Abstract:Unobtrusive sleep position classification is essential for sleep monitoring and closed-loop intervention systems that initiate position changes. In this paper, we present a novel unobtrusive under-mattress optical tactile sensor for sleep position classification. The sensor uses a camera to track particles embedded in a soft silicone layer, inferring the deformation of the silicone and therefore providing information about the pressure and shear distributions applied to its surface.We characterized the sensitivity of the sensor after placing it under a conventional mattress and applying different weights (258 g, 500 g, 5000 g) on top of the mattress in various predefined locations. Moreover, we collected multiple recordings from a person lying in supine, lateral left, lateral right, and prone positions. As a proof-of-concept, we trained a neural network based on convolutional layers and residual blocks that classified the lying positions based on the images from the tactile sensor.We observed a high sensitivity of the optical tactile sensor: Even after placing the sensor below a conventional mattress, we were able to detect our lowest test weight of 258 g. Using the neural network, we were able to classify the four sleep positions, lateral left, lateral right, prone, and supine with a classification accuracy of 91.2 %.The high sensitivity of the sensor, as well as the good performance in the classification task, demonstrate the feasibility of using such a sensor in a robotic bed setup.Clinical Relevance- Positional Obstructive Sleep Apnea is highly prevalent across the general population. Today's gold standard treatment of using CPAP ventilation is often not accepted, leading to unwanted treatment cessations. Alternative treatments, such as positional interventions through robotic beds are highly promising. However, these beds require reliable detection of the lying position. In this paper, we present a novel, scalable and completely unobtrusive sensor that is concealed under the mattress while classifying sleeping positions with high accuracy.

A Method for Sleep Position Identification Based on Back Propagation Neural Network

A Method to Recognize Sleeping Position Using an CNN Model Based on Human Body Pressure Image

A Study on Sleep Position Recognition of Body Pressure Image Based on KPCA and SVM

Deep Residual Networks for Sleep Posture Recognition with Unobtrusive Miniature Scale Smart Mat System

Human Sleep Posture Recognition Method Based on Interactive Learning of Ultra-long Short-term Information

[A Mattress System of Recognizing Sleep Postures Based on BCG Signal].

A Real-Time Patient-Specific Sleeping Posture Recognition System Using Pressure Sensitive Conductive Sheet and Transfer Learning

In-bed posture classification using pressure sensor data and spiking neural network

A dual fusion recognition model for sleep posture based on air mattress pressure detection

Unobtrusive Smart Mat System for Sleep Posture Recognition.

A Novel Body Posture Recognition System on Bed

Sleep and Wake Classification Based on Heart Rate and Respiration Rate

Unobtrusive Sleep Position Classification Using a Novel Optical Tactile Sensor

An Intelligent Head Posture Recognition Pillow

In-Bed Posture Classification Using Deep Neural Network

Design and Implementation of a Noncontact Sleep Monitoring System Using Infrared Cameras and Motion Sensor.

Identity and Posture Recognition in Smart Beds with Deep Multitask Learning

Lightweight Neural Network for Sleep Posture Classification Using Pressure Sensing Mat at Various Sensor Densities

A Feasibility Study on Smart Mattresses to Improve Sleep Quality

Sleep Position Detection with a Wireless Audio-Motion Sensor—A Validation Study

Automatic Sleep Stage Classification Using Nasal Pressure Decoding Based on a Multi-Kernel Convolutional BiLSTM Network