FIG. 4 illustrates  a pulse  oximeter  of  the  type wherein

15  light of  two different wavelengths  is passed  through    any pulsatile tissue bed, such as a side of a face or the scalp, so as to be modulated by the pulsatile component of arterial blood therein, and thereby allowing an indication of oxygen saturation,  blood  perfusion  and  heart  rate.  The  level  of

20  incident light is continually adjusted for optimal detection of the pulsatile component, while permitting accommodation to variable attenuations due to skin color, flesh thickness and other invariants. At significant slope reversal of the pulsatile component to negative (indicating a wave maximum), wave 25   form analysis of blood flow  occurs.

A quotient of the pulsatile component of light transmis- sion is measured for each of two wavelengths by direct digital tracking. The respective quotients are thereafter con-

30 verted to a ratio, which ratio may be thereafter fitted to a curve of independently derived of oxygen saturation for the

purpose of calibration. The saturation versus ratio calibra- tion curve may be characterized by various mathematical techniques  including  polynomial  expansion  whereby  the

35 coefficients of the polynomial specify the curve. An output of pulse rate, pulsatile flow and oxygen saturation is given.

An incident light source duty cycle is chosen to be at  least 1 in 4 so that noise, inevitably present in the signal, may be substantially eliminated and filtered.

40 In FIG. 4, a part-schematic, part-perspective view of the optical sensor is shown. A flexible base material 150 is provided. Incorporated into base material 150 at suitably spaced  intervals  are the  electrical  components  of sensor.

Photoelectric sensor 130 is attached to the outside of    base 45 150 and protrudes slightly from the underside of base 150. Sensor 130 has ground wire G and lead wire 131. Light emitting  diode  110, typically  emitting  frequencies  in the

infrared range of the spectrum, is mounted to and pierces base 150 in a similar manner to sensor 130 and at a distance 50 from sensor 130 of approximately several centimeters or less. LED 110 is connected to ground wire G and has input lead wire 111. Placed in proximity to LED 110 is a second LED 120, typically having wavelength emission  character-istics in the red range of the spectrum. LED 120 attaches to55 ground wire G and has input lead wire 121. Resistor 140 is shown mounted to base 150 between sensor 130 and   LED110. However, the physical location of resistor 140 is not important and it may be mounted to sensor at any other convenient  location. Resistor  140 has input  lead wire 14160 and is connected to ground wire G. Wires G, IM, 121, 131, 141 lead to connector 152 so that sensor may be readily disconnected from the processor electronics 160. In an alternative embodiment, the LEDs may not be wired together or may be contained in separate base  components.

65 The sensor of FIG. 4 is constructed in  the  following manner: LEDs 110, 120 are selected from batches of LEDs with generally known wavelength characteristics. The exactwavelength characteristics of the specific LEDs M0, 120 chosen are determined at this time through readily available metering means. Resistor 140 or a similar impedance ref- erence is then selected to have an impedance or specificallya resistance  whose  amount  is exactly specified  by  a table  5made available to the factor technician for this purpose, of all possible wavelength combinations which may be expected to be encountered from the available supplies of LEDs. The following table is an example  of how  a   singleresistor 140 might be selected for any hypothetical combi- 10 nation of LED’s 110, 120 in a case where each has only two possible wavelengths: