[Design Tips] Heart rate and blood oxygen level measurement technology for portable and wearable devices

The medical and fitness fields, and the electronic devices associated with these fields, are truly changing at a rapid pace. The demands of today’s healthcare equipment market are not only huge, diverse, but also extremely challenging. Devices that were previously used primarily in hospitals are now also being used for home medical and health monitoring.

For example, it is common to see devices that measure heart rate and blood oxygen levels in consumer products today. Pulse oximeters perform both of these measurements and are currently sold in two forms: as home medical devices and as an integrated component of wrist-worn fitness activity trackers.

This article introduces the basics of pulse oximeters for medical and fitness applications. It also provides a pulse oximeter design example that describes how to measure heart rate and blood oxygen levels.

Oximetry measures the oxygen saturation of the blood, usually expressed as a percentage. A pulse oximeter is a non-invasive device used to measure a person's blood oxygen saturation and heart rate. Pulse oximeters are easily identified by the device's clip-like probe, which usually clips onto the patient's finger.

Pulse oximeters can be stand-alone devices or part of a patient monitoring system, or they can be integrated into a wearable fitness tracker. Accordingly, pulse oximeters can be used by nurses in hospitals, outpatients returning home, fitness enthusiasts in fitness centers, or even pilots operating in low-pressure environments.

Blood oxygen saturation is measured by measuring hemoglobin, the pigment in red blood cells that carries oxygen, which is why they appear red. Hemoglobin transports oxygen to various tissues in the body and exists in two forms. The first is called oxyhemoglobin, expressed as HbO2 (i.e. with oxygen). The other is called deoxyhemoglobin, expressed as Hb (i.e. without oxygen).

Therefore, blood oxygen saturation (SpO2) is the ratio of oxygenated hemoglobin to deoxygenated hemoglobin. It can also be expressed as:

The value of blood oxygen saturation is expressed as a percentage. Normal readings are usually 97% or higher.

An interesting aspect of hemoglobin is the way it reflects and absorbs light. For example, Hb absorbs more (and reflects less) visible red light. HbO2, on the other hand, absorbs more (and reflects less) infrared light. Since blood oxygen saturation can be determined by comparing the Hb and HbO2 values, one way to measure this is to pass red and infrared LED light through a part of the body (such as a finger or wrist) and then compare the relative intensities of the two lights. There are two common methods for doing this: (1) measuring the intensity of light transmitted through tissue is called transmission oximetry, and (2) measuring the intensity of light reflected by tissue is called reflectance oximetry (see Figure 1).

Figure 1: Two methods of oximetry

Hospitals generally use transmissive pulse oximetry. Typically, most hospital patient monitoring systems have integrated transmissive pulse oximeters. However, many of the newer, high-end wearable fitness devices use reflective pulse oximetry.

When the heart beats, it pumps blood throughout the body. Each time the heart contracts, it squeezes blood into the capillaries, causing them to increase slightly in volume. When the heart relaxes, the capillaries decrease in volume. This change in volume affects the amount of light, such as red and infrared light, that is transmitted through tissue. Although this fluctuation is small, it can be measured by a pulse oximeter, using the same device used to measure blood oxygen saturation.

A typical pulse oximeter monitors the oxygen saturation (SpO2) of human blood based on the absorption characteristics of oxygenated hemoglobin (HbO2) and deoxygenated hemoglobin (Hb) to red light (using a wavelength of 600-750 nm) and infrared light (using a wavelength of 850-1000 nm). This type of pulse oximeter alternately transmits red and infrared light through a body part (such as a finger) to a photodiode sensor.

A photodiode is typically used to receive the unabsorbed light from each LED. This signal is then inverted by an inverting operational amplifier (or op amp). The resulting signal represents the light absorbed by the finger, as shown in Figure 2.

Figure 2: Real-time red and infrared (IR) light pulsation signals captured by an oscilloscope

The pulse amplitude (Vpp) of the red and infrared light signals was measured and converted to Vrms to calculate the ratio using the following formula:

SpO2 can be determined from this ratio and an empirically set lookup table. Heart rate can be calculated from the number of samples and the sampling rate of the pulse oximeter's analog-to-digital converter (ADC).

The lookup table is an important part of the pulse oximeter. The lookup table corresponds to the specific oximeter design and is usually based on a calibration curve drawn from a large number of subjects with different SpO2 levels. Figure 3 shows an example of a calibration curve.

Figure 3: Example calibration curve

The following example will detail the different parts of a transmissive pulse oximeter design. As shown in Figure 4, the design demonstrates the measurement of heart rate and blood oxygen saturation levels.

Figure 4: Transmission pulse oximeter system block diagram

The SpO2 sensor used in this example is an off-the-shelf finger clip that integrates a red LED, an infrared LED, and a photodiode. These LEDs are controlled by an LED driver circuit.

The signal conditioning circuit detects the red and infrared light passing through the finger and feeds it into a 12-bit ADC module integrated in the digital signal controller (DSC) to calculate the SpO2 percentage.

The two PWM signals of the DSC drive a dual-channel single-pole double-throw analog switch to alternately switch the red LED and the infrared LED. In order to obtain the appropriate number of ADC samples and have enough time to process the data before the next LED turns on, we control the on and off of the LED according to the timing diagram shown in Figure 5.

Figure 5: Timing diagram

The LED current/intensity is controlled via a 12-bit digital-to-analog converter (DAC) driven by the DSC.

The signal conditioning circuit consists of two stages. The first stage is a transimpedance amplifier and the second stage is a gain amplifier. A high-pass filter is placed between these two stages.

The transimpedance amplifier converts the microampere current produced by the photodiode into a voltage of a few millivolts. The signal received by the first-stage amplifier then passes through a high-pass filter to reduce background light interference.

The signal output from the high-pass filter is then sent to the second-stage amplifier with a gain of 22 and a DC bias of 220 mV. The gain and DC bias of this amplifier should be set to appropriate values ​​so that the output signal level of the gain amplifier is within the ADC range of the MCU.

The output of the analog signal conditioning circuit is connected to the integrated 12-bit ADC module of the DSC. For this example, we used a dsPIC® DSC from Microchip Technology. Using the dsPIC33FJ128GP802 in this design allows us to take advantage of its integrated DSP capabilities while also having easy access to Microchip's digital filter design tools.

An ADC sample is taken during each LED on-time and during both LED off-times. Since it is difficult to directly measure the amount of light passing through body tissue, a filter design tool is used to implement a 513th order FIR bandpass digital filter so that the ADC data can be filtered. The pulse amplitude is then calculated using the filtered data, as shown in Figure 6.

The specifications of the FIR bandpass filter are as follows:

Sampling frequency (Hz): 500 Bandpass ripple (-dB): 0.1

Passband Frequency (Hz): 1 & 5 Stopband Ripple (-dB): 50

Stopband frequency (Hz): 0.05 & 25 Filter length: 513

FIR Window: Kaiser

Figure 6: Input and filtered data

The red curve 1 is the input signal of the FIR filter.

The green curve in Figure 2 is the output signal of the FIR filter.

The X-axis represents the number of ADC samples

The Y-axis represents the ADC code value

The home medical and fitness markets are growing rapidly. The demand for devices that can measure heart rate and blood oxygen levels is set to rise in the coming years. Pulse oximeter reference designs, such as the one described in this article, can be a great help to designers of medical and fitness equipment, helping them get a head start in turning their designs into products and bringing them to market.