| 研究課題/領域番号 |
19K14589
|
|---|
| 研究種目 |
若手研究
|
| 配分区分 | 基金 |
|---|
| 審査区分 |
小区分12040:応用数学および統計数学関連
|
|---|
| 研究機関 | 東京理科大学 |
|---|
研究代表者 |
スヴィリドヴァ ニーナ 東京理科大学, 工学部情報工学科, 助教 (70782829)
|
|---|
| 研究期間 (年度) |
2019-04-01 – 2024-03-31
|
| 研究課題ステータス |
交付 (2022年度)
|
| 配分額 *注記 |
4,030千円 (直接経費: 3,100千円、間接経費: 930千円)
2022年度: 780千円 (直接経費: 600千円、間接経費: 180千円)
2021年度: 780千円 (直接経費: 600千円、間接経費: 180千円)
2020年度: 780千円 (直接経費: 600千円、間接経費: 180千円)
2019年度: 1,690千円 (直接経費: 1,300千円、間接経費: 390千円)
|
|---|
| キーワード | photoplethysmogram / time series analysis / nonlinear dynamics / chaos / chaos theory / recurrence plot / Photoplethysmogram / deterministinc chaos / filtering / image processing / biomedical signal |
|---|
| 研究開始時の研究の概要 |
Photoplethysmogram (PPG) is widely used for health monitoring. Recent studies demonstrated that imaging PPG can be obtained from the conventional camera's video. It has great potential for health monitoring applications, but noise is a significant problem. Common filtration methods decrease PPG dynamics complexity. PPG dynamical characteristics can be used for health monitoring, so it is essential to preserve signal dynamics. This study aims to develop dynamics preserving method for coping with noise issue for data extraction from video recordings obtained from a conventional camera.
|
| 研究実績の概要 |
In FY 2022, the dynamics of the photoplethysmogram signals were furtherly investigated. Thoughtful investigation of the data collection parameters effects, such as time series length and sampling, were studied. The results demonstrated the dependence of the dynamical characteristics, extracted by nonlinear time series analysis, from the time series length and sampling rate. Moreover, it was found that the obtained results can be used to detect the minimal time series length required for reliable feature extraction. Furthermore, these results were validated on complex nonlinear models; thus, these results' applicability can be extended to other nonlinear complex signals besides the photoplethysmogram.
Additionally, in this FY imaging photoplethysmogram (iPPG) signals were investigated. First, iPPG signals were successfully extracted from video recordings obtained from an iPad camera. Then its dynamical characteristics were investigated. The results demonstrated significant similarities between reference signals recorded by a conventional pulse oximeter and signals extracted from an iPad camera. This unveils that a smartphone/iPad camera can be used to record photoplethysmographic signals with dynamical characteristics similar to those of conventional pulse oximeters, thus enabling possibilities for photoplethysmogram-based health monitoring applications, such as physiological and mental health monitoring, without any specialized devices.
|
| 現在までの達成度 (区分) |
現在までの達成度 (区分)
2: おおむね順調に進展している
理由
The iPPG signals that preserve important chaotic properties, such as positive Lyapunov exponent, were successfully extracted from videos recorded by a conventional iPad camera. Extracted signal properties were carefully investigated and found to be comparable to those of the conventional photoplethysmograms recorded by designated pulse oximeters. Moreover, a fundamental question related to the minimal length of photoplethysmographic signal that can provide reliable analysis results was thoughtfully investigated. Results revealed minimal requirements for the signal length that provide reliable information on the dynamical properties of the signal. These results provide essential insight for the further development of wearable photoplethysmographic devices for personal health monitoring applications.
|
| 今後の研究の推進方策 |
Further research will be conducted to deepen the understanding of iPPGs. First, it is necessary to confirm whether the same quality signals and results on the dynamical characteristics can be obtained from other conventional cameras besides the iPad, to ensure results applicability despite the camera choice. Then the results obtained for conventional PPG signals, such as timeseries length effect on the dynamical features extraction, need to be validated for iPPG signals, as the time series length and especially sampling rate can be limited in the case of data recordings by conventional cameras. These results are expected to be important for establishing affordable daily health monitoring.
|
