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 2020-01-29 13:20
Imbalanced Subarachnoid Hemorrhage data automatic detection by using SMOTE algorithm based on deep learning
Zhongyang Lu, Masahiro Oda, Yuichiro Hayashi, Hayato Ito (Nagoya Univ), Takeyuki Watadani, Osamu Abe (Department of Radiology,The Univ of Tokyo Hospital), Masahiro Hashimoto, Masahiro Jinzaki (Department of Radiology,Keio Univ School of Medicine), Kensaku Mori (Nagoya Univ) MI2019-75
Abstract (in Japanese) (See Japanese page) 
(in English) Based on deep learning techniques, the performance of image classification has made significant progress. Especially in the medical image processing field, the CNNs are broadly utilized. However, for the most problems in the real world, the numbers of every class in the data set are not equal, called imbalanced data. It causes a low recall of the minority class. In this paper, we apply the SMOTE method to alleviate the imbalanced problem on anomaly detection. Sequentially, we utilize this strategy on the Subarachnoid Hemorrhage (SAH) detection with imbalanced data based on deep learning techniques and discuss the efficiency. In this study, 33 cases of SAH data combined with 33 cases, and 100 cases of normal brain CT, respectively, are applied to support our experiments. We utilize F-measure and ROC curve for evaluating the trained models. Trained on the 33 cases SAH and 100 cases normal dataset, the model got 0.731 AUC score without SMOTE processing. With SMOTE, acquired 0.830 AUC score, and during SMOTE and data augmentation, the performance was improved into a 0.875 AUC score.
Keyword (in Japanese) (See Japanese page) 
(in English) Deep learning / Imbalanced classification / Subarachnoid Hemorrhage / data augmentation / / / /  
Reference Info. IEICE Tech. Rep., vol. 119, no. 399, MI2019-75, pp. 47-52, Jan. 2020.
Paper # MI2019-75 
Date of Issue 2020-01-22 (MI) 
ISSN 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 MI2019-75

Conference Information
Committee MI  
Conference Date 2020-01-29 - 2020-01-30 
Place (in Japanese) (See Japanese page) 
Place (in English) OKINAWAKEN SEINENKAIKAN 
Topics (in Japanese) (See Japanese page) 
Topics (in English) Medical Image Engineering, Analysis, Recognition, etc. 
Paper Information
Registration To MI 
Conference Code 2020-01-MI 
Language English (Japanese title is available) 
Title (in Japanese) (See Japanese page) 
Sub Title (in Japanese) (See Japanese page) 
Title (in English) Imbalanced Subarachnoid Hemorrhage data automatic detection by using SMOTE algorithm based on deep learning 
Sub Title (in English)  
Keyword(1) Deep learning  
Keyword(2) Imbalanced classification  
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 Nagoya University (Nagoya Univ)
2nd Author's Name Masahiro Oda  
2nd Author's Affiliation Nagoya University (Nagoya Univ)
3rd Author's Name Yuichiro Hayashi  
3rd Author's Affiliation Nagoya University (Nagoya Univ)
4th Author's Name Hayato Ito  
4th Author's Affiliation Nagoya University (Nagoya Univ)
5th Author's Name Takeyuki Watadani  
5th Author's Affiliation Department of Radiology,The University of Tokyo Hospital (Department of Radiology,The Univ of Tokyo Hospital)
6th Author's Name Osamu Abe  
6th Author's Affiliation Department of Radiology,The University of Tokyo Hospital (Department of Radiology,The Univ of Tokyo Hospital)
7th Author's Name Masahiro Hashimoto  
7th Author's Affiliation Department of Radiology,Keio University School of Medicine (Department of Radiology,Keio Univ School of Medicine)
8th Author's Name Masahiro Jinzaki  
8th Author's Affiliation Department of Radiology,Keio University School of Medicine (Department of Radiology,Keio Univ School of Medicine)
9th Author's Name Kensaku Mori  
9th Author's Affiliation Nagoya University (Nagoya Univ)
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 2020-01-29 13:20:00 
Presentation Time 30 minutes 
Registration for MI 
Paper # MI2019-75 
Volume (vol) vol.119 
Number (no) no.399 
Page pp.47-52 
#Pages
Date of Issue 2020-01-22 (MI) 


[Return to Top Page]

[Return to IEICE Web Page]


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