There are some exiting studies that came out in the last 1-2 weeks. Topol's team compared the accuracy of deep learning algorithms to that of medical professionals in making diagnoses on medical images. Besides giving a comprehensive overview of the field (without any hype), they pointed out 5 major issues with those studies. 1) most studies took the approach of assessing diagnostic accuracy in isolation not reflecting clinical practice 2) there were very few prospective studies done in real clinical environments 3) a wide range of metrics were employed to report diagnostic performance 4) there is inconsistency over key terminology 5) most studies did not do an out-of-sample validation for both health-care professionals and deep learning algorithms Read it here: https://lnkd.in/gFVPT4u Also, some other news: Machine learning could offer faster, more precise results for cardiac MRI scans http://bit.ly/2oPXnqA Prediction system for risk of allograft loss in patients receiving kidney transplants: international derivation and valhttps://lnkd.in/gyTbbUU #themedicalfuturist #digitalhealth #future #healthcare #medicine #technology #AI #artificialintelligence #chess #deeplearning #machinelearning …see more