PPG reliability can be compromised by ambient light and poor sensor-skin contact. This study introduces the Contact Quality Index (CQI), a spectral-based metric to assess PPG signal degradation caused by ambient light and suboptimal sensor placement.
Data were collected from 50 participants wearing a wrist band for 24 hours which provided three PPG channels and one ambient light channel, all sampled at 55 Hz. The ambient signal was segmented into 5-second windows and analyzed using Welch's PSD method with 50 % overlap, 256-point Hamming windows to mitigate spectral leakage, and a 2048-point FFT. The CQI was calculated by integrating the PSD curve using a trapezoidal method with unit spacing. The area was log-transformed, normalized between 0 and 1 and inverted to yield the CQI, with higher scores indicating cleaner signal conditions and better sensor contact.
Proper contact between the device and the skin showed PSD peaks, reaching 80 dB/Hz, indicating minimal interference. Poor contact caused increased peaks power value up to 100 dB/Hz, reflecting the presence of ambient light artifacts. Average CQI values of the segments were high (>0.8) during good contact and lower (<0.4) during poor contact.
The effectiveness of the CQI was validated by comparing it to the SNR of the corresponding PPG segments. Those with minimal interference showed an average SNR of 28.03 dB and an average CQI of 0.8. In contrast, segments affected by ambient light exhibited a lower average SNR of 14.5 dB, with a corresponding average CQI of 0.3.
The CQI provides a computationally efficient and interpretable approach to evaluating PPG signal contamination in the presence of ambient light. The method is lightweight, independent of physiological variations, and suitable for integration into wearable systems. Future work will explore the integration of CQI into adaptive preprocessing pipelines to improve signal robustness under dynamic environmental conditions.