Background: Patients with ER-positive/HER2-negative early breast cancer of low or intermediate grade are generally considered to have a favourable prognosis, yet a number will recur. This risk is poorly captured by standard clinicopathological factors. AI-based medical devices for whole slide image (WSI) analyses used for risk-stratification often predict on slide-level thereby overlooking spatial localisation of the biomarkers.
A spatial AI-based biomarker with independent prognostic value in early low- and intermediate- risk breast cancer
calendar_today
October 9, 2026
person
Boissin, C., Xiang, Y., Liu, B., Vallon-Christersson, J., Walker, E., Hartman, J., Rantalainen, M.
domain
medrxiv