Fig. 4: Performance of the deep learning model to assess patient prognosis in the training and validation cohorts. | npj Precision Oncology

Fig. 4: Performance of the deep learning model to assess patient prognosis in the training and validation cohorts.

From: Biologically interpretable multi-task deep learning pipeline predicts molecular alterations, grade, and prognosis in glioma patients

Fig. 4: Performance of the deep learning model to assess patient prognosis in the training and validation cohorts.

a Kaplan–Meier curves show the overall survival (OS) of patients in the low-DPS and high-DPS subtypes. The OS of patients in the high-DPS group was significantly shorter compared to the low-DPS group in the training and validation cohorts. b Calibration curves depict the agreement between the actual and predicted survival probabilities. c Time-dependent receiver operating characteristic curves at the given points of 1 year, 3 year, and 5 year. DPS, deep prognosis score.

Back to article page