Prediction models for diagnosis and prognosis of covid-19 infection: systematic review and critical appraisal | The BMJ

amarashar's bookmarks 2020-05-13

Summary:

Preliminary research suggests that most models using AI for COVID-19 diagnosis and prognosis are at high risk of bias and likely misleading. The researchers reviewed 31 models that use AI for online diagnosis, diagnosis based on CT scans and case prognosis prediction. According to their assessment, the reported performance of all 31 models was “probably optimistic.” Given the lack of external validation of most models, the authors recommend against relying on any current AI-based COVID-19 diagnostic tools.

Link:

https://www.bmj.com/content/369/bmj.m1328

From feeds:

Ethics/Gov of AI » amarashar's bookmarks

Tags:

covid

Date tagged:

05/13/2020, 09:56

Date published:

05/13/2020, 05:56