Disparate Censorship & Undertesting: A Source of Label Bias in Clinical Machine Learning

Open in new window