Abstract #2214

Differentiation of glioblastoma and primary central nervous system lymphoma by using MR image-based texture features

Akira Kunimatsu¹, Kouhei Kamiya², Yasushi Watanabe², Yuichi Suzuki², Natsuko Kunimatsu³, Harushi Mori¹, and Osamu Abe¹

¹Department of Radiology, The University of Tokyo, Tokyo, Japan, ²Department of Radiology, The University of Tokyo Hospital, Tokyo, Japan, ³Department of Radiology, International University of Health and Welfare Mita Hospital, Tokyo, Japan

We evaluated the feasibility of machine learning-based differentiation between glioblastoma and primary central nervous system lymphoma by using texture features of post-contrast MR images. Cross validation showed that more than 80% of teacher data were correctly assigned. Trial data comprised of atypical image variants were correctly assigned in up to 78.6% by the best classifiers.

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

Join Here