Cambp: A camera-based, non-contact blood pressure monitor

Convenient monitoring of vital signs, particularly blood pressure(BP), is critical to improve the effectiveness of healthcare and prevent chronic diseases. This study presents a user-friendly, low-cost, real-time, and non-contact technique for BP measurement based on the detection of photoplethysmography (PPG) using a regular webcam. Leveraging features extracted from photoplethysmograph, an individual's BP can be estimated using a neural network. Experiments were performed on 20 human participants during three different daytime slots given the influence of background illumination. Compared against the systolic blood pressure and diastolic blood pressure readings collected from a commercially available BP monitor, the proposed technique achieves an average error rate of 9.62% (Systolic BP) and 11.63% (Diastolic BP) for the afternoon session, and 8.4% (Systolic BP) and 11.18% (Diastolic BP) for the evening session. The proposed technique can be easily extended to the camera on any mobile device and thus be widely used in a pervasive manner. Copyright held by the owner/author(s).