Deakin University

File(s) not publicly available

Segmentation of perivascular spaces using vascular features and structured random forest from 7T MR image

conference contribution
posted on 2022-09-28, 05:56 authored by Jun ZhangJun Zhang, Y Gao, S H Park, X Zong, W Lin, D Shen
Quantitative analysis of perivascular spaces (PVSs) is important to reveal the correlations between cerebrovascular lesions and neurodegenerative diseases. In this study, we propose a learning-based segmentation framework to extract the PVSs from high-resolution 7T MR images. Specifically, we integrate three types of vascular filter responses into a structured random forest for classifying voxels into PVS and background. In addition, we also propose a novel entropy-based sampling strategy to extract informative samples in the background for training the classification model. Since various vascular features can be extracted by the three vascular filters, even thin and low-contrast structures can be effectively extracted from the noisy background. Moreover, continuous and smooth segmentation results can be obtained by utilizing the patch-based structured labels. The segmentation performance is evaluated on 19 subjects with 7T MR images, and the experimental results demonstrate that the joint use of entropy-based sampling strategy, vascular features and structured learning improves the segmentation accuracy, with the Dice similarity coefficient reaching 66%.



10019 LNCS


61 - 68







Publication classification

E1.1 Full written paper - refereed

Title of proceedings

Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)