Abstract #3070

Radiomics with 3D MR fingerprinting: Influence of dictionary design on radiomic features and a potential mitigation strategy

Shohei Fujita^1,2, Akifumi Hagiwara¹, Koichiro Yasaka³, Hiroyuki Akai³, Akira Kunimatsu³, Shigeru Kiryu⁴, Issei Fukunaga¹, Shimpei Kato^1,2, Toshiaki Akashi¹, Koji Kamagata¹, Akihiko Wada¹, Osamu Abe², and Shigeki Aoki¹

¹Department of Radiology, Juntendo University, Tokyo, Japan, ²Department of Radiology, The University of Tokyo, Tokyo, Japan, ³Department of Radiology, The Institute of Medical Science, The University of Tokyo, Tokyo, Japan, ⁴Department of Radiological Sciences, International University of Health and Welfare, Narita, Japan

Understanding the dependencies of radiomic features on the dictionary step size in MR fingerprinting is crucial in the clinical implementation of MR fingerprinting-based radiomics. We investigated the influence of dictionary design in the radiomic analysis of MR fingerprinting and observed that radiomic features vary significantly for different designs. Moreover, special attention is required when using datasets containing maps generated from different dictionaries. We also demonstrated that an inverse quantization process can mitigate the influence of dictionary design on the radiomic features obtained with MR fingerprinting.

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

Join Here