In Regard to Valdes et al

To the Editor: We read with interest the article by Valdes et al in which machine learning methods were used to predict biochemical failure after salvage brachytherapy for prostate cancer (1). Although the article addresses a relevant clinical issue, we are concerned that there are issues with the statistical methods that undermine the conclusion of good predictive performance. Given the many recent papers in our field applying machine learning to clinical questions, the themes discussed in this letter may apply more broadly.

https://ift.tt/2RWUqxS