Low-Cost Pulse Oximeter for Improving Accuracy on Different Skin Pigmentation

Na, Si Qi (2026) Low-Cost Pulse Oximeter for Improving Accuracy on Different Skin Pigmentation. Final Year Project (Diploma), Tunku Abdul Rahman University of Management and Technology.

Text Na Si Qi_Thesis.pdf Restricted to Registered users only Download (5MB)

Abstract

Pulse oximeters are essential tools for monitoring oxygen levels (SpO₂) and heart rate, but their accuracy can be affected by skin pigmentation. Many commercial devices tend to overestimate or underestimate SpO₂ in people with darker skin tones, which may lead to incorrect readings and risks in medical decision-making. This project focuses on developing a low-cost transmittance pulse oximeter (TPO) that reduces this bias and provides more reliable results for users with different skin tones. The prototype is built using an Arduino Uno, a BPW34 photodiode with 660nm and 940nm LEDs and an OLED display to show SpO₂ in real time. A software-based calibration method is introduced, where users can manually select their skin tone (light, medium, or dark) equations using a simple button interface. The device then applies the corresponding correction equation to improve measurement accuracy. To evaluate the prototype's performance, an experiment was conducted involving three healthy participants with three different skin colors. The participants' skin tones were categorized using the Monk Skin Tone (MST) scale. When compared to a commercial TPO oximeter reading, the skin-specific equations successfully corrected the overestimation occur in darker skin tones, though the device exhibited a higher signal variance due to the lack of a stabilized finger enclosure. With its affordability and algorithmic calibration, this project delivers a practical solution for equitable community healthcare monitoring