Peter Attia· MD
he built it off around 16700 MRI biopsy linked cases but now that has grown because it's it's always learning
Loading…
The headline is broadly defensible, but the qualifications matter. Effect sizes vary by population, the strongest claims rest on shorter trials, and credible voices push back on how it's typically framed.
he built it off around 16700 MRI biopsy linked cases but now that has grown because it's it's always learning
Every Sunday: the week’s new conflicts and verdict changes — and nothing else.
Native comments, Twitter mentions, and Reddit threads about this claim — surfaced together so the conversation isn't fragmented across platforms.
Bookmarking — the dossier-vs-overview split is the right call. Most of the time I want overview; sometimes I want receipts.
Would love a "what would change this verdict" RSS feed. Sign me up if it exists.
he analyzed and built a neural network real-time predictor for all patients not just Northwestern Medicine patients but we used all the Northwestern Medicine patients in our system who had had Phi MRI and a biopsy and looked at all their outcomes so there's some selection by us because we didn't include people who didn't have a biopsy but in general we took all these people and we created a neural network real-time predictor for what's your absolute risk for having prostate cancer in general but more specifically prostate cancers that would require treatment