Fig. 2 | Scientific Reports

Fig. 2

From: AI-derived CT biomarker score for robust COVID-19 mortality prediction across multiple waves and regions using machine learning

Fig. 2

Diagnostic performance of M3-score in Training Set (n = 252, UZ Brussel, COVID-19 Wave 1, March 1 till May 15 2020). (A) Receiver Operating Characteristic (ROC) curve. (B) Violin plots of calculated M3-scores for 213 surviving patients and 39 deceased patients, with interpretative result intervals, associated likelihood ratios (LRs), and calculated posttest probabilities using mortality prevalence as the pretest probability. (C) Pie charts showing the proportion of samples categorized by LR group. (D) Posttest probability curves for mortality as a function of pretest probability and M3-score result intervals. Arrows indicate posttest probabilities calculated from mortality prevalence as the pretest probability.

AUC, area under the curve; CI, confidence interval; LR, likelihood ratio; PPRE, pretest probability; PPOST, posttest probability.

Back to article page