An analog VLSI front end for pulse oximetry

Abstract

Pulse oximetry is a fast, noninvasive, easy-to-use, and continuous method for monitoring the oxygen saturation of a patient's blood. In modem medical practice, blood oxygen level is considered one of the important vital signs of the body. The pulse oximeter system consists of an optoelectronic sensor that is normally placed on the subject's finger and a signal processing unit that computes the oxygen saturation. It uses red and infrared LEDs to illuminate the subject's finger. We present an advanced logarithmic photoreceptor which takes advantage of techniques such as distributed (cascaded) amplification, automatic loop gain control, and parasitic capacitance unilateralization to improve the performance and ameliorate certain shortcomings of existing logarithmic photoreceptors. These improvements allow us to reduce LED power significantly because of a more sensitive photoreceptor. Furthermore, the exploitation of the logarithmic nonlinearity inherent in transistors eliminates the need of performing some of the mathematical operations which are traditionally done in digital domain to calculate oxygen saturation and allows for a very area-efficient all-analog implementation. The need for an ADC and a DSP is thus completely eliminated.

(cont.) We show that our analog pulse oximeter constructed with red and infrared LEDs and our novel photoreceptor at its front end consumes 4.8mW of power whereas a custom-designed ASIC digital implementation (employing a conventional linear photoreceptor) and the best commercial pulse oximeter are estimated to dissipate 15.7mW and 55mW, respectively. The direct result of such power efficiency is that while the batteries in this commercial oximeter need replacement every 5 days (assuming four "AAA" 1.5V batteries are used), our analog pulse oximeter allows 2 months of operation. Therefore, our oximeter is well suited for portable medical applications such as continuous home-care monitoring for elderly or chronic patients, emergency patient transport, remote soldier monitoring, and wireless medical sensing.