Paige launched an AI tool that screens 505 genes directly from H and E stained pathology slides, without requiring next generation sequencing as a first step. The model predicts a comprehensive genomic profile from routine morphology alone and flags potential biomarkers for follow up. Rather than replacing sequencing, it is positioned as a triage layer that decides which cases warrant the full molecular workup. Paige is applying the same foundation model approach it built for cancer detection to molecular prediction.
Our readPathology labs and oncology practices can screen for hundreds of mutations off slides they already produce, compressing the time between biopsy and a targeted therapy decision. The economic argument is sequencing avoidance: labs run the expensive molecular test on the subset the model flags rather than on everything. Buyers should ask which of the 505 genes have been validated against sequencing ground truth and at what sensitivity, because a screening tool that misses a targetable mutation carries a different risk profile than one that over refers.