Smartphone camera and flash could help measure blood oxygen levels


A standard modern smartphone is capable of detecting blood oxygen saturation levels of up to 70%, according to a study by researchers from the University of Washington and the University of California.

The research team demonstrated a technique that involves participants placing their finger on the camera and flashing a phone, after which a deep learning algorithm deciphers blood oxygen levels from blood flow patterns captured in the resulting video.

When we inhale, our lungs fill with oxygen, which is distributed to our red blood cells to be transported throughout our body. Our bodies need a lot of oxygen to function and healthy people always have an oxygen saturation of at least 95%.

Conditions like asthma or Covid-19 make it harder for bodies to absorb oxygen from the lungs. This leads to oxygen saturation percentages that drop to 90% or less – an indication that medical attention may be needed.

In a clinic, doctors monitor oxygen saturation using pulse oximeters: clips placed on a patient’s fingertip or ear. Being able to monitor oxygen saturation at home several times a day could, for example, help patients keep tabs on Covid symptoms.

In their proof-of-principle study, researchers from the University of Washington and the University of California, San Diego showed that smartphones are able to detect blood oxygen saturation levels of up to 70%. This is the lowest value that pulse oximeters should be able to measure, as recommended by the US Food and Drug Administration.

The team recruited six participants between the ages of 20 and 34. Three identified as female; three identified as male. One participant identified as African American, while the others identified as Caucasian.

To collect data to train and test the algorithm, the researchers asked each participant to wear a standard pulse oximeter on one finger, then place another finger on the same hand on the camera and flash it. a smartphone. Each participant had this same configuration on both hands simultaneously.

Using the smartphone camera technique described above, the team administered a controlled mixture of nitrogen and oxygen to six subjects to artificially lower their blood oxygen levels. The smartphone was able to correctly predict if the subject had low blood oxygen 80% of the time.

Image credit: Dennis Wise/University of Washington

“Other smartphone apps that do this have been developed by asking people to hold their breath. But people get very uncomfortable and have to breathe after about a minute and that’s before their oxygen levels in the blood has declined sufficiently to represent the full range of clinically relevant data,” said co-lead author Jason Hoffman, a UW doctoral candidate at the Paul G. Allen School of Computer Science & Engineering.

“With our test, we are able to collect 15 minutes of data on each subject. Our data shows that smartphones could perform well in the critical threshold range.”

Another advantage of measuring blood oxygen levels on a smartphone is that almost everyone already has such a device.

“That way, you could have multiple measurements with your own device for free or at low cost,” said co-author Dr. Matthew Thompson, professor of family medicine at UW School of Medicine. “In an ideal world, this information could flow seamlessly to a doctor’s office. It would be really beneficial for telemedicine appointments or for triage nurses to be able to quickly determine if patients need to go to the ER. or if they can continue to rest at home and make an appointment with their primary care provider later.”

Edward Wang, lead author of the research paper, and now an assistant professor at UC San Diego’s Design Lab and Department of Electrical and Computer Engineering, said, “The camera records video: Every time your heart beats, fresh blood flows through the part illuminated by the flash.

“The camera records the amount of light from the flash absorbed by the blood in each of the three color channels it measures: red, green and blue. We can then incorporate these intensity measurements into our training model by depth.”

Each test participant breathed a controlled mixture of oxygen and nitrogen to slowly lower their oxygen levels. The process took about 15 minutes. For the six participants, the team acquired more than 10,000 blood oxygen level readings between 61% and 100%.

The researchers used data from four of the participants to train a deep learning algorithm to extract blood oxygen levels. The rest of the data was used to validate the method and then tested it to see how it performed on new subjects.

“Light from the smartphone can be scattered by all these other components of your finger, which means there’s a lot of noise in the data we’re looking at,” said co-lead author Varun Viswanath, an alumnus of the ‘UW and now Ph.D. by Wang at UC San Diego. “Deep learning is a very useful technique here because it can see these really complex and nuanced features and helps you find patterns that you might not otherwise see.”

The team hopes to continue this research by testing the algorithm on more people.

“One of our subjects had thick calluses on his fingers, which made it more difficult for our algorithm to accurately determine his blood oxygen levels,” Hoffman said. “If we were to extend this study to more subjects, we would likely see more people with calluses and more people with different skin tones. Then we could potentially have an algorithm that is complex enough to be able