First Evaluation of a Newly Constructed Underwater Pulse Oximeter for Use in Breath-Holding Activities

Studying risk factors in freediving, such as hypoxic blackout, requires development of new methods to enable remote underwater monitoring of physiological variables. We aimed to construct and evaluate a new water- and pressure proof pulse oximeter for use in freediving research. The study consisted of three parts: (I) A submersible pulse oximeter (SUB) was developed on a ruggedized platform for recording of physiological parameters in challenging environments. Two MAX30102 sensors were used to record plethysmograms, and included red and infra-red emitters, diode drivers, photodiode, photodiode amplifier, analog to digital converter, and controller. (II) We equipped 20 volunteers with two transmission pulse oximeters (TPULS) and SUB to the fingers. Arterial oxygen saturation (SpO(2)) and heart rate (HR) were recorded, while breathing room air (21% O(2)) and subsequently a hypoxic gas (10.7% O(2)) at rest in dry conditions. Bland-Altman analysis was used to evaluate bias and precision of SUB relative to SpO(2) values from TPULS. (III) Six freedivers were monitored with one TPULS and SUB placed at the forehead, during a maximal effort immersed static apnea. For dry baseline measurements (n = 20), SpO(2) bias ranged between −0.8 and −0.6%, precision between 1.0 and 1.5%; HR bias ranged between 1.1 and 1.0 bpm, precision between 1.4 and 1.9 bpm. For the hypoxic episode, SpO(2) bias ranged between −2.5 and −3.6%, precision between 3.6 and 3.7%; HR bias ranged between 1.4 and 1.9 bpm, precision between 2.0 and 2.1 bpm. Freedivers (n = 6) performed an apnea of 184 ± 53 s. Desaturation- and resaturation response time of SpO(2) was approximately 15 and 12 s shorter in SUB compared to TPULS, respectively. Lowest SpO(2) values were 76 ± 10% for TPULS and 74 ± 13% for SUB. HR traces for both pulse oximeters showed similar patterns. For static apneas, dropout rate was larger for SUB (18%) than for TPULS (<1%). SUB produced similar SpO(2) and HR values as TPULS, both during normoxic and hypoxic breathing (n = 20), and submersed static apneas (n = 6). SUB responds more quickly to changes in oxygen saturation when sensors were placed at the forehead. Further development of SUB is needed to limit signal loss, and its function should be tested at greater depth and lower saturation.