AI Colored Patches (IMAGE)
Caption
The AI model estimates recurrence risk by integrating information from digitized histopathology slides and patient clinical data. Regions containing cancer are extracted from whole-slide images and the model aggregates information across many tissue patches into a patient-level risk representation. The single green or orange patch is a schematic illustration of this aggregated patient-level representation. Patients are assigned to low-risk (green) or high-risk (orange) groups based on their AI-derived recurrence risk score. In the evaluation cohort, patients classified as high risk had substantially worse recurrence-free outcomes over time, with a 2.8-fold higher risk of recurrence than patients classified as low risk.
Credit
Ken Zeng, Ataraxis AI
Usage Restrictions
Use with this study only
License
Original content