Abstract

Motivation: 21-gene assay is recommend to guide decision on the use of adjuvant chemotherapy. However, this test is expensive and time-consuming, so it is not widely used in clinic. Goal(s): To develop a radiomics model for predicting 21-gene recurrence score based on MRI intratumoral and peritumoral features. Approach: Prediction models in tumor and different peritumoral regions (2 mm, 4 mm, 6 mm, 8 mm, 10 mm) were established using machine learning method. Feature-fusion and logistics-regression methods were used to fuse information. Results: Combining 4-mm peritumoral model on T2WI (AUC = 0.66) with intratumoral and clinical-imaging features, the fusion model performed best (AUC = 0.75). Impact: To provide an alternative for patients who cannot afford Oncotype 21-gene assay and to reduce the medical costs for those who could afford it.