A Dual-Tube Actuated POF Sensor with Double End-Face Misalignment for Mattress-Embedded Respiratory Monitoring and Abnormal Pattern Recognition

Abstract This paper presents a dual-tube actuated plastic optical fiber (POF) end-face coupling sensor integrated into a dual-channel mattress for non-invasive monitoring of thoracic and abdominal respiratory motions. The sensing unit employs a three-segment fiber architecture within a nested tube assembly: the middle POF segment is embedded in a flexible inner tube, while the input and output fibers are inserted into the ends of this inner tube with well-defined gaps from the middle fiber’s end faces. This inner assembly is further sheathed in a rigid outer tube. When subjected to cyclic compression from human respiration, the rigid outer tube bears the pressure and drives the middle segment to undergo overall vertical displacement, while the input/output fibers remain stationary. This induces simultaneous lateral misalignment at both end-face pairs (input–middle and middle–output), which magnifies the change in optical coupling efficiency and results in enhanced output intensity variations with signal stability. The transmission loss of the sensor exhibited good linearity with applied force over the range of 0–8 N, yielding R2 = 0.9824 and a sensitivity of 1.362 dB/N. In dynamic tests with simulated respiratory signals, real-time extraction of respiratory rate and amplitude using a peak detection algorithm gave maximum relative errors of 2.6% and 5.2%, respectively. After integration into the mattress, the system was evaluated in a human-subject experiment during respiratory measurements. The thoracic and abdominal channels showed high measurement accuracy, close agreement with a reference respiratory belt, and robust signal quality. Furthermore, several simulated abnormal respiratory states were recorded, and the acquired waveforms captured key features of these abnormal respiratory patterns. These results demonstrate that the proposed mattress can provide respiratory rate, amplitude, and thoracoabdominal phase difference data, offering multidimensional parameters for the assessment of abnormal respiratory patterns.