Contactless Vital Signs Measurement and Analysis Systems

本文提出的项目旨在开发一种非侵入性、非接触式系统，用于测量、记录、分析和解释心肺声音以及由心肺引起的机械性身体运动。为了实现这一目标，我们将使用敏感的压力和声音传感器来测量心脏诱发的身体运动，也称为心跳心动图(BCG)、心音记录仪(心音)、呼吸诱导的身体运动和呼吸声音。传感器将被放置在受试者的床垫下，并在受试者躺在床上的整个过程中不断监测他/她。除了心脏和肺部产生的声音和运动外，传感器还会记录其他身体运动(如打呼噜、打滚、腿部运动等)产生的不必要的声音和运动。我们将使用先进的信号处理和基于机器学习的过滤/分类方案来消除不需要的信号的影响，并分离出重叠的心音和肺音、心脏运动、呼吸运动和其他声音。机器学习算法还将从记录中提取与健康相关的信息，如心率、呼吸频率和肺活量。除了拟议的技术能够改善甚至延长老年人的寿命外，我们的非接触式监测系统在睡眠障碍诊断和治疗监测、呼吸事件和心脏活动监测方面也有几个潜在的应用。

The project proposed herein aims to develop a non-invasive, contactless system for measuring, recording, analysing and interpreting heart and lung sounds and mechanical body movement induced by the heart and lungs. To achieve this aim we will utilize sensitive pressure and sound sensors to measure heart induced body movement, also known as ballistocardiograph (BCG), phonocaridograph (heart sounds), respiratory induced body movements and respiratory sounds. The sensors will be placed under the person's mattress and will constantly monitor him/her throughout the time the subject is resting on the bed. Besides heart and lung generated sounds and movements the sensors will record unwanted sounds and movements resulting from other body movements (such as snoring, rolling, leg movements, etc.). We will use advanced signal processing and machine learning-based filtering/classification schemes to remove the effects of the unwanted signals, and to separate out the overlapping heart and lung sounds, heart induced movement, respiration movement and other sounds. The machine learning algorithms will also extract health related information such as heart rate, breathing rate, and lung capacity from the recordings. Beyond the proposed technology's capacity to improve, and perhaps extend, the lives of senior citizens, our contactless monitoring system has several potential applications in sleep disorders diagnosis and treatment monitoring, respiratory events, and heart activity monitoring.