New Pulse Oximeter Technology Adapts to Different Skin Tones

Researchers led by Valencia Koomson have developed ChromaSense, a prototype pulse oximeter designed to eliminate skin-tone bias. Unlike traditional devices, which misdiagnosed occult hypoxemia in up to 17% of Black patients in a 2020 study, ChromaSense adjusts its light emission based on individual skin reflectance. In clinical testing at UC San Francisco, the wrist-worn device achieved oxygen-saturation accuracy within 2.87% of reference standards across diverse skin tones, meeting FDA requirements.

ChromaSense pulse oximeter prototype

• ▪The ChromaSense device dynamically adjusts its light emission levels and signal-processing parameters after first measuring a user's individual skin reflectance profile.
• ▪The ChromaSense device utilizes a watch-sized box worn on the wrist to measure reflected light from skin and tissue rather than transmitting light through a finger.
• ▪A prototype pulse oximeter called ChromaSense, developed by Valencia Koomson's lab, measures blood oxygen saturation, heart rate, and respiration rate across diverse skin tones.

Skin tone measurement bias

• ▪A 2020 study published in the New England Journal of Medicine found that up to 17% of Black patients had undetected occult hypoxemia despite reassuring pulse oximeter readings.
• ▪The 2020 New England Journal of Medicine study showed that Black patients experienced undetected low arterial oxygen at a rate more than three times that of white patients.

Photoplethysmography signal interference

• ▪Photoplethysmography waveforms can be used to determine heart rate from peak spacing, oxygen levels from light absorption, and breathing rate from pulse amplitude changes.
• ▪Photoplethysmography technology measures the rising and falling pulsing pattern of reflected light to represent arterial blood volume changes in microscopic blood vessels.

Melanin light absorption effects

• ▪Melanin pigment in darker skin absorbs and scatters light, which can weaken the optical signal or distort the ratio used by traditional pulse oximeters to calculate oxygen saturation.
• ▪Variations in skin pigmentation, blood flow, and age can alter the faint optical signals that light-based wearable monitors depend on, potentially causing clinical errors.

Clinical accuracy testing results

• ▪In initial testing across 50 participants, the ChromaSense device achieved an oxygen-saturation measurement accuracy within 1.4% of a standard reference oximeter.
• ▪In a hypoxia study at the University of California, San Francisco, ChromaSense demonstrated oxygen-saturation accuracy within 2.87% of a reference oximeter with no tone-dependent bias.

Machine learning blood pressure

• ▪Researchers used machine-learning models to estimate systolic and diastolic blood pressure from photoplethysmography waveforms of 2,315 adult intensive care unit patients.
• ▪The machine-learning models achieved up to 90% accuracy for blood pressure estimation across diverse subgroups, though this technology is not yet integrated into ChromaSense.