期刊
2020 25TH INTERNATIONAL CONFERENCE ON PATTERN RECOGNITION (ICPR)
卷 -, 期 -, 页码 5214-5221出版社
IEEE COMPUTER SOC
DOI: 10.1109/ICPR48806.2021.9412934
关键词
magnetic resonance images; convolutional neural networks; deep learning; image semantic segmentation
类别
资金
- Regional Ministry of Health of the Valencian Region, under MIDAS project from BIMCV-Generalitat Valenciana [ACIF/2018/285]
This study focuses on automatically segmenting MR images of the lumbar spine using convolutional neural networks, aiming to detect and delimit structural elements like vertebrae, intervertebral discs, nerves, and blood vessels. The proposed network architectures are variants of the U-Net, incorporating spatial attention models, deep supervision, multi-kernels, and inception-based blocks to achieve better results than the standard U-Net.
This work addresses the problem of automatically segmenting the MR images corresponding to the lumbar spine. The purpose is to detect and delimit the different structural elements like vertebrae, intervertebral discs, nerves, blood vessels, etc. This task is known as semantic segmentation. The approach proposed in this work is based on convolutional neural networks whose output is a mask where each pixel from the input image is classified into one of the possible classes. Classes were defined by radiologists and correspond to structural elements and tissues. The proposed network architectures are variants of the U-Net. Several complementary blocks were used to define the variants: spatial attention models, deep supervision and multi-kernels at input, this last block type is based on the idea of inception. Those architectures which got the best results are described in this paper, and their results are discussed. Two of the proposed architectures outperform the standard U-Net used as baseline.
作者
我是这篇论文的作者
点击您的名字以认领此论文并将其添加到您的个人资料中。
推荐
暂无数据