Validation of a spectroscopic sensor for the continuous, noninvasive measurement of muscle oxygen saturation and pH

• Published in: Physiological measurement 2013-08, Vol.34 (8), p.859-871
• Main Authors: Ellerby, G E C, Smith, C P, Zou, F, Scott, P, Soller, B R
• Format: Article
• Language: English
• Subjects: acidosis; Animals; Female; Humans; Hydrogen-Ion Concentration; In Vitro Techniques; Male; Monitoring, Physiologic - instrumentation; Muscle, Skeletal - metabolism; near infrared spectroscopy; noninvasive sensors; Oxygen - metabolism; patient monitoring; Perfusion; Reproducibility of Results; Spectroscopy, Near-Infrared; Sus scrofa; tissue perfusion
• ISSN: 0967-3334; 1361-6579
• DOI: 10.1088/0967-3334/34/8/859

New patient monitoring technologies can noninvasively and directly provide an assessment of the adequacy of tissue perfusion through the simultaneous determination of muscle oxygen saturation (SmO2) and muscle pH (pHm). Non-pulsatile near infrared spectroscopy is used to determine these microvascular parameters. Two separate studies were conducted using an isolated perfused swine limb preparation to widely vary venous blood oxygen saturation (SviO2) and pH (pHvi) to assess the accuracy of a noninvasive sensor with the capability to simultaneously measure both parameters. The isolated limb model is necessary to establish equilibrium between the venous output of the perfusion circuit and the venule measurement of the spectroscopic sensor. The average absolute difference between SmO2 and SviO2 determined over 50 conditions of SviO2 between 13% and 83% on 3 pig limbs was 3.8% and the coefficient of determination (R2) was 0.95. The average absolute difference between pHm and pHvi determined over 69 conditions of pHvi between pHvi 6.9 and pHvi 7.5 on 3 pig limbs was 0.045 pH units with an R2 of 0.92. Measured accuracy was acceptable to support clinically relevant decision making for the assessment of impaired tissue perfusion and acidosis. Sensors were also evaluated on human subjects. There was no statistical difference in SmO2 by gender or location when multiple sensors were evaluated on the right and left calf, deltoid, and thigh of resting men and women (N = 33). SmO2 precision for subjects at rest was 5.6% over the six locations with four different sensors.