Fig. 2From: Development of a machine learning-based radiomics signature for estimating breast cancer TME phenotypes and predicting anti-PD-1/PD-L1 immunotherapy responseLesion segmentation. The core tumor area in MRI imaging was measured and outlined manually by ITK-SNAP software in different slices. The boundary of the tumor was automatically identified by Python. A peripheral region was then established inside and outside 2 mm away from the defined boundary, consequently forming a 4-mm thickness ring to represent the extension of the tumor infiltrationBack to article page