Optical Fibre-Based Pulse Oximetry Sensor with Contact Force Detection

Abstract

A novel optical sensor probe combining monitoring of blood oxygen saturation (S_pO₂) with contact pressure is presented. This is beneficial as contact pressure is known to affect S_pO₂ measurement. The sensor consists of three plastic optical fibres (POF) used to deliver and collect light for pulse oximetry, and a fibre Bragg grating (FBG) sensor to measure contact pressure. All optical fibres are housed in a biocompatible epoxy patch which serves two purposes: (i) to reduce motion artefacts in the photoplethysmogram (PPG), and (ii) to transduce transverse loading into an axial strain in the FBG. Test results show that using a combination of pressure measuring FBG with a reference FBG, reliable results are possible with low hysteresis which are relatively immune to the effects of temperature. The sensor is used to measure the S_pO₂ of ten volunteers under different contact pressures with perfusion and skewness indices applied to assess the quality of the PPG. The study revealed that the contact force ranging from 5 to 15 kPa provides errors of <2%. The combined probe has the potential to improve the reliability of reflectance oximeters. In particular, in wearable technology, the probe should find use in optimising the fitting of garments incorporating this technology.

Keywords: contact pressure; fibre Bragg grating; oxygen saturation; photoplethysmography; plastic optical fibre; pulse oximetry; signal quality index.

Figure 1

(a) Plan view of the probe. Red and orange frames represent transmit POFs connected to two LEDs (660 nm and 850 nm). The yellow frame represents the receiver POF connected to the photo-diode. (b) Side view of the probe. The probe consists of two epoxy rectangular patches which are connected using epoxy resin. The top patch is the pulse oximeter whilst the bottom is the FBG pressure sensor. Two FBGs are fabricated in the core of the silica fibre which are connected to the FBG interrogator (see Section 2.2). The brown line represents the skin surface where the POFs are located.

Figure 2

Opto-electronics system for the designed sensor. Light sources are two different wavelength Thorlabs fibre-coupled LEDs (660 nm and 850 nm). The photodetector is a Thorlabs PDA36AEC. The data-acquisition system is a National Instruments myDAQ at 16 bits and 20 kHz sampling frequency. The FBG interrogator is a Smart Fibres SmartScan FBG interrogator with 0.8 pm resolution. (I/V—current to voltage converter, DAC—digital-to-analogue converter, ADC—analogue-to-digital converter and PC—personal computer).

Figure 3

Set up for loading pressure.

Figure 4

The patch beneath the index finger is the designed epoxy probe. The commercial sensor is taped on the middle finger. Two thermocouples are used to measure the environment and the index finger temperatures.

Figure 5

Comparison of optical fibre sensor with commercial pulse oximeter (a) During normal breathing both devices record stable values with the absolute error 0.443 ± 0.466%. (b) S_pO₂ test result with a desaturation event. Phase 1: Breathe normally (80 s). Phase 2: Breathe out (14 s). Phase 3: Hold breath (19 s). Phase 4: Breathe normally (66 s). Both devices identify the desaturation event with the absolute error 1.16 ± 0.423% across the trace.

Figure 6

Step pressure increasing/decreasing experiment. (a) Increase the pressure step by step until 25 kPa. Then, the pressure decreased step by step to 0 kPa. This process was repeated three times. (b) Resulting FBG peak wavelength shift versus loaded pressures from Figure 6a demonstrating a linear relationship, repeatable results and low hysteresis.

Figure 7

(a) Infrared PPG signals (red) under different contact pressures (blue). (b) Absolute S_PO₂ error (red) compared to an unloaded commercial device (c) skewness index of infrared PPG (red) (d) Perfusion index of infrared PPG (red).

Figure 8

Effect of contact pressure on S_pO₂ error between optical fibre sensor and a commercial pulse oximeter. When the contact pressure is higher than 25 kPa, the S_pO₂ error is >10%. When the contact pressure is lower than 15 kPa, the S_pO₂ error is lower than 2%. The S_pO₂ error reached the lowest value while the contact pressure is in the range from 5 to 15 kPa.