How to compare the time-dependent precision recall (PR) receiver operating curve (ROC) values for two cox regression models at multiple time points?

To compare two time-dependent AUC values, I would use the compare function of the pROC R library. For a long time I thought DeLong's test as an alternative but it was originally developed for comparing standard (binary classification) AUROCs, where the outcome is binary and independent of time. Time-dependent AUROCs, however, incorporate censoring and time-to-event information.

The PRROC R library works well to calculate time-dependent ROC values for PR curves (no statistical comparison) and the pr.test function of the usefun R library to compare two PR values (not time-dependent!). However, there are no well-documented libraries to combine both so to statistically compare the time-dependent PRROCs.

Any ideas how to implement this in R or python? Could not find any other threads addressing this specific issue.

How to compare the time-dependent precision recall (PR) receiver operating curve (ROC) values for two cox regression models at multiple time points?

To compare two time-dependent AUC values, I would use the compare function of the timeROC R library. For a long time I thought DeLong's test as an alternative but it was originally developed for comparing standard (binary classification) AUROCs, where the outcome is binary and independent of time. Time-dependent AUROCs, however, incorporate censoring and time-to-event information.

The PRROC R library works well to calculate time-dependent ROC values for PR curves (no statistical comparison) and the pr.test function of the usefun R library to compare two PR values (not time-dependent!). However, there are no well-documented libraries to combine both so to statistically compare the time-dependent PRROCs.

Any ideas how to implement this in R or python? Could not find any other threads addressing this specific issue.