Meeting Banner
Abstract #2832

Radiomics and machine learning analysis of brain lesions in MOG-ab-positive and AQP4-ab-positive patients

Liqin Yang1, Wei Xia2, Haiqing Li1, Daoying Geng1, and Yuxin Li1

1Department of Radiology, Huashan Hospital, Fudan University, Shanghai, China, 2Academy for Engineering & Technology, Fudan University, Shanghai, China

Little is known about the differences between the lesion features of the recently discovered MOG-ab-positive and well-demonstrated AQP4-ab-positive patients till now. We studied the radiomics features of 747 lesions from AQP4 patients, and 295 lesions from MOG patients. Seventy radiomic features were calculated and compared. Features with significant between-group discrimination ability input to the classifier and trained. A radiomics signature was obtained for the discrimination of MOG-ab-positive and AQP4-ab-positive patients. These results provide valuable information for understanding of pathogenesis and imaging-based initial diagnosis in the two subsets of patients.

This abstract and the presentation materials are available to members only; a login is required.

Join Here