Adaptive filter theory (2nd ed.)
Adaptive filter theory (2nd ed.)
IEEE Transactions on Information Technology in Biomedicine
| Hi-index | 0.00 |
The implementation of wearable diagnostic devices would enable real-time remote physiological assessment and facilitate triage of injured persons operating in dangerous and high-risk environments. For example, a wearable pulse oximeter to monitor arterial oxygen saturation (SpO2) and heart rate (HR) could enable field medics to monitor vital physiological information following critical injuries, thereby prioritizing live saving medical intervention. However, since photoplethysmographic (PPG) signals, from which SpO2 and HR measurements are derived, are compromised during movements, the applications of commercially available pulse oximeters require the wearer to remain at rest. This study was undertaken to investigate if accelerometry-based adaptive noise cancellation is effective in minimizing SpO2 and HR errors induced during jogging exercises to simulate artifacts expected to occur in the field. Since transmission-type pulse oximeters utilizing fingertip sensors are highly vulnerable to motion artifacts, measurements were performed using a custom, forehead-mounted pulse oximeter. Preliminary tests revealed that processing the motion corrupted PPG signals by a simple least mean squared (LMS) adaptive noise cancellation (ANC) algorithm can improve the signal-to-noise ratio of motion-corrupted PPG signals, thereby reducing SpO2 and HR errors during jogging.