Scheme of heart rate and blood oxygen detection by intelligent wearable device

Nowadays, there are more and more forms of intelligent wear, and the additional functions are more and more abundant. For example, heart rate and blood oxygen almost become the standard functions of intelligent wear. In the 11th article, taking smart bracelets and watches as examples, it discusses how to detect heart rate and blood oxygen in intelligent wear.

There are two methods for routine blood oxygen test

First, the traditional blood oxygen saturation test method is advanced human blood sampling, using the blood gas analyzer for electrochemical analysis, measuring the partial pressure of oxygen PO2, calculating the concentration of bleeding oxygen, this method is troublesome and can not be checked continuously.

Secondly, a finger sleeve photoelectric sensor is used to cover the human finger. The finger is used as a transparent container containing hemoglobin. Red light with wavelength of 660nm and near-infrared light of 940nm are used to enter the light source. The light sensing intensity of tissue bed is measured, and the hemoglobin concentration and oxygen saturation are calculated.

Scheme of heart rate and blood oxygen detection by intelligent wearable device

Definition of blood oxygen saturation:

Human red blood cells attach hemoglobin, that is Hb, easily combine with oxygen to form oxygenated hemoglobin, that is HBO2, and hemoglobin reduction. When the artery returns to the vein through the capillary, oxygen molecules fall off, which is red blood cells. Red blood cells are attached with hemoglobin, which contains four oxygen molecules, which are saturated hemoglobin molecules. Blood oxygen saturation, a typical value for healthy people, the normal range of this value is 90% - 100%.

As shown in the figure below, when hemoglobin has no oxygen molecule, the blue line Hb in the figure has strong red line absorption; when hemoglobin with weak infrared absorption has oxygen molecule, the red line HBO2 in the figure has weak red line absorption and strong infrared absorption.

Usually, two wavelengths are selected for blood oxygen analyzer. According to the absorption curve, our light source is close to the absorption peak, so 660nm for red light and 940nm for infrared light. It can be seen from the figure that the two curves meet at 800 nm, and the absorption of the two curves is the same, so this frequency band cannot be selected.

High end blood oxygen test, even up to 8 wavelengths, the main reason is that human hemoglobin in addition to reduced hemoglobin and oxygenated hemoglobin, there are other hemoglobin, we often see is carboxyhemoglobin, more wavelengths are conducive to your better accuracy.

How does smart wear measure heart rate?

We know that human blood is red, mainly because of the strong absorption of green. Because of the periodic pacing of blood vessel, the distance from blood vessel to photoelectric sensor changes periodically. As long as the sensor continuously emits green light, the receiver will receive periodic absorption peak, and the heart rate curve can be obtained by processing the signal.

Our consumer intelligent wearable devices, such as smart watches, are very limited in size. Because the traditional penetration photoelectric testing method is not practical, we generally use reflective photoelectric testing scheme. As shown in the figure below, on the one hand, the light source emits led, on the other hand, it is the CMOS receiver. Through the change of reflected light, the inspection results can be confirmed in real time. In order to ensure the stability and continuity of the data, the normal transmission frequency is 3000 ~ 4000Hz, and the occupancy rate of LED is about 50%.

After understanding the detection principle of blood oxygen and heart rate, we found that the biggest problem when the actual product landed was that the wrist and watch could not be in close contact. For example, when walking, if the arm shakes, the relative position of the sensor and the arm may be different, and the test result is unstable, and the difference between the result and the calibration value is very large.

In order to overcome this problem, we want to know whether it is possible to use multiple sets of sensors and place them in each direction to reduce the possibility of the problem. For example, apple and Mio started using a set of transmit and receive, and then used two transmitters and two receivers, as shown in the figure below

The two groups do have an advantage over one, but the problem cannot be solved. Some people think that I can use multiple sets of emitters with receiver in the middle (the receiver integrates visible light reception and infrared light reception). For example, the image below shows the surrounding emitter. From the actual inspection results, it is indeed improved. However, synchronization brings new problems, such as large power consumption and serious heating, which will also reduce the data stability of the receiver. (in the figure, the square is the receiving end, and the circle is the transmitting end)

The working current of each LED is 10 ~ 30mA. If it is monitored in real time, the power consumption of multiple lamps working together is not low. The battery capacity of smart watch is only about 500mA, which will affect the continuous capacity. The designer is very depressed. Why can't we put the transmitter in the middle? There are receivers around. I didn't expect that. The key point is that the standby current of the receiver is only μ a, so the power consumption can be ignored, and the serious heating problem is solved. Therefore, it has become the mainstream design trend