Chris Suter’s Post

[#Video] A new video of our QuanTIM Webinars is online! 💬 Prediction models in healthcare: a playground for researchers 🗣 Richard Riley (Institute of Applied Health Research, University of Birmingham, UK) 💻 https://lnkd.in/djvJz57c ⬇ Abstract ⬇ Clinical prediction models estimate an individual’s risk of a particular health outcome. In this talk, I suggest that prediction model research has become the academic’s playground, and that machine learning and AI have only exacerbated this problem. I highlight common pitfalls, including model instability and poor calibration, and propose some solutions moving forwards. The talk is aimed at a broad audience. #PublicHealth #ClinicalPredictionModels #MachineLearning #AI #PredictiveAlgorithms #Stability Cc. Aix-Marseille Université, Inserm, IRD, Santé publique France, Société Française de Santé Publique (SFSP), Collège de Liaison des Internes de Santé Publique (CLISP), EHESP - École des hautes études en santé publique, EUPHA - European Public Health Association

I’m honored to have been invited to participate on the Industry and Academic Partnership panel later this afternoon at the Inaugural UT AI System Symposium in Health Care at UT Southwestern Medical Center . This event has been an incredible two days of frank, insightful discussion on the application of AI in Healthcare. As far as we’ve come, there are many unanswered questions. It’s going to take a united multidisciplinary effort to realize the promise of AI in the real world setting. Walking around the poster floor, you can see the incredible innovation happening in the labs. As we think about “models” there is a clear recognition that there will in fact be many models created from all corners of academia and industry. Considering an unbounded multi model future, I keep coming back to the need for organizations to build the technical infrastructure, data governance policies and local validation frameworks for implementing these models in their own local environments. https://lnkd.in/gwPW3bSi

[#Webinar] D-8 until our next QuanTIM Webinar! ⤵ 💬 Prediction models in healthcare: a playground for researchers 🗣 Richard Riley (University of Birmingham) 💻 https://lnkd.in/dixBFbaU ------- Interested in a QuanTIM Webinar's panelist spot and a funding for a scientific conference or a research stay in one of the SESSTIM’s teams? Check-out our PhD Student Award! 🎓🏅 ➡ https://lnkd.in/dvJatmMx ------- ⬇⬇ Abstract ⬇⬇ Clinical prediction models estimate an individual’s risk of a particular health outcome. In this talk, I suggest that prediction model research has become the academic’s playground, and that machine learning and AI have only exacerbated this problem. I highlight common pitfalls, including model instability and poor calibration, and propose some solutions moving forwards. The talk is aimed at a broad audience. #ClinicalPredictionModels #MachineMearning #AI #PredictiveAlgorithms #Stability Cc. Aix-Marseille Université, IRD, Santé publique France, ISSPAM, Société Française de Santé Publique (SFSP), Collège de Liaison des Internes de Santé Publique (CLISP), EUPHA - European Public Health Association, EuroHealthNet, European Public Health Alliance (EPHA), Young WFPHA

My research vision is to produce AI algorithms that are both technically cutting edge and at the same time have real clinical applications that make a difference for patients. I'm super happy that recently my team (inc Ling Luo) had two major successes! My PhD student Xuan May's paper “ShapeFormer: Shapelet Transformer for Multivariate Time Series Classification” has been accepted by the top data mining conference KDD 2024. Her research improves time series classification by leveraging both class-specific and generic features. The ShapeFormer algorithm is currently being tested on ventilator data from the Austin hospital with the aim to help Motor Neurone Disease Patients breathe better. My PhD student Mojgan Kouhounestani received an Honourable Mention Best Paper at one of the top health informatics conferences - the International Conference on Information and Communication Technologies for Ageing Well and e-Health (ICT4AWE). Her work is about “Enhancing predictive Modelling in Emergency Departments” and showcases a new algorithm that can handle very high (>10k) dimensional categorical features. The work is based on our collaboration with the Royal Melbourne Hospital and will inform the triage process at their Emergency Department. These kinds of results keep motivating me to dig deeper! If you would like a copy of the articles, please contact me. #AI #Healthinformatics #unimelb

