Paper Abstract and Keywords |
Presentation |
2011-09-06 09:25
Study of input voxel size on voxel-based classifier for lesion detection Yukihiro Nomura, Mitsutaka Nemoto, Yoshitaka Masutani, Shouhei Hanaoka, Soichiro Miki, Takeharu Yoshikawa, Naoto Hayashi, Kuni Ohtomo (The Univ. of Tokyo) MI2011-48 |
Abstract |
(in Japanese) |
(See Japanese page) |
(in English) |
To improve detection accuracy for small lesion, it is desirable that a resolution of original volume data should be as high as possible. However, resolution of clinical image is limited due to hardware limitation or size of imaging range. Volume upsampling by interpolation may be used instead of high resolution volume data. In this short report, we investigated relation between voxel size after interpolation and detection accuracy by using cerebral aneurysm detection in MR angiograms based on voxel based classifier. In the experimental study with 83 clinical cases, the results show that volume upsampling by interpolation have potential to improve detection accuracy for small lesion. |
Keyword |
(in Japanese) |
(See Japanese page) |
(in English) |
computer assisted detection (CAD) / voxel-based classifier / voxel size / interpolation / / / / |
Reference Info. |
IEICE Tech. Rep., vol. 111, no. 199, MI2011-48, pp. 7-10, Sept. 2011. |
Paper # |
MI2011-48 |
Date of Issue |
2011-08-30 (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 |
MI2011-48 |
|