ScanRacer makes data annotation in #MedicalImaging fun thanks to @FAU_Germany's students and @DataDonors! #DeepLearning #MachineLearning @SpirosMargaris @informatikradar @GaryMarcus @drfeifei @antgrasso @SwissCognitive @Montreal_AI @ipfconline1 @TamaraMcCleary @KirkDBorne https://t.co/Nw9xZaDQnK
The great irony of #MachineLearning for medical applications: On one hand, everything is "data-driven", "makes no assumptions" etc. On the other, diagnostic labels are taken as ground truth.