Abstract
In a present aging and stressful society, it is important to conduct a daily health check for detection of disease in early stages. Therefore, the vital sign measurements in daily life attract much attention. In this study, an electrocardiogram (ECG) acquisition system is built into a chair to measure and record ECG with clothes on. A capacitive connection between body and the system consisting of the clothes and a metal plate on a chair was used as an electrode. However, noise arising from mechanical movement of the back caused by respiration and heart beat superimposes on the ECG. Quite a large baseline wander prevents from a detailed observation of the ECG for 20 seconds or more after seating. It is difficult to remove these noises by a fixed-frequency response filter because of noise having unstable and overlapping frequency components on ECG. This study demonstrates the usefulness of adaptive filter approach for removing a movement-related noise from the signal obtained by capacitive electrodes. A separate noise-related signal required for adaptive noise canceller was measured by other capacitive electrodes simultaneously with ECG. The results indicated that the use of an adaptive noise canceller sufficiently removed the noise components in the capacitive ECG. Baseline wander was suppressed immediately after seating by noise cancellation. It is possible to detect ECG in 0.9 seconds. Although this method sufficiently removed noise, it was not yet examined if the quality of the ECG is enough for diagnosis based on such as ST segment changes because waveform was distorted by the filters in the system. As the next step, quality of the waveform must be improved to be acceptable for diagnosis by optimization of the system parameters including analog circuits. After finishing improvement, evaluation on diagnosis and long term test should be performed to show the real usefulness of capacitive ECG measurement system compensated by adaptive filter.
