Hepatic proton density fat fraction (PDFF) measurement plays an important role in the assessment of chronic diffuse liver diseases, while it is always time consuming and lack of good reproducibility and repeatability. To address this problem, we introduced a five-ROI sampling approach based on a convolutional neural network that provided high dice coefficient (DC) of whole liver segmentation, good correlation and quicker compared with manual operation.
This abstract and the presentation materials are available to members only; a login is required.