ITE Technical Group Submission System
Conference Paper's Information
Online Proceedings
[Sign in]
 Go Top Page Go Previous   [Japanese] / [English] 

Paper Abstract and Keywords
Presentation 2022-05-20 16:40
3D Medical Image Segmentation Using 2.5D Deformable Convolutional CNN
Yuya Okumura, Kudo Hiroyuki, Takizawa Hotaka (Tsukuba Univ.)
Abstract (in Japanese) (See Japanese page) 
(in English) An effective method to improve the accuracy of 3D medical image segmentation using deep learning is to use deformable convolutional CNN, which can absorb individual differences in organ structure and misalignment using displacement vector fields. However, the natural extension from 2D to 3D is impractical due to the huge amount of computation required to calculate and store the displacement vector fields. In this study, we propose a 2.5D method to solve this problem, in which a deformable convolutional CNN is used to perform segmentation in 2D cross sections of xy, yz, and xz horizontal sections, and the results are integrated by majority voting to obtain 3D segmentation results. Experimental results on a real CT image dataset of the abdomen show that the proposed method is more accurate than conventional deep learning methods due to the introduction of deformable convolution, and the computational complexity of the proposed method is realistic for a 2.5D method.
Keyword (in Japanese) (See Japanese page) 
(in English) CT images / Deep learning / Convolutional Neural Networks / 3D CT images / Computer-aided Detection Systems / Automatic recognition and detection of anatomical structures / /  
Reference Info. ITE Tech. Rep.
Paper #  
Date of Issue  
Download PDF

Conference Information
Conference Date 2022-05-19 - 2022-05-20 
Place (in Japanese) (See Japanese page) 
Place (in English) Kumamoto University 
Topics (in Japanese) (See Japanese page) 
Topics (in English)  
Paper Information
Registration To IEICE-MI 
Conference Code 2022-05-SIP-BioX-IE-MI-IST-ME 
Language Japanese 
Title (in Japanese) (See Japanese page) 
Sub Title (in Japanese) (See Japanese page) 
Title (in English) 3D Medical Image Segmentation Using 2.5D Deformable Convolutional CNN 
Sub Title (in English)  
Keyword(1) CT images  
Keyword(2) Deep learning  
Keyword(3) Convolutional Neural Networks  
Keyword(4) 3D CT images  
Keyword(5) Computer-aided Detection Systems  
Keyword(6) Automatic recognition and detection of anatomical structures  
1st Author's Name Yuya Okumura  
1st Author's Affiliation University of Tsukuba (Tsukuba Univ.)
2nd Author's Name Kudo Hiroyuki  
2nd Author's Affiliation University of Tsukuba (Tsukuba Univ.)
3rd Author's Name Takizawa Hotaka  
3rd Author's Affiliation University of Tsukuba (Tsukuba Univ.)
4th Author's Name  
4th Author's Affiliation ()
5th Author's Name  
5th Author's Affiliation ()
6th Author's Name  
6th Author's Affiliation ()
7th Author's Name  
7th Author's Affiliation ()
8th Author's Name  
8th Author's Affiliation ()
9th Author's Name  
9th Author's Affiliation ()
10th Author's Name  
10th Author's Affiliation ()
11th Author's Name  
11th Author's Affiliation ()
12th Author's Name  
12th Author's Affiliation ()
13th Author's Name  
13th Author's Affiliation ()
14th Author's Name  
14th Author's Affiliation ()
15th Author's Name  
15th Author's Affiliation ()
16th Author's Name  
16th Author's Affiliation ()
17th Author's Name  
17th Author's Affiliation ()
18th Author's Name  
18th Author's Affiliation ()
19th Author's Name  
19th Author's Affiliation ()
20th Author's Name  
20th Author's Affiliation ()
Speaker Author-1 
Date Time 2022-05-20 16:40:00 
Presentation Time 20 minutes 
Registration for IEICE-MI 
Paper #  
Volume (vol) vol.46 
Number (no)  
Date of Issue  

[Return to Top Page]

[Return to ITE Web Page]

The Institute of Image Information and Television Engineers (ITE), Japan