Paper Abstract and Keywords |
Presentation |
2009-05-22 11:10
An Attempt of a Novel Calibration Method for Pulse Oximetry Using Support Vector Machines Non-Linear Regression Hirotaka Nomoto, Mitsuhiro Ogawa (Kanawaza Univ), Yasuhiro Yamakoshi (yu.sys Corp.), Masamichi Nogawa, Takehiro Yamakoshi, Kosuke Motoi, Shinobu Tanaka, Ken-ichi Yamakoshi (Kanawaza Univ) MBE2009-2 |
Abstract |
(in Japanese) |
(See Japanese page) |
(in English) |
A new calibration method using a non linear multivariate regression method, support vector machines regression (SVMsR) of pulse oximetry was attempted. Two photo plethysmograms (PPGs) by red and IR were measured on five subjects. Simultaneously, oxygen saturation (SpO_2) level was measured by a commercial pulse oxymeter. SpO_2 level was controlled by breathing 10% oxygen gas for obtaining 98-75% SpO2 level. Sequentially, feature points of two PPGs were extracted in beat by beat. Then, SVMsR were attempted to obtain regression model between SpO_2 by meter and the feature points of PPG. Finally, the regression model was evaluated by unknown data. The root mean squared error is 1.90 [SpO_2 level %]. In conclusion, SVMsR might be applicable on calibration for pulse oximetry. |
Keyword |
(in Japanese) |
(See Japanese page) |
(in English) |
Pulse oximeter / Photoplethysmogram (PPG) / Multivariate Analysis / Support Vector Machines Regression / Nonlinear Problem / / / |
Reference Info. |
IEICE Tech. Rep., vol. 109, no. 50, MBE2009-2, pp. 5-8, May 2009. |
Paper # |
MBE2009-2 |
Date of Issue |
2009-05-15 (MBE) |
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 |
MBE2009-2 |
|