Abstract #3068

Enhanced isocitrate dehydrogenase mutation classification in lower grade gliomas with deep learning and CEST and MTC MRI

Zuo Wang^1,2, Meiyappan Solaiyappan¹, Qihong Rui³, Hye-Young Heo¹, Zhibo Wen³, Gregory Hager², Jinyuan Zhou¹, and Shanshan Jiang¹

¹Department of Radiology, Johns Hopkins University， School of Medicine, Baltimore, MD, United States, ²Whiting School of Engineering, Johns Hopkins University, Baltimore, MD, United States, ³Zhujiang Hospital,Southern Medical University, guangzhou, China

we assess the feasibility of using molecular MRI with deep learning to differentiate IDH mutation status in patients with lower grade gliomas. Two separate deep learning models were used to analyze routine MRI and molecular MRI, and then, a combined model was also devised. 18% and 11% higher AUCs were obtained by the combined system, with respect to the routine MRI subsystem and the molecular MRI subsystem, respectively. Molecular MRI with deep learning algorithm demonstrated a great potential to diagnose IDH mutation status, which could be implemented as a robust approach to enhance routine MRI classification performance.

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

Join Here