Deep learning-based method for the continuous detection of heart rate in signals from a multi-fiber Bragg grating sensor compatible with magnetic resonance imaging

Abstract

A method for the continuous detection of heart rate (HR) in signals acquired from patients using a sensor mat comprising a nine-element array of fiber Bragg gratings during routine magnetic resonance imaging (MRI) procedures is proposed. The method is based on a deep learning neural network model, which learned from signals acquired from 153 MRI patients. In addition, signals from 343 MRI patients were used for result verification. The proposed method provides automatic continuous extraction of HR with the root mean square error of 2.67 bpm, and the limits of agreement were -4.98–5.45 bpm relative to the reference HR.

High-Precision Vital Signs Monitoring Method Using a FMCW Millimeter-Wave Sensor

“…J proposed to convert the constructed heartbeat waveform into an image and used VGG for identification [ 25 ]. For heart rate monitoring, convolutional neural network recognition technology based on the camera-based remote photoplethysmography method [ 26 ] and the fiber-optic sensor [ 27 ] has achieved high-precision detection. However, there are few related studies based on radar sensors, which may be because deep learning currently mainly relies on image training, and it is difficult to convert the electrical signals propagated by radar into images, which is also a challenge for the future.…”

Machine Learning Applications in Optical Fiber Sensing: A Research Agenda

“…There was a significant amount of interest in the application of this type of sensor to measure a variety of different physical and chemical parameters, such as fluid flow estimation [56], mechanical impact [57], refractive index [42,58], liquid identification [59], gas detection [60], and so on. FBS was also widely used in this year [57,61]. However, other techniques such as Fabry-Perot interferometer [60], Brillouin optical time domain analyzer (BOTDA) [62], phase-sensitive optical time domain reflectometry (φ-OTDR) [51], and photonic crystal fibers (PCFs) [63] were also implemented in combination with ML techniques to optimize the measurements.…”