Join us for a groundbreaking presentation on Foresight-SDE: A revolutionary #generativeAI model transforming healthcare prediction! Chris Tomlinson, MD, Senior Research Fellow, University College London, is leading a collaboration between British Heart Foundation, King's College London, NHS England and UCL to develop a first-of-its-kind #AI system that combines large language models with population-scale health data to revolutionize how to predict #health outcomes across and beyond the COVID-19 pandemic. What makes Foresight unique? • Universal risk prediction from a single model • Learns from comprehensive medical histories • Can simulate "what-if" healthcare scenarios Imagine being able to predict how modifying risk factors might affect a patient's future health risks - that's what Foresight can do! For example, it can model how optimising cardiovascular medications could impact an individual's risk of stroke or severe COVID-19. What: AWS’ Genomics and Cloud Computing: Accelerating Scientific Research conference - register here: https://lnkd.in/ejQKSNfg When: Wednesday, 11th December Where: The Orchard, Level 8, One Great Cumberland Place, London, W1H 7AL Don't miss this opportunity to learn about the future of personalized medicine and preventive healthcare. The possibilities for clinical applications are truly exciting. #Healthcare #Research #innovation #digitalmarketing #technology --- Join the conversation: How do you think AI will transform preventive medicine? Share your thoughts below! 👇

Digital innovation: the practical application of AI in healthcare. 🩺 Henley’s Professor of Informatics and Digital Health, Weizi (Vicky) Li’s work focuses on using informatics, data science, machine learning and digital information systems to solve real-world healthcare challenges – specifically how AI can better detect the onset of Inflammatory Arthritis (IA), a disease that 10 million people in the UK currently suffer with. 🏥 Professor Li recently took part as a guest speaker in a series of seminars across Asia and the US delving into case studies of developing AI applications in the NHS, as well as discussing AI challenges in the real-world healthcare ecosystem, including: ✅ AI usability and integration within healthcare ✅ Effective leveraging of healthcare data #AIinHealthcare #HealthTech #Innovation #AIBreakthroughs #UKHealthcare #AIApplications #HealthcareData #AIImplementation

Revolutionizing Disease Prediction with Cutting-Edge AI! As healthcare professionals, we're constantly seeking innovative ways to improve patient outcomes. That's why I'm excited to share the groundbreaking findings of a recent study that could transform the future of disease prediction. The research paper, published in the prestigious Journal of Electrical Systems, introduces a pioneering multi-disease prediction model that leverages the power of machine learning and deep learning techniques. Led by a team of renowned researchers, this study tackles a crucial challenge: accurately predicting the onset of various health conditions, even in complex, multi-morbidity scenarios. At the heart of this model lies the Stabilized Energy Valley Optimization with Enhanced Bounds (SEV-EB) algorithm, a novel approach to feature selection that enhances the model's stability and accuracy. By seamlessly integrating statistical features, deep features, and optimally selected features, the model is able to capture the nuanced patterns within healthcare data. But the real game-changer is the HSC-AttentionNet architecture. This innovative network combines the power of deep temporal convolution and long-term memory capabilities, allowing the model to predict diseases with unparalleled precision. The results are nothing short of remarkable. Tested on comprehensive datasets, the model achieved a staggering 95% accuracy and 94% F1-score in predicting a wide range of disorders. This groundbreaking performance surpasses traditional methods, setting a new standard in the field of disease prediction. The implications of this research extend far beyond the academic realm. By harnessing the wealth of information embedded in Electronic Health Records (EHR), this model presents a paradigm shift in healthcare interventions. Clinicians and healthcare providers can now access timely and accurate insights, empowering them to implement personalized treatment plans and proactive preventive measures. As we continue to navigate the ever-evolving landscape of healthcare, innovations like this hold the key to revolutionizing patient outcomes. I'm excited to see how this research will inspire and shape the future of predictive analytics in the medical field. Image source: https://lnkd.in/e3mNd3ci Follow Leon Zheng and let's come together and explore the vast potential of this breakthrough technology. I'm eager to hear your thoughts and perspectives on how we can leverage these advances to transform the way we deliver healthcare. #healthcare #machinelearning #diseaseprediction #AI #innovation #research #patientcare

