3444

View Abstract / View Presentation

Interpreting a machine learning model: radiomics in cervical spondylotic myelopathy postoperative recovery prediction

Mengze Zhang¹, Hanqiang Ouyang¹, Dan Jing¹, Jiangfang Liu¹, Chunjie Wang¹, Huishu Yuan¹, and Liang Jiang¹
¹Peking University Third Hospital, Beijing, China

Radiomics features are potential indicators for predicting CSM patients’ postoperative recovery. The run variance was the most important feature in the radiomic-feature-based model.

Fig 1. The pipeline of segment, normalization, scaling, feature extraction, feature selection, and model building.

Fig 4. The partial dependence plots, the local dependence plots, and the accumulated dependence plots for each feature.