3D Dense Separated Convolution Module for Volumetric Medical Image Analysis
With the thriving of deep learning, 3D convolutional neural networks have become a popular choice in volumetric image analysis due to their impressive 3D context mining ability. However, the 3D convolutional kernels will introduce a significant increase in the amount of trainable parameters. Conside...
Main Authors: | , , |
---|---|
Format: | Article |
Language: | English |
Published: |
MDPI AG
2020-01-01
|
Series: | Applied Sciences |
Subjects: | |
Online Access: | https://www.mdpi.com/2076-3417/10/2/485 |
id |
doaj-b8801918ef7b4046aa7fa5b5f1a42467 |
---|---|
record_format |
Article |
spelling |
doaj-b8801918ef7b4046aa7fa5b5f1a424672020-11-25T01:27:50ZengMDPI AGApplied Sciences2076-34172020-01-0110248510.3390/app10020485app100204853D Dense Separated Convolution Module for Volumetric Medical Image AnalysisLei Qu0Changfeng Wu1Liang Zou2School of Electronics and Information Engineering, Anhui University, Hefei 236601, ChinaSchool of Electronics and Information Engineering, Anhui University, Hefei 236601, ChinaDepartment of Electrical and Computer Engineering, The University of British Columbia, Vancouver, BC V6T 1Z4, CanadaWith the thriving of deep learning, 3D convolutional neural networks have become a popular choice in volumetric image analysis due to their impressive 3D context mining ability. However, the 3D convolutional kernels will introduce a significant increase in the amount of trainable parameters. Considering the training data are often limited in biomedical tasks, a trade-off has to be made between model size and its representational power. To address this concern, in this paper, we propose a novel 3D Dense Separated Convolution (3D-DSC) module to replace the original 3D convolutional kernels. The 3D-DSC module is constructed by a series of densely connected 1D filters. The decomposition of 3D kernel into 1D filters reduces the risk of overfitting by removing the redundancy of 3D kernels in a topologically constrained manner, while providing the infrastructure for deepening the network. By further introducing nonlinear layers and dense connections between 1D filters, the network’s representational power can be significantly improved while maintaining a compact architecture. We demonstrate the superiority of 3D-DSC on volumetric medical image classification and segmentation, which are two challenging tasks often encountered in biomedical image computing.https://www.mdpi.com/2076-3417/10/2/485convolutional neural networksbiomedical imagingimage segmentationmedical diagnosis |
collection |
DOAJ |
language |
English |
format |
Article |
sources |
DOAJ |
author |
Lei Qu Changfeng Wu Liang Zou |
spellingShingle |
Lei Qu Changfeng Wu Liang Zou 3D Dense Separated Convolution Module for Volumetric Medical Image Analysis Applied Sciences convolutional neural networks biomedical imaging image segmentation medical diagnosis |
author_facet |
Lei Qu Changfeng Wu Liang Zou |
author_sort |
Lei Qu |
title |
3D Dense Separated Convolution Module for Volumetric Medical Image Analysis |
title_short |
3D Dense Separated Convolution Module for Volumetric Medical Image Analysis |
title_full |
3D Dense Separated Convolution Module for Volumetric Medical Image Analysis |
title_fullStr |
3D Dense Separated Convolution Module for Volumetric Medical Image Analysis |
title_full_unstemmed |
3D Dense Separated Convolution Module for Volumetric Medical Image Analysis |
title_sort |
3d dense separated convolution module for volumetric medical image analysis |
publisher |
MDPI AG |
series |
Applied Sciences |
issn |
2076-3417 |
publishDate |
2020-01-01 |
description |
With the thriving of deep learning, 3D convolutional neural networks have become a popular choice in volumetric image analysis due to their impressive 3D context mining ability. However, the 3D convolutional kernels will introduce a significant increase in the amount of trainable parameters. Considering the training data are often limited in biomedical tasks, a trade-off has to be made between model size and its representational power. To address this concern, in this paper, we propose a novel 3D Dense Separated Convolution (3D-DSC) module to replace the original 3D convolutional kernels. The 3D-DSC module is constructed by a series of densely connected 1D filters. The decomposition of 3D kernel into 1D filters reduces the risk of overfitting by removing the redundancy of 3D kernels in a topologically constrained manner, while providing the infrastructure for deepening the network. By further introducing nonlinear layers and dense connections between 1D filters, the network’s representational power can be significantly improved while maintaining a compact architecture. We demonstrate the superiority of 3D-DSC on volumetric medical image classification and segmentation, which are two challenging tasks often encountered in biomedical image computing. |
topic |
convolutional neural networks biomedical imaging image segmentation medical diagnosis |
url |
https://www.mdpi.com/2076-3417/10/2/485 |
work_keys_str_mv |
AT leiqu 3ddenseseparatedconvolutionmoduleforvolumetricmedicalimageanalysis AT changfengwu 3ddenseseparatedconvolutionmoduleforvolumetricmedicalimageanalysis AT liangzou 3ddenseseparatedconvolutionmoduleforvolumetricmedicalimageanalysis |
_version_ |
1725102970301841408 |