Evaluation of a Novel Method for Camera-Based Pulse Oximetry

Abstract

This study aimed to ascertain whether the sensitivity of a CMOS channel to a specific LED can be characterized by a single constant value, addressing a fundamental question in the optimization of CMOS sensors for medical applications. Utilizing a rigorous experimental setup, the research employed varying brightness levels of red and infrared LEDs to assess the response consistency of a CMOS sensor. The methodology cantered around controlled laboratory experiments where LED brightness was systematically varied, and the sensor's response was captured and analysed using custom-developed Python scripts in Google Colab (Mehrabi et al., 2018 & Konstantinidis et al., 2012). The key findings revealed a high degree of linearity in the sensor's response across a range of LED brightness levels, indicating that a single constant value could effectively approximate the sensitivity for each type of LED. Specifically, the study demonstrated that the CMOS sensor could reliably detect and measure light intensities with a high coefficient of determination (R²), supporting the potential for simplifying calibration processes in medical devices that rely on optical sensing (Migaud et al., 2007 & Liu et al., 2015). The conclusion drawn from this research underscores the potential of CMOS sensors to enhance the accuracy and reliability of medical imaging and diagnostic tools. By confirming the linear and predictable nature of sensor responses to controlled light exposure, this study contributes valuable insights into the design and development of future non-invasive medical devices, potentially leading to broader applications and improved patient outcomes (Shin, Lee and Choi, 2011 & Raghda Al-Halawani et al., 2023).