28 min
#070 AI Sepsis Prediction — Dr Michael Moor (ETH Zurich) Big Picture Medicine
-
- Medicine
Dr Michael Moor is a medical doctor doing his PhD in the Machine Learning and Computational Biology Lab at ETH Zurich.
Some of his most interesting research looks at predicting sepsis using machine learning approaches.
In case you’re unfamiliar, sepsis is a life threatening condition in which the body’s natural defences against an infection go into overdrive.
Mortality from septic shock can be up to 50% and every minute counts. In fact, for every hour treatment is delayed — mortality increases by 7.6%.
We started off by talking about a paper published in JAMA earlier this year. It assessed a popular sepsis prediction tool and found that it wasn’t very good. It was delivering so many false positives, that doctors would need to assess up to 109 patients flagged by the tool just to find one patient who was actually septic…
Michael's Preprint: Predicting Sepsis in Multi-site, Multi-national Intensive Care Cohorts Using Deep Learning
Michael Moor*, Nicolas Bennet∗, Drago Plecko∗, Max Horn∗,Bastian Rieck, Nicolai Meinshausen, Peter Bühlmann and Karsten Borgwardt.
JAMA 2021 Paper: External Validation of a Widely Implemented Proprietary Sepsis Prediction Model in Hospitalized Patients
Andrew Wong, Erkin Otles, Meng; John P. Donnelly, Andrew Krumm, Jeffrey Mccullough, Olivia Detroyer-cooley, Justin Pestrue, Mecon Marie Phillips, Judy Konye, Carleen Penoza, Muhammad Ghous, Karandeep Singh
You can find me on Twitter @MustafaSultan and subscribe to my newsletter on www.musty.io
Dr Michael Moor is a medical doctor doing his PhD in the Machine Learning and Computational Biology Lab at ETH Zurich.
Some of his most interesting research looks at predicting sepsis using machine learning approaches.
In case you’re unfamiliar, sepsis is a life threatening condition in which the body’s natural defences against an infection go into overdrive.
Mortality from septic shock can be up to 50% and every minute counts. In fact, for every hour treatment is delayed — mortality increases by 7.6%.
We started off by talking about a paper published in JAMA earlier this year. It assessed a popular sepsis prediction tool and found that it wasn’t very good. It was delivering so many false positives, that doctors would need to assess up to 109 patients flagged by the tool just to find one patient who was actually septic…
Michael's Preprint: Predicting Sepsis in Multi-site, Multi-national Intensive Care Cohorts Using Deep Learning
Michael Moor*, Nicolas Bennet∗, Drago Plecko∗, Max Horn∗,Bastian Rieck, Nicolai Meinshausen, Peter Bühlmann and Karsten Borgwardt.
JAMA 2021 Paper: External Validation of a Widely Implemented Proprietary Sepsis Prediction Model in Hospitalized Patients
Andrew Wong, Erkin Otles, Meng; John P. Donnelly, Andrew Krumm, Jeffrey Mccullough, Olivia Detroyer-cooley, Justin Pestrue, Mecon Marie Phillips, Judy Konye, Carleen Penoza, Muhammad Ghous, Karandeep Singh
You can find me on Twitter @MustafaSultan and subscribe to my newsletter on www.musty.io
28 min