MRI-based radiomics model for predicting pathologic response to neoadjuvant therapy in HER2-positive breast cancer

Motivation: To investigate the value of an MRI-based radiomics model in evaluating the pathological complete response status of HER2-positive breast cancer after neoadjuvant therapy. Goal(s): Is the MRI-based radiomics model better than experienced radiologists for diagnosing pathologic complete response in patients with HER2-positive breast cancer after neoadjuvant therapy. Approach: Two experienced radiologists visually assessed the pCR status on post-neoadjuvant therapy breast MRI, and their diagnostic performance was compared with the MRI radiomics model. Results: An MRI-based radiomics model showed better classification performance than experienced radiologists for diagnosing pathologic complete response in patients with HER2-positive breast cancer after neoadjuvant therapy. Impact: MRI-based radiomics models may serve as a noninvasive predictor of pathological complete response. Accurate prediction of pCR is of great significance for timely termination of neoadjuvant therapy to avoid toxic and side-effects, and even for future exemption of breast surgery.