In vivo investigations of photoplethysmograms and arterial oxygen saturation from the auditory canal in conditions of compromised peripheral perfusion

Pulse oximeters rely on the technique of photoplethysmography (PPG) to estimate arterial oxygen saturation (SpO2). In conditions of poor peripheral perfusion such as hypotension, hypothermia, and vasoconstriction, the PPG signals detected are often small and noisy, or in some cases unobtainable. Hence, pulse oximeters produce erroneous SpO2 readings in these circumstances. The problem arises as most commercial pulse oximeter probes are designed to be attached to peripheral sites such as the finger or toes, which are easily affected by vasoconstriction. In order to overcome this problem, the ear canal was investigated as an alternative site for measuring reliable SpO2 on the hypothesis that blood flow to this central site is preferentially preserved. Novel miniature ear canal PPG sensors were developed along with a state of the art PPG processing unit and a data acquisition system to allow for PPG measurements from different depths and surfaces of the ear canal. A preliminary in vivo investigation on seven healthy volunteers has revealed that good quality PPG signals with high amplitude can be obtained from the posterior surface of the outer ear canal. Based on these observations, a second prototype probe suitable for acquisition of PPGs from the posterior surface of the outer ear canal was developed. A pilot study was then carried out on 15 healthy volunteers to validate the feasibility of measuring PPGs and SpO2 from the ear canal in conditions of induced local peripheral vasoconstriction (right hand immersion in ice water). The PPG signals acquired from the ear canal probe were compared with those obtained simultaneously from finger probes attached to the left and the right index fingers. Significant drop (p 45%) and right (> 50%) index fingers during the ice water immersion, while good quality PPG signals with relatively constant amplitude were obtained from the ear canal. Also, the SpO2 values showed that the ear canal pulse oximeter performed better than the two finger pulse oximeters (mean failure rate 30%). A second in vivo investigation was carried out in 15 healthy volunteers, where hypoperfusion was induced more naturally by exposing the volunteer to cold temperatures of 10C for 10min. Normalised Pulse Amplitude (NPA) and SpO2 was calculated from the PPG signals acquired from the ear canal, the finger and the earlobe. By the end of the cold exposure, a mean drop of > 80% was found in the NPA of finger PPGs. The % drop in the NPA of red and infrared earlobe PPG signals was 20% and 26% respectively. Contrarily to both these sites, the NPA of the ear canal PPGs had only dropped by 0.2% and 13% respectively. The SpO2 estimated from the finger sensor was below 90% in 5 volunteers (failure) by the end of the cold exposure. The earlobe pulse oximeter failed in 3 volunteers. The ear canal sensor on the other hand had only failed in 1 volunteer. These results strongly suggest that the ear canal may be used as a suitable alternative site for reliable monitoring of PPGs and SpO2 in cases of compromised peripheral perfusion