Abstract #1707

Discrimination of Malignant and Benign Breast Lesions Using Machine Learning on Non-Gaussian Diffusion MRI Parameters

Muge Karaman^1,2, Yangyang Bu^3,4, Zheng Zhong^1,2, Shiwei Wang^3,4, Changyu Zhou^3,4, Weihong Hu^3,4, Mark Balich¹, Maosheng Xu^3,4, and Xiaohong Joe Zhou^1,2,5

¹Center for Magnetic Resonance Research, University of Illinois at Chicago, Chicago, IL, United States, ²Department of Bioengineering, University of Illinois at Chicago, Chicago, IL, United States, ³The First Clinical Medical College, Zhejiang Chinese Medical University, Hangzhou, China, ⁴Department of Radiology, The 1st Affiliated Hospital of Zhejiang Chinese Medical University, Hangzhou, China, ⁵Departments of Radiology and Neurosurgery, University of Illinois at Chicago, Chicago, IL, United States

Breast cancer is the second leading cause of cancer death among women in the US. Recognizing the complexity of cancerous tissue, several non-Gaussian diffusion MRI models, such as the continuous-time random-walk (CTRW) model, were suggested to probe the underlying tissue environment. In this study, we employed a support-vector-machine-based analysis on the histogram features of CTRW model parameters to differentiate malignant and benign breast lesions. This multi-parameter multi-feature approach provided the best diagnostic performance compared to the conventional single-parameter or single-feature analysis techniques. The combination of machine-learning with non-Gaussian diffusion MRI can facilitate comparable diagnostic performance to that of dynamic-contrast-enhanced MRI.

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

Join Here