2021 Fiscal Year Research-status Report
Development of dynamics preserving method for extraction of imaging photoplethysmogram signal from a conventional camera's video recordings.
| Project/Area Number |
19K14589
|
|---|
| Research Institution | Tokyo University of Science |
|---|
Principal Investigator |
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)
|
|---|
| Project Period (FY) |
2019-04-01 – 2023-03-31
|
| Keywords | photoplethysmogram / nonlinear dynamics / time series analysis / chaos theory / recurrence plot |
|---|
| Outline of Annual Research Achievements |
This fiscal year, attention was paid to the systematic and thorough investigation of the changes in dynamical properties of complex dynamics due to the data limitations such as the data resolution and length, which are essential factors when dealing with the experimentally measured photoplethysmogram (PPG), and especially in the case of the video recording-based imaging PPG.
First, dynamics changes were investigated in a nonlinear model with properties resembling those of the PPG. Then conventional PPGs were investigated. As a result, the effect of PPG recording length on its dynamical characteristics was clarified, and the minimum length for robust dynamical properties estimation was evaluated. In particular, it was found that such important dynamical properties as determinism and the complexity expressed by the entropy can be correctly estimated even from very short recordings.
Preliminary results also demonstrated that the PPG dynamics can be potentially reconstructed from the highly sparse marked point process data generated from conventional PPG data, which is essential for the investigation of imaging PPG dynamical properties.
|
| Current Status of Research Progress |
Current Status of Research Progress
2: Research has progressed on the whole more than it was originally planned.
Reason
Systematic investigation of the chaotic models and photoplethysmogram dynamical properties under data collection-related limitations was conducted, thus creating a solid background for continuing this research project.
|
| Strategy for Future Research Activity |
Continue investigation of dynamical properties changes between original conventional PPGs and sparse unequal sampling interval data (i.e. marked point process data) generated from conventional PPG recordings. Conduct a comparison between marked point processes data due to its generation method to find which generation method provides the best dynamical properties preservation. Then based on the obtained results conduct evaluation of the difference between dynamical properties observed in conventional PPG based sparse data and those of the video recordings-based data.
The obtained results are expected to create the base for the identification of the best dynamical properties preserving method.
|
| Causes of Carryover |
This fiscal year the significant amount was reserved for international conference participation. However, due to the effect of COVID-19 international conferences were held online, thus creating a significant amount unused. The remained amount will be used for conference participation in the next fiscal year and to cover journal publication expenses, such as the open access fee.
|
