Unconstrained Identification of the Positions of Chest and Abdomen and Detection of Respiratory Motions in Sleep by Using a Bed-Size Tactile Sensor Sheet

Respiratory diseases can be effectively screened by monitoring chest and abdominal motions during respiration. Respiratory inductance plethysmography (RIP) is commonly used for this purpose in clinics. However, RIP has strong constraints on human body, limiting its daily use. Several unconstrained methods have been developed, but they cannot replace RIP due to their inability to accurately distinguish between chest and abdominal motions. The authors’ group proposed an unconstrained method that has the potential to replace RIP by using a bed-size tactile sensor sheet. To increase its accuracy, the present work investigates the personalized identification method of optimal measurement regions of the chest and abdomen based on the body pressure distribution measured by the tactile sensor sheet. A new discriminative feature procedure and a corresponding mathematical model were proposed, and tests were conducted on 16 healthy subjects. The results demonstrate that the proposed method can enhance the measurement accuracy of the sensor sheet method, providing a convenient and practical way to early screening of respiratory diseases.