IEICE Technical Committee Submission System
Conference Paper's Information
Online Proceedings
[Sign in]
Tech. Rep. Archives
 Go Top Page Go Previous   [Japanese] / [English] 

Paper Abstract and Keywords
Presentation 2019-05-24 15:15
On the influence of data augmentation and network structures in bleeding detection from brain CT images using deep learning
Zhongyang Lu, Masahiro Oda, Tong Zheng, Chen Shen, Tao Hu (Graduate School of Informatics, Nagoya Univ), Takeyuki Watadani, Osamu Abe (Department of Radiology,The University of Tokyo Hospital), Masahiro Hashimoto, Masahiro Jinzaki (Department of Radiology,Keio University School of Medicine), Kensaku Mori (Graduate School of Informatics, Nagoya Univ) SIP2019-15 IE2019-15 MI2019-15
Abstract (in Japanese) (See Japanese page) 
(in English) Based on deep learning technique, the performance of image classification has made great progress. However, their state-of-the-art results are based on enormous data. In this paper, we discuss the effectiveness of data augmentation. Several networks, such as VGG-16, GoogLeNet, Resnet-50, and DenseNet-121 are applied to examine the effectiveness of data augmentation and image cropping on the classification. In this study, 33 cases, a total of 1,220 2D axial CT slices with the condition of Subarachnoid Hemorrhage (SAH) are used in our experiment. This paper utilizes the standard performance measures covering recall, precision, F-measure, and ROC curve for evaluating the trained models. For acquiring more reliable results, 5-fold cross-validation method is employed. Trained on the original dataset, the models got 74.99% F1 score and 0.8220 AUC score, respectively. With the circumstance of augmented dataset, the models got 76.23% F1 score and 0.8396 AUC score.
Keyword (in Japanese) (See Japanese page) 
(in English) Deep learning / brain CT / Subarachnoid Hemorrhage / data augmentation / / / /  
Reference Info. IEICE Tech. Rep., vol. 119, no. 51, MI2019-15, pp. 65-70, May 2019.
Paper # MI2019-15 
Date of Issue 2019-05-16 (SIP, IE, MI) 
ISSN Print edition: ISSN 0913-5685  Online edition: ISSN 2432-6380
Copyright
and
reproduction
All rights are reserved and no part of this publication may be reproduced or transmitted in any form or by any means, electronic or mechanical, including photocopy, recording, or any information storage and retrieval system, without permission in writing from the publisher. Notwithstanding, instructors are permitted to photocopy isolated articles for noncommercial classroom use without fee. (License No.: 10GA0019/12GB0052/13GB0056/17GB0034/18GB0034)
Download PDF SIP2019-15 IE2019-15 MI2019-15

Conference Information
Committee SIP MI IE  
Conference Date 2019-05-23 - 2019-05-24 
Place (in Japanese) (See Japanese page) 
Place (in English)  
Topics (in Japanese) (See Japanese page) 
Topics (in English)  
Paper Information
Registration To MI 
Conference Code 2019-05-SIP-MI-IE 
Language English (Japanese title is available) 
Title (in Japanese) (See Japanese page) 
Sub Title (in Japanese) (See Japanese page) 
Title (in English) On the influence of data augmentation and network structures in bleeding detection from brain CT images using deep learning 
Sub Title (in English)  
Keyword(1) Deep learning  
Keyword(2) brain CT  
Keyword(3) Subarachnoid Hemorrhage  
Keyword(4) data augmentation  
Keyword(5)  
Keyword(6)  
Keyword(7)  
Keyword(8)  
1st Author's Name Zhongyang Lu  
1st Author's Affiliation Graduate School of Informatics, Nagoya University (Graduate School of Informatics, Nagoya Univ)
2nd Author's Name Masahiro Oda  
2nd Author's Affiliation Graduate School of Informatics, Nagoya University (Graduate School of Informatics, Nagoya Univ)
3rd Author's Name Tong Zheng  
3rd Author's Affiliation Graduate School of Informatics, Nagoya University (Graduate School of Informatics, Nagoya Univ)
4th Author's Name Chen Shen  
4th Author's Affiliation Graduate School of Informatics, Nagoya University (Graduate School of Informatics, Nagoya Univ)
5th Author's Name Tao Hu  
5th Author's Affiliation Graduate School of Informatics, Nagoya University (Graduate School of Informatics, Nagoya Univ)
6th Author's Name Takeyuki Watadani  
6th Author's Affiliation Department of Radiology,The University of Tokyo Hospital (Department of Radiology,The University of Tokyo Hospital)
7th Author's Name Osamu Abe  
7th Author's Affiliation Department of Radiology,The University of Tokyo Hospital (Department of Radiology,The University of Tokyo Hospital)
8th Author's Name Masahiro Hashimoto  
8th Author's Affiliation Department of Radiology,Keio University School of Medicine (Department of Radiology,Keio University School of Medicine)
9th Author's Name Masahiro Jinzaki  
9th Author's Affiliation Department of Radiology,Keio University School of Medicine (Department of Radiology,Keio University School of Medicine)
10th Author's Name Kensaku Mori  
10th Author's Affiliation Graduate School of Informatics, Nagoya University (Graduate School of Informatics, Nagoya Univ)
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
Date Time 2019-05-24 15:15:00 
Presentation Time 25 
Registration for MI 
Paper # IEICE-SIP2019-15,IEICE-IE2019-15,IEICE-MI2019-15 
Volume (vol) IEICE-119 
Number (no) no.49(SIP), no.50(IE), no.51(MI) 
Page pp.65-70 
#Pages IEICE-6 
Date of Issue IEICE-SIP-2019-05-16,IEICE-IE-2019-05-16,IEICE-MI-2019-05-16 


[Return to Top Page]

[Return to IEICE Web Page]


The Institute of Electronics, Information and Communication Engineers (IEICE), Japan