Milou Penders, Fontys University of Applied Sciences
Proving the feasibility of a new chest strap measurement principle to get a robust and reliable read out of breathing rate and depth, making breathing measurements more accessible to a larger audience to accelerate advancements in breathing research.
Prototyping a new concept quickly by sampling the raw signal from an optical sensor on a chest strap with the NI myDAQ device and implementing a simple, efficient detection method that converts the periodic breathing signal using the standard fast Fourier transform (FFT).
Milou Penders - Fontys University of Applied Sciences
Sofie Mc Luskey - Fontys University of Applied Sciences
Niels van den Broek - Fontys University of Applied Sciences
Bram Slaats - Fontys University of Applied Sciences
Sjoerd Peters - Fontys University of Applied Sciences
Saskia Blom - Fontys University of Applied Sciences
Geert Langereis - Fontys University of Applied Sciences
At Fontys University of Applied Sciences, located in Eindhoven, Netherlands, we stress the importance of putting theory into practice so students can successfully start their professional careers. By linking our education and research to innovation processes of companies and institutes in our region that develop high-tech equipment for consumer electronics, healthcare, and our connected society, we have become an innovation engine. Our education and research impacts almost all sectors of society. Fontys trains designers in engineering and science by getting them to do scientific validation and design, from concept to functional products.
The Fontys University of Applied Sciences engineering and applied physics departments chose a combination of LabVIEW software with myDAQ hardware as the standard educational platform for measurement automation. Using the myDAQ system with our laptops has changed the way we teach by making education more independent of time and location. Now students can achieve more with every project by using myDAQ and LabVIEW for rapid prototyping and signal evaluation before implementing a more final configuration with microcontrollers or analogue electronics.
Measuring breathing is useful for sports, training, healthcare, and meditation. Developers cannot access current methods of verifying lung functionality and monitoring the health of athletes as open source systems. To accelerate research in breathing measurements, two departments at Fontys have come together to develop a new method to measure breathing that is applicable for all developers. Electronic solutions placed close to the human chest appear to be the most robust and inexpensive solution.
In previous projects at our university, we applied optical ear clip sensors for heart rate monitoring through photoplethysmography. This clip-based method is a good vehicle for students to explore signal processing and sensor interfacing. The clip we used (HRM2511-B, displayed in Figure 1 and made by KYTO Electronic Co) detects the variation of light transmission through an earlobe.
In our new breathing measurement method, we use the same clip to measure transmission of light through a piece of textile as an indication of the amount of chest stretching. We mount the clip on a textile belt in a way that ensures that we can easily separate it from the electronic clip, which combines comfort and hygiene. Using this method, we can measure both breathing rate (time domain) and breathing depth (amplitude domain), a combination that is required to estimate pulmonary output.
The optical ear clip circuitry consists of a power supply, two resistors, and two optoelectronic components (see Figure 2). The myDAQ analogue input directly samples the measured voltage at the phototransistor and uses it as an output signal.
Figure 3 shows how the ear clip connects to a chest belt. A small part of the belt (10 cm) is highly stretchable and the longer part of the belt is made of a stiffer textile and adjusts to the chest size. We kept the stretchable piece short to enhance optical modulation on the sensor. Figure 4 shows how a test subject wears the belt.
After we recorded and digitized the first signal, we needed to start optimization of the detection feature. To deal with noise, motion artefacts, and offset fluctuation of the signal, we considered multiple methods. We decided on a detection technique that assumes a periodic signal due to the periodic nature of breathing. We chose a spectral method because of the insensitivity to motion and noise. We could derive a complete spectrum at once for fingerprinting all periodic signals in the measured data using LabVIEW.
Figure 5 shows a screenshot of the LabVIEW program during a test measurement. We performed prototyping mainly using LabVIEW quick start functions, known as Express VIs. Figure 6 shows our LabVIEW block diagram. If we want to optimize the algorithm further, we can use low-level acquisition and signal conditioning VIs when we know more about the signals.
The three signals represented on the front panel for inspection illustrate our methodology. First, we filtered the signal using a bandpass filter to remove high-frequency noise and very low-frequency sensor drift. Next, we did a fast Fourier transform (FFT) to derive the spectrum. We used the dominant periodic peak in the spectrum as the indicator for the breathing signal. As a reference and for feedback, we reconstructed one single periodic signal. We then logged the acquired breathing frequency over a long period for later analyses.
We prefer spending our time on innovation and the essence of the problem, rather than setting up systems. With entry-level LabVIEW functions, we can quickly set up all needed processing to favour innovation. Designing and verifying a read-out method for optical breath rate detection was an example where the creative process was not limited by technological shortcomings thanks to LabVIEW.
Fontys University of Applied Sciences