Skip to main content
. 2023 Jan 25;25(4):776–787. doi: 10.1007/s11307-023-01803-y

Fig. 1.

Fig. 1

Flowchart illustrating the workflow for radiomics feature extraction and machine learning–augmented analysis. Example depicts a 50-year-old female with undifferentiated pleomorphic sarcoma of the left calf pre- and post-NAC. Following manual segmentation of 3D-ROIs, data were then extracted from co-registered sequences of interest via our institutional radiomics pipeline. Machine learning models using random forest (RF) and real adaptive boosting (AdaBoost) methods were constructed using a tenfold cross-validation procedure. NAC, neoadjuvant chemotherapy; ROI, region of interest; RF, random forest; AdaBoost, real adaptive boosting.