Derek Driggs - Barriers to Deploying Deep Learning Models During the COVID-19 Pandemic
Presentation given by Derek Driggs on 28 April 2021 in the one world seminar on the mathematics of machine learning on the topic "Barriers to Deploying Deep Learning Models During the COVID-19 Pandemic".
Abstract: A promising application for deep learning models is in assisting clinicians with interpreting X-ray and CT scans, especially when treating respiratory diseases. At the onset of the COVID-19 pandemic, radiologists had to quickly learn how to identify a new disease on chest X-rays and CT scans, and use this information to decide how to allocate scarce resources like ventilators. Researchers around the world developed deep learning models to help clinicians with these decisions, and some models were deployed after only three weeks of testing.
Our group reviewed over 1,000 studies that introduce deep learning models for interpreting chest X-rays or CT scans of COVID-19 patients to determine which models, if any, have the potential to help clinicians during the pandemic. In this talk, I will present our findings and discuss how this pandemic could inform researchers creating deployable deep learning models in healthcare.
This talk is based on the paper [1].
[1] Roberts, M., Driggs, D., Thorpe, M., and the AIX-COVNET Collaboration. "Common pitfalls and recommendations for using machine learning to detect and prognosticate for COVID-19 using chest radiographs and CT scans”. Nat. Mach. Intel. 3, 199–217 (2021).
Видео Derek Driggs - Barriers to Deploying Deep Learning Models During the COVID-19 Pandemic канала One world theoretical machine learning
Abstract: A promising application for deep learning models is in assisting clinicians with interpreting X-ray and CT scans, especially when treating respiratory diseases. At the onset of the COVID-19 pandemic, radiologists had to quickly learn how to identify a new disease on chest X-rays and CT scans, and use this information to decide how to allocate scarce resources like ventilators. Researchers around the world developed deep learning models to help clinicians with these decisions, and some models were deployed after only three weeks of testing.
Our group reviewed over 1,000 studies that introduce deep learning models for interpreting chest X-rays or CT scans of COVID-19 patients to determine which models, if any, have the potential to help clinicians during the pandemic. In this talk, I will present our findings and discuss how this pandemic could inform researchers creating deployable deep learning models in healthcare.
This talk is based on the paper [1].
[1] Roberts, M., Driggs, D., Thorpe, M., and the AIX-COVNET Collaboration. "Common pitfalls and recommendations for using machine learning to detect and prognosticate for COVID-19 using chest radiographs and CT scans”. Nat. Mach. Intel. 3, 199–217 (2021).
Видео Derek Driggs - Barriers to Deploying Deep Learning Models During the COVID-19 Pandemic канала One world theoretical machine learning
Показать
Комментарии отсутствуют
Информация о видео
29 апреля 2021 г. 3:22:57
00:53:23
Другие видео канала
Soheil Kolouri - Wasserstein Embeddings in the Deep Learning EraThe first 20 hours -- how to learn anything | Josh Kaufman | TEDxCSUDeep Learning: A Crash Course (2018) | SIGGRAPH CoursesMatthew Colbrook - Smale’s 18th Problem and the Barriers of Deep LearningEKG Interpretation | Master Fundamentals of ECG | Electrocardiography | Dr Najeeb LecturesMatteo Bianchi (U. Pisa) - Model based and data driven approaches for human sensory-motor systemThe State of Healthcare Policy: from COVID-19 to Medicare for AllShoulder Impingement, Pain and Injury Rehabilitation - Full Seminar Presentation | Tim KeeleyRobert Nowak - What Kinds of Functions Do Neural Networks Learn?Tatiana Bubba (University of Bath) - Learning a microlocal prior for limited angle tomographyHongyang Zhang - Understanding and improving generalization in multitask and transfer learningCAIDA Talk - June 21, 2022 - Justin HartTyrus Berry - Beyond Regression: Operators and Extrapolation in Machine LearningAlessandro Scagliotti - Deep Learning Approximation of Diffeomorphisms via Linear-Control SystemsAndrew Stuart - Learning Linear OperatorsLoucas Pillaud-Vivien - the role of noise in non-convex machine learning dynamicsChris Budd and Simone Appella - R-Adaptivity, Deep Learning and Optimal TransportNicola Pedreschi: The temporal rich club phenomenon