The pulse oximeter is a non-immersive instrument used to monitor the patient’s blood oxygen saturation. It is benefiting from an expensive discrete component solution to a more integrated design. However, integration means making difficult choices about which processing architecture to use.
Pulse oximeters rely on pulse intensity or pulsatile flow for measurement, so the area to be monitored must have good blood flow, and any blockage may cause measurement errors.
Pulse oximeters can also calculate the pulse frequency, so most pulse oximeters have this function. The measuring range of a typical pulse oximeter is between 70% and 100% saturation, and readings below 70% are unreliable.
The working principle of pulse oximeter is based on the change of light absorption during arterial pulse. Two light sources located in the visible red light spectrum (660 nanometers) and the infrared spectrum (940 nanometers) alternately illuminate the area under test (generally fingertips or earlobe). The amount of light absorbed during these pulsations is related to the amount of oxygen in the blood. The microprocessor calculates the ratio of the two spectra absorbed, and compares the result with the saturation value table stored in the memory to obtain the hemorrhagic oxygen saturation.
Typical components of a pulse oximeter include: a microprocessor, memory (EPROM and RAM), two digital-to-analog converters that control the LED, devices that filter and amplify the signal received by the photodiode, and digitize the received signal to provide Analog-to-digital converter for the microprocessor. The LED and photodiode are placed in a small probe that comes into contact with the patient’s fingertip or earlobe. The pulse oximeter also includes a display and a plotter, which are generally discrete components.
Several manufacturers already provide integrated analog microcontrollers with A / D and D / A converters, microcontroller cores, memory and other peripheral functions. Only filtering, signal conditioning, and amplification require external components. If the function of the selected MCU is strong enough, the signal adjustment can be completed in the digital domain, thereby simplifying the system further.
However, the choice of integrated solutions is quite challenging, because there are many types of analog microcontrollers, such as 8-bit, 16-bit and 32-bit CISC and RISC architectures, as well as multiple A / D, D / A and Other peripheral options.
For MCUs, designers should first consider an 8-bit core, such as the traditional 8051. These kernels are well-known and easy to understand and use. In addition, a large number of codes and proven tools are available.
Although 16-bit and 32-bit MCU cores are not popular in data acquisition systems, they are gradually being accepted. Using these wide-bus MCUs is easier to process wide-byte data than using 8-bit MCUs, so they can calculate 12-bit, 16-bit, and 24-bit A / D data easier and faster. With their faster operating speeds, these wide-bus MCUs are more efficient than 8-bit MCUs in performing complex algorithms. Tool providers have begun to provide toolkits and support for these 16 / 32-bit integrated solutions.
For a simple pulse oximeter, the 8-bit MCU core is sufficient, but if more functions such as complex filtering, mathematical calculations, or data processing are required, then 16 / 32-bit devices may be a better choice. The ideal solution would be: a 16 / 32-bit MCU with internal program flash and RAM (for storing operating programs, saturation lookup tables, and captured data), two D / A converters for driving light sources, A multi-channel A / D converter with a resolution of at least 12 bits to quantify the data received by the photodiode (it can also be used to monitor other parameters such as battery life). It would be better if there were some configurable glue logic on the board.
