Paper Abstract and Keywords |
Presentation |
2020-03-06 11:35
Performance improvement by bone removal based on watershed algorithm and texture analysis in extravasation detection using contrast CT images Hiroki Kimura, Kumiko Arai, Yuichiro Yoshimura, Takaaki Nakada, Shigeto Oda, Toshiya Nakaguchi (Chiba Univ) IMQ2019-34 IE2019-116 MVE2019-55 |
Abstract |
(in Japanese) |
(See Japanese page) |
(in English) |
We are investigating an automatic detection method for extravasation using contrast-enhanced CT images in order to reduce the burden on doctors in emergency medicine. In this paper, we propose a bone removal method based on the watershed algorithm for bone removal, which is one of the important factors in the detection process, and improve its performance. In addition, the number of false positive was reduced by using a random forest learner to classify candidate area using texture feature. By using the proposed bone removal method, the sensitivity was improved and the detectability was improved compared to the conventional method. In addition, false positives were reduced by about 35% compared to previous studies by classifying candidate regions corresponding to bone misdetection. |
Keyword |
(in Japanese) |
(See Japanese page) |
(in English) |
contrast-enhanced CT images / extravasation / machine learning / Random Forest method / texture analysis / / / |
Reference Info. |
IEICE Tech. Rep., vol. 119, no. 454, IMQ2019-34, pp. 93-96, March 2020. |
Paper # |
IMQ2019-34 |
Date of Issue |
2020-02-27 (IMQ, IE, MVE) |
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 |
IMQ2019-34 IE2019-116 MVE2019-55 |
|