Paper Abstract and Keywords |
Presentation |
2019-07-06 11:55
[Poster Presentation]
Basic study of left atrial appendage segmentation from cardiac CT images Itaru Takayashiki, Doi Akio, Toru Kato, Hiroki Takahashi (Iwate Prefectural Univ.), Shoto Sekimura (ISP), Maiko Hozawa, Yoshihiro Morino (Iwate Medical Univ.) MI2019-30 |
Abstract |
(in Japanese) |
(See Japanese page) |
(in English) |
In this study, we propose a method to automatically extract the left atrial appendage region from the cardiac CT image for the purpose of facilitating the diagnosis of the left atrial appendage closure procedure. Generally, in the case of a cardiac CT image, it is difficult to automatically classify the left atrial appendage region because the heart region is a very complicated organ. Therefore, in addition to the semantic segmentation method using Fully Convolutional Neural Networks (FCN), we performed an automatic extraction of only the left atrial appendage region using mini-batch and Adversarial training, for heart contrast CT containing contrast medium with constant condition for CT measurement. With this method, it becomes possible to automatically obtain information necessary for preoperative plan support of left atrial appendage closure from cardiac CT images. |
Keyword |
(in Japanese) |
(See Japanese page) |
(in English) |
Left Atrial Appendage Region / Heart CT image / Deep Learning / Fully Convolutional Neural Networks / Segmentation / Left Atrial Appendage Occlusion / / |
Reference Info. |
IEICE Tech. Rep., vol. 119, no. 104, MI2019-30, pp. 43-48, July 2019. |
Paper # |
MI2019-30 |
Date of Issue |
2019-06-28 (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 |
MI2019-30 |
|