of 2019 novel ‘‘Big data for infectious disease surveillance and modeling,’’ J. Infectious
coronavirus infection in China,’’ MedRxiv, Feb. 2020, doi: Diseases, vol. 214, no. 4, pp. S375–S379, Dec. 2016.
10.1101/2020.02.06.20020974. [118] M. Eisenstein, ‘‘Infection forecasts powered by big data,’’ Nature,
[98] J. Lei, J. Li, X. Li, and X. Qi, ‘‘CT imaging of the 2019 novel coro- vol. 555, no. 7695, pp. S2–S4, Mar. 2018.
navirus (2019-nCoV) pneumonia,’’ Radiology, vol. 298, no. 1, 2020, [119] S. Jacob, V. G. Menon, F. Al-Turjman, P. G. Vinoj, and L. Mostarda,
Art. no. 200236. ‘‘Artificial muscle intelligence system with deep learning for post-stroke
[99] L. Bi, J. Kim, E. Ahn, and D. Feng, ‘‘Automatic skin lesion analysis using assistance and rehabilitation,’’ IEEE Access, vol. 7, pp. 133463–133473,
large-scale dermoscopy images and deep residual networks,’’ Mar. 2017, 2019.
arXiv:1703.04197. [Online]. Available: https://arxiv.org/abs/1703.04197 [120] A. Waheed, M. Goyal, D. Gupta, A. Khanna, F. Al-Turjman, and
P. R. Pinheiro, ‘‘CovidGAN: Data augmentation using auxiliary classi-
[100] A. Mahbod, G. Schaefer, I. Ellinger, R. Ecker, A. Pitiot, and C. Wang,
fier GAN for improved COVID-19 detection,’’ IEEE Access, vol. 8,
‘‘Fusing fine-tuned deep features for skin lesion classification,’’ Comput.
pp. 91916–91923, 2020, doi: 10.1109/ACCESS.2020.2994762.
Med. Imag. Graph., vol. 71, pp. 19–29, Jan. 2019.
[121] C. Buckee. (2020). Improving Epidemic Surveillance and
[101] J. R. Hagerty, R. J. Stanley, H. A. Almubarak, N. Lama, R. Kasmi, P. Guo,
Response: Big Data is Dead, Long Live Big Data. [Online].
R. J. Drugge, H. S. Rabinovitz, M. Oliviero, and W. V. Stoecker, ‘‘Deep
Available: https://www.thelancet.com/journals/landig/article/PIIS2589-
learning and handcrafted method fusion: Higher diagnostic accuracy
7500(20)30059-5/fulltext
for melanoma dermoscopy images,’’ IEEE J. Biomed. Health Informat.,
[122] R. Bean. (2020). Big Data in the Time of Coronavirus (COVID-19).
vol. 23, no. 4, pp. 1385–1391, Jul. 2019.
[Online]. Available: https://www.forbes.com/sites/ciocentral/2020/03/
[102] D. Ardila, A. P. Kiraly, S. Bharadwaj, B. Choi, J. J. Reicher, L. Peng,
30/big-data-in-the-time-of-coronavirus-fCOVID-19g/
D. Tse, M. Etemadi, W. Ye, G. Corrado, D. P. Naidich, and S. Shetty
[123] J. Kent. (2020). Understanding the COVID-19 Pandemic as a Big
‘‘End-to-end lung cancer screening with three-dimensional deep learning
Data Analytics Issue. [Online]. Available: https://healthitanalytics.
on low-dose chest computed tomography,’’ Nature Med., vol. 25, no. 6,
com/news/understanding-the-fCOVID-19g-pandemic-as-a-big-data-
pp. 954–961, 2019.
analytics-issue
[103] S. Serte and H. Demirel, ‘‘Wavelet-based deep learning for skin
[124] V. Kumar. (2020). Is Big Data Effective in Response to Coronavirus Out-
lesion classification,’’ IET Image Process., vol. 14, no. 4, pp. 720–726,
break? [Online]. Available: https://www.analyticsinsight.net/big-data-
Mar. 2020.
effective-response-coronavirus-outbreak/
[104] D. Ravì, C. Wong, F. Deligianni, M. Berthelot, J. Andreu-Perez, B. Lo, [125] X. Zhao, X. Liu, and X. Li, ‘‘Tracking the spread of novel coro-
and G.-Z. Yang, ‘‘Deep learning for health informatics,’’ IEEE J. Biomed. navirus (2019-nCoV) based on big data,’’ medRxiv, Feb. 2020, doi:
Health Informat., vol. 21, no. 1, pp. 4–21, Jan. 2017. 10.1101/2020.02.07.20021196.
[105] M. Chung, A. Bernheim, X. Mei, N. Zhang, M. Huang, X. Zeng, [126] C. Zhou et al., ‘‘COVID-19: Challenges to GIS with big data,’’ Geogr.
J. Cui, W. Xu, Y. Yang, Z. A. Fayad, A. Jacobi, K. Li, S. Li, and Sustainability, vol. 1, no. 1, pp. 77–87, Mar. 2020.
H. Shan, ‘‘CT imaging features of 2019 novel coronavirus (2019- [127] J. C. M. Than, L. Saba, N. M. Noor, O. M. Rijal, R. M. Kassim, A. Yunus,
nCoV),’’ Radiology, vol. 295, no. 1, Apr. 2020, Art. no. 200230, doi: H. S. Suri, M. Porcu, and J. S. Suri, ‘‘Lung disease stratification using
10.1148/radiol.2020200230. amalgamation of riesz and Gabor transforms in machine learning frame-
[106] P. Kantor, Foundations of Statistical Natural Language Processing. work,’’ Comput. Biol. Med., vol. 89, pp. 197–211, Oct. 2017.
Cambridge, MA, USA: MIT Press, 1999, pp. 91–92. [128] Y. Gu, X. Lu, L. Yang, B. Zhang, D. Yu, Y. Zhao, L. Gao, L. Wu, and
[107] V. M. Castro, D. Dligach, S. Finan, S. Yu, A. Can, M. Abd-El-Barr, T. Zhou, ‘‘Automatic lung nodule detection using a 3D deep convolutional
V. Gainer, N. A. Shadick, S. Murphy, T. Cai, G. Savova, S. T. Weiss, neural network combined with a multi-scale prediction strategy in chest
and R. Du, ‘‘Large-scale identification of patients with cerebral CTs,’’ Comput. Biol. Med., vol. 103, pp. 220–231, Dec. 2018.
aneurysms using natural language processing,’’ Neurology, vol. 88, no. 2, [129] W. Zhang, X. Wang, X. Li, and J. Chen, ‘‘3D skeletonization feature based
pp. 164–168, 2017. computer-aided detection system for pulmonary nodules in CT datasets,’’
[108] N. Afzal, S. Sohn, S. Abram, C. G. Scott, R. Chaudhry, H. Liu, I. J. Kullo, Comput. Biol. Med., vol. 92, pp. 64–72, Jan. 2018.
and A. M. Arruda-Olson ‘‘Mining peripheral arterial disease cases from [130] L. Sun, Z. Wang, H. Pu, G. Yuan, L. Guo, T. Pu, and Z. Peng, ‘‘Spectral
narrative clinical notes using natural language processing,’’ J. Vascular analysis for pulmonary nodule detection using the optimal fractional S-
Surg., vol. 65, no. 6, pp. 1753–1761, 2017. transform,’’ Comput. Biol. Med., vol. 119, Apr. 2020, Art. no. 103675.
[109] M. Fiszman, W. W. Chapman, D. Aronsky, R. S. Evans, and P. J. Haug, [131] W. Sun, B. Zheng, and W. Qian, ‘‘Automatic feature learning using mul-
‘‘Automatic detection of acute bacterial pneumonia from chest X-ray tichannel ROI based on deep structured algorithms for computerized lung
reports,’’ J. Amer. Med. Inform. Assoc., vol. 7, no. 6, pp. 593–604, cancer diagnosis,’’ Comput. Biol. Med., vol. 89, pp. 530–539, Oct. 2017.
Nov. 2000. [132] V. Gulshan, L. Peng, and M. Coram, M. C. Stumpe, D. Wu,
[110] Y. Lecun, L. Bottou, Y. Bengio, and P. Haffner, ‘‘Gradient-based learn- A. Narayanaswamy, S. Venugopalan, K. Widner, T. Madams, J. Cuadros,
ing applied to document recognition,’’ Proc. IEEE, vol. 86, no. 11, R. Kim, R. Raman, P. C. Nelson, J. L. Mega, and D. R. Webster,
pp. 2278–2324, Nov. 1998. ‘‘Development and validation of a deep learning algorithm for detection
[111] (2017). Research BA Caffe. [Online]. Available: http://caffe. of diabetic retinopathy in retinal fundus photographs,’’ J. Amer. Med.
berkeleyvision.org/ Assoc., vol. 316, pp. 2402–2410, Nov. 2016.
[133] C. Zheng, X. Deng, Q. Fu, Q. Zhou, J. Feng, H. Ma, W. Liu, and X. Wang, OUNS BOUACHIR (Member, IEEE) received
‘‘Deep learning-based detection for COVID-19 from chest CT using weak the degree in telecommunications engineering
label,’’ medRxiv, Mar. 2020, doi: 10.1101/2020.03.12.20027185. from the Higher School of Communications
[134] Z. Ullah, F. Al-Turjman, L. Mostarda, and R. Gagliardi, ‘‘Applications (Sup’Com), Tunis, Tunisia, and the Ph.D. degree
of artificial intelligence and machine learning in smart cities,’’ Comput. in computer engineering from the University Paul
Commun., vol. 154, pp. 313–323, Mar. 2020. Sabatier (Toulouse 3), Toulouse, France. She
[135] S. Oueida, M. Aloqaily, and S. Ionescu, ‘‘A smart healthcare reward was an Assistant Professor and a Postdoctoral
model for resource allocation in smart city,’’ Multimedia Tools Appl.,
Researcher with Canadian University Dubai. She
vol. 78, no. 17, pp. 24573–24594, Sep. 2019.
is currently an Assistant Professor of computer
[136] S. Oueida, Y. Kotb, M. Aloqaily, Y. Jararweh, and T. Baker, ‘‘An edge
computing based smart healthcare framework for resource management,’’ engineering with the College of Technological
Sensors, vol. 18, no. 12, p. 4307, Dec. 2018. Innovation, Zayed University, United Arab Emirates. Her current research
[137] B. D. Deebak, F. Al-Turjman, M. Aloqaily, and O. Alfandi, ‘‘IoT- interests include connected and autonomous vehicles, block chain solutions,
BSFCAN: A smart context-aware system in IoT-cloud using mobile- the Internet of Things (IoT), energy optimization, AI, and ML. She is a
fogging,’’ Future Gener. Comput. Syst., vol. 109, pp. 368–381, Aug. 2020. part of the technical program committee of numerous journals, conferences,
[138] C. Lawrence. (2020). Is Cloud Computing the Superhero of COVID- and workshops, such as BCCA2020, ITCVT-NOMS2020, PEDISWESA-
19? [Online]. Available: https://www.codemotion.com/magazine/dev- ISCC2020, and AICSSA2019.
hub/cloud-manager/cloud-computing-covid-19/
[139] C. Morrison and H. Archer. (2020). Coronavirus Resilience Planning: FADI AL-TURJMAN (Member, IEEE) received
Use of Near Me Video Consulting in GP Practices Scottish Gov- the Ph.D. degree in computer science from
ernment Technology Enabled Care Programme. [Online]. Available: Queen’s University, Kingston, ON, Canada,
https://www.sehd.scot.nhs.uk/pca/PCA2020(M)03 in 2011. He is currently a Full Professor and a
[140] A. Pludderman, R. Hobbs, and K. R. Mahtani. (2020). Rapid Research Center Director with Near East Uni-
Diagnosis of Community-Acquired Pneumonia for Clinicians Rapid versity, Nicosia, Cyprus. He is also a Leading
Review Oxford COVID-19 Evidence Series University of Oxford. Authority in the areas of smart/intelligent, wire-
[Online]. Available: https://www.cebm.net/rapid-diagnosis-strategy-of-
less, and mobile networks’ architectures, proto-
community-acquired-pneumonia-for-clinician/
cols, deployments, and performance evaluation.
[141] B. Mitchell. (2019). Computer Network Storage Explained. [Online].
Available: https://www.lifewire.com/what-is-computer-networking- His publication history spans over 250 publica-
816249 tions in journals, conferences, patents, books, book chapters, numerous
[142] U. Ibrahim, M. Ozsos, and F. Al-Turjman, ‘‘Futuristic CRISPR-based keynotes, and plenary talks at flagship venues. He has authored and edited
biosensing in the cloud and Internet of Things era: An overview,’’ Multi- more than 25 books about cognition, security, and wireless sensor networks’
media Tools Appl., Jun. 2020, doi: 10.1007/s11042-020-09010-5. deployments in smart environments, published by Taylor and Francis,
[143] M. Al-Zinati, R. Alrashdan, B. Al-Duwairi, and M. Aloqaily, ‘‘A re- Elsevier, and Springer. He has led a number of international symposia and
organizing biosurveillance framework based on fog and mobile edge workshops in flagship communication society conferences. He received
computing,’’ Multimedia Tools Appl., vol. 1, pp. 1–21, May 2020. several recognitions and best papers’ awards at top international confer-
[144] M. Al-Khafajiy et al., ‘‘Intelligent control and security of fog resources in ences. He also received the prestigious Best Research Paper Award from
healthcare systems via a cognitive fog model,’’ Int. J. Wireless Inf. Netw., Elsevier Computer Communications Journal, from 2015 to 2018, and the Top
vol. 27, pp. 467–483, 2020, doi: 10.1007/s10776-020-00491-7. Researcher Award from Antalya Bilim University, Turkey, in 2018. He serves
[145] B. D. Deebak, F. Al-Turjman, M. Aloqaily, and O. Alfandi, as an Associate Editor and a Lead Guest/Associate Editor for several well
‘‘An authentic-based privacy preservation protocol for smart e-Healthcare reputed journals, including the IEEE COMMUNICATIONS SURVEYS AND TUTORIALS
systems in IoT,’’ IEEE Access, vol. 7, pp. 135632–135649, 2019.
(IF 22.9) and Sustainable Cities and Society (Elsevier) (IF 4.7).