Next steps in research for AI prostate cancer mapping

In this video, Shyam Natarajan, PhD, discusses future research involving Unfold AI. He is an author of the recent BJUI Compass study, “Extracapsular extension risk assessment using an artificial intelligence prostate cancer mapping algorithm.” Natarajan is the founder and CEO of Avenda Health.

Transcription:

What are the next steps in developing and refining AI-generated cancer maps for prostate cancer?

[Although] we were very excited about the finding in this study showing that Unfold AI could be superior to MRI in detecting extracapsular disease, our next step is to validate this in a multi-institutional, prospective manner. [Although] Unfold AI was trained on multi-institutional data, this particular study was concerning a single institution's experience, and so the next research phase for AI cancer maps and technologies like Unfold AI really is to understand, now we have a way to map prostate cancer, sort of within and potentially around the prostate, what other types of correlates or covariates can we predict? Can we predict biochemical recurrence? Can we predict long-term success or failure? I think those are very fruitful avenues of research.

Is there anything you would like to add?

This is exciting because Unfold AI is one of the first multimodal AI systems for patients with prostate cancer. I think it's a testament to our collaborators and the senior author of the study, Dr. Wayne Brisbane, of being able to acquire a very robust and large data set to perform this kind of unique analysis, because typically, AI-based systems use consensus panels of experts as their sort of final ground truth or gold standard, and that can only show you, are you as good as an expert clinician? Whereas in this study and other studies of this nature, we actually use the surgical pathology, sort of the ground truth, tissue information as our gold standard, showing that we could potentially be superior to the standard of care.

This transcript was edited for clarity.