Paper: 3D intracranial artery segmentation using a convolutional autoencoder clatfd

Paper: 3D intracranial artery segmentation using a convolutional autoencoder

Posted on 2017-10-10 by admin 1820 0

3D intracranial artery segmentation using a convolutional autoencoder

Li Chen¹, Yanjun Xie², Jie Sun³, Niranjan Balu³, Mahmud Mossa-Basha³, Kristi Pimentel³, Thomas S. Hatsukami⁴, Jenq-Neng Hwang¹, Chun Yuan^3†

Department of 1. Electrical Engineering; 2. Mechanical Engineering; 3. Radiology; 4. Surgery

University of Washington

Seattle, WA, 98195, USA

{cluw, yanjunx. sunjie, ninja, mmossab, kristidb, tomhat, hwang, cyuan }@uw.edu

†Corresponding author

Abstract— Automated segmentation of intracranial arteries on magnetic resonance angiography (MRA) allows for quantification of cerebrovascular features, which provides tools for understanding aging and pathophysiological adaptations of the cerebrovascular system. Using a convolutional autoencoder (CAE) for segmentation is promising as it takes advantage of the autoencoder structure in effective noise reduction and feature extraction by representing high dimensional information with low dimensional latent variables. In this paper, we trained an 8-layer CAE to learn a 3D segmentation model of intracranial arteries from 49 cases of MRA data. After parameter optimization and prediction refinement, our trained model was shown to perform better than the three traditional segmentation methods in both binary classification and visual evaluation.

Keywords- convolutional autoencoder; deep neural network; machine learning; artery segmentation; Magnetic Resonance Angiography

The IEEE International Conference on Bioinformatics and Biomedicine (BIBM) 2017

http://ieeexplore.ieee.org/document/8217741/

B245.pdf