AI Model Offers Faster, Cheaper Prediction of Breast Cancer Recurrence

Researchers at New York University have created a multi‑modal artificial‑intelligence test that predicts the risk of breast‑cancer recurrence using routine pathology slides combined with basic clinical data such as tumor stage, age and hormone‑receptor status. Evaluated on more than 3,500 patients drawn from 15 populations across seven countries, the model achieved accuracy comparable to or better than widely used genomic assays, while delivering results in hours instead of weeks and preserving tissue for future testing. The approach performed well for hormone‑receptor‑positive disease and also showed promise for triple‑negative and HER2‑positive tumors, sub‑types that currently lack reliable genomic tests. The authors note that the test must still be validated in completed randomized clinical trials before it can become standard care. Several investigators hold equity in Ataraxis AI, the company that will commercialise the technology, and NYU retains intellectual‑property interests.

The development could reduce the cost and turnaround time of recurrence risk assessment, potentially guiding more precise treatment decisions for the millions of women diagnosed with breast cancer each year.