Meeting Banner
Abstract #3110

Development and validation of an MRI-based radiomics nomogram for distinguishing multiple myeloma from bone metastasis of breast cancer

Miyuki Takasu1, Konagi Takeda1, Saki Kawai1, Nobuko Tanitame1, Akihisa Tamura1, Makoto Iida1, Yuji Akiyama2, and Kazuo Awai2
1Diagnostic Radiology, Hiroshima City Hiroshima Citizens Hospital, Hiroshima, Japan, 2Diagnostic Radiology, Hiroshima University Hospital, Hiroshima, Japan

Synopsis

We established and validated the feasibility of an MRI-based radiomics nomogram to differentiate between multiple myeloma (MM) and bone metastasis of breast cancer (BC). Regions of interest were drawn for the largest bone lesion per patient on T1- and T2-weighted images of patients with MM (n = 85) and BC (n = 70) from two institutions. The machine learning model with logistic regression resulted in the best performance for differentiating MM from BC with both sequences in the test cases. Our proposed clinical radiomics analysis shows promise in differentiating MM from bone metastasis of BC.

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

Join Here