Abstract #4113

Determine Breast Density from T1-weighted Images using a 3D Convolutional Neural Network

Mario Serrano-Sosa¹, Jie Ding¹, Karl Spuhler¹, and Chuan Huang^2,3

¹Biomedical Engineering, Stony Brook University, Stony Brook, NY, United States, ²Psychiatry, Stony Brook Medicine, Stony Brook, NY, United States, ³Radiology, Stony Brook Radiology, Stony Brook, NY, United States

Using MRI-based breast density (BD) measurements, we trained a convolutional neural network to take a T1-weighted MRI and output a BD value. Our CNN yielded a 3.1% error for the entire data set and 4.1% error in the testing set. Having a CNN work on T1-weighted MRI to quantify BD will allow BD to be a part of every practice.

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

Join Here