Development of Simplified Diagnosis System for Sleep Apnea Syndrome by Breath Sound
| Project/Area Number |
24500213
|
|---|
| Research Category |
Grant-in-Aid for Scientific Research (C)
|
| Allocation Type | Multi-year Fund |
|---|
| Section | 一般 |
|---|
| Research Field |
Perception information processing/Intelligent robotics
|
|---|
| Research Institution | Kogakuin University |
|---|
Principal Investigator |
|
|---|
| Co-Investigator(Kenkyū-buntansha) |
YAMAGUCHI Yasuhiro 東京大学, 医学部, 講師 (60376473)
|
|---|
| Project Period (FY) |
2012-04-01 – 2016-03-31
|
| Project Status |
Completed (Fiscal Year 2015)
|
| Budget Amount *help |
¥4,810,000 (Direct Cost: ¥3,700,000、Indirect Cost: ¥1,110,000)
Fiscal Year 2015: ¥650,000 (Direct Cost: ¥500,000、Indirect Cost: ¥150,000)
Fiscal Year 2014: ¥910,000 (Direct Cost: ¥700,000、Indirect Cost: ¥210,000)
Fiscal Year 2013: ¥1,170,000 (Direct Cost: ¥900,000、Indirect Cost: ¥270,000)
Fiscal Year 2012: ¥2,080,000 (Direct Cost: ¥1,600,000、Indirect Cost: ¥480,000)
|
|---|
| Keywords | 睡眠時無呼吸症候群 / 無呼吸低呼吸指数 / 呼吸音 / いびき / 音響特徴量 / 機械学習 / なし |
|---|
| Outline of Final Research Achievements |
Sleep Apnea Syndrome (SAS) is a type of disorder that causes behavioral effects such as airow obstruction and/or neural abnormality due to Apnea, and lower oxygen saturation (SpO2) based on abnormal ventilatory function due to Hypopnea. As an aim of this research, a simple and low-cost SAS diagnosis system based on breath sounds was proposed including a prediction model for Apnea Hypopnea Index (AHI) as a part of the system. The feature of this model is to detect a snore sound during sleep and to estimate AHI as a criterion for SAS diagnosis by means of a weighting histogram for the duration of estimated Apnea. As a result of extracting acoustic characteristics for the detected snoring sounds, the correlation between Apnea periods and Hypopnea periods was found on the histogram for the duration of estimated Apnea. AHI prediction model by the histogram was designed. Verifying the efficacy of the prediction model, the average relative error was 37%.
|
Report
(5 results)
Research Products
(8 results)