🌟 Thrilled to have presented my research on early sepsis detection at the Global AI Summit 2024, organized by Bennett University and the Times Group! 🌟Our paper, "Detection of Sepsis using Machine Learning: A Comprehensive Evaluation of Model Performance ," highlights the potential of predictive modeling to enhance patient outcomes by identifying sepsis at an early stage. Leveraging advanced ML models, including Decision Trees, Random Forest, and Extra Trees, we achieved promising results that can contribute to improving healthcare analytics. A big thank you to my incredible co-authors mallemputi sai teja, Rohan Satya Balaji Bandaru and faculty mentor Dr. Sophiya E for their guidance and support throughout this journey. Their insights have been invaluable in making this research a success. 🙏Looking forward to exploring more in the realm of AI-driven healthcare innovations! 🚀 #AIML #HealthcareAnalytics #SepsisDetection #GlobalAISummit2024 #MachineLearning #PredictiveModeling #DataScience #BennettUniversity #Research

[#Webinar] D-1 until our next QuanTIM Webinar! ⤵ 💬 Prediction models in healthcare: a playground for researchers 🗣 Richard Riley (University of Birmingham) 💻 https://lnkd.in/dixBFbaU ------- Interested in a QuanTIM Webinar's panelist spot and a funding for a scientific conference or a research stay in one of the SESSTIM’s teams? Check-out our PhD Student Award! 🎓🏅 ➡ https://lnkd.in/dvJatmMx ------- ⬇⬇ Abstract ⬇⬇ Clinical prediction models estimate an individual’s risk of a particular health outcome. In this talk, I suggest that prediction model research has become the academic’s playground, and that machine learning and AI have only exacerbated this problem. I highlight common pitfalls, including model instability and poor calibration, and propose some solutions moving forwards. The talk is aimed at a broad audience. #ClinicalPredictionModels #MachineMearning #AI #PredictiveAlgorithms #Stability Cc. Aix-Marseille Université, IRD, Santé publique France, ISSPAM, Société Française de Santé Publique (SFSP), Collège de Liaison des Internes de Santé Publique (CLISP), EUPHA - European Public Health Association, EuroHealthNet, European Public Health Alliance (EPHA), Young WFPHA

I’m pleased to announce that I have completed the course on 𝗙𝗮𝗶𝗿𝗻𝗲𝘀𝘀 𝗮𝗻𝗱 𝗥𝗲𝘀𝗽𝗼𝗻𝘀𝗶𝗯𝗶𝗹𝗶𝘁𝘆 𝗶𝗻 𝗛𝘂𝗺𝗮𝗻-𝗔𝗜 𝗜𝗻𝘁𝗲𝗿𝗮𝗰𝘁𝗶𝗼𝗻 𝗶𝗻 𝗠𝗲𝗱𝗶𝗰𝗮𝗹 𝗦𝗲𝘁𝘁𝗶𝗻𝗴𝘀 by The Alan Turing Institute. This program highlighted the transformative potential of AI in healthcare, particularly in enhancing patient care and alleviating workload pressures. 𝗞𝗲𝘆 𝗶𝗻𝘀𝗶𝗴𝗵𝘁𝘀 included the challenges of 𝗮𝗹𝗴𝗼𝗿𝗶𝘁𝗵𝗺𝗶𝗰 𝘁𝗿𝘂𝘀𝘁𝘄𝗼𝗿𝘁𝗵𝗶𝗻𝗲𝘀𝘀, 𝗺𝗼𝗱𝗲𝗹 𝗯𝗶𝗮𝘀, 𝗮𝗻𝗱 𝗰𝗼𝗴𝗻𝗶𝘁𝗶𝘃𝗲 𝗯𝗶𝗮𝘀𝗲𝘀 that can impact AI-augmented clinical reasoning and decision-making. UK NHS AI-Lab and Health Education England underscores the need for greater education among healthcare professionals to build confidence and ensure fairness in AI applications. GE HealthCare Sanofi Novartis Google Health Philips Siemens Healthineers Mayo Clinic NHS #AI #Health #Human #Interaction #Medical #Settings #ResponsibleAI #Bias #TrustworthyAI