Authors:
Songlin Li
1
;
Julien Denoulet
1
;
Olivier Tsiakaka
2
;
3
and
Sylvain Feruglio
1
Affiliations:
1
CNRS UMR 7606, LIP6, Sorbonne Université, Paris, France
;
2
Biomedical Microsystems Laboratory, Université Laval, Quebec, Canada
;
3
CERVO, University of Laval, Quebec, Canada
Keyword(s):
Heterogeneous System, Modeling, Monte-Carlo Simulation, Multi-domain, Near-Infrared Spectroscopy (NIRS), Oxygen Saturation (SpO2), PhotoPlethysmoGraphy (PPG), PVT (Process, Voltage, Temperature), SystemC(-AMS).
Abstract:
Oxygen concentration quantification in the blood (SpO2) has been used as a mean to diagnose and prevent critical medical conditions thanks to pulse oximetry. In spite of its theoretical precision, this method suffers from intrinsic deviations from the components used in such monitoring devices (PVT) that can lead to SpO2 measurement errors. In this paper, we propose a multi-domain modeling of a NIRS-based blood oxygen saturation monitoring system and its biological environment using SystemC(-AMS) for virtual prototyping, to analyze the impact of PVT variations of opto-electrical components, thanks to Monte-Carlo simulation correlated with transient analysis. We simulated the blood flow of the finger tissue and the dynamic attenuation of the red and infrared light passing through the tissue. The Monte-Carlo simulation method was used to analyze different PVT parameters that may cause measurement deviations separately. Finally, we found that the red/IR LED peak wavelength deviation and
the temperature of the system have an important impact on the SpO2 quantification, especially red LED peak wavelength deviation. This result shows that the choice of the red-light source is of prime importance for accurate SpO2 quantification.
(More)