Biomarker Diagnostic Threshold — ROC Curve Explorer

Every blood-test cutoff — troponin, PSA, HbA1c, a tumour marker — is a line drawn through two overlapping bell curves: one for healthy patients, one for diseased ones. This simulator draws both populations as real 3D point clouds spread along a biomarker-concentration axis, lets you drag the diagnostic threshold through the overlap, and recomputes the confusion matrix, sensitivity, specificity and PPV directly from the sampled points on every move. A separate analytic panel reports the theoretical AUC from the underlying Gaussian parameters, so you can see exactly how widening the gap between population means or tightening their spread turns a mediocre test into a decisive one — and why no single cutoff can maximise both sensitivity and specificity at once.