Emre Yekedüz, Research Fellow in Medicine at Dana-Farber and Medical Oncologist and Faculty Member at Ankara Üniversitesi, shared an article by Yu Zhao on X:
“DeepGEM: A ground breaking AI model for predicting gene mutations directly from routine histological slides in lung cancer.
A large multi-center dataset (16 hospitals, 3637 patients) and TCGA data, DeepGEM demonstrated high accuracy (AUC up to 0.97) across excisional & aspiration biopsies, even generalizing to lymph node metastases.
This cost-effective, annotation-free method could revolutionize precision oncology by providing timely, accessible, and accurate mutation predictions—potentially guiding targeted therapy decisions.”
Deep learning using histological images for gene mutation prediction in lung cancer: a multicentre retrospective study.
Authors: Yu Zhao, et al.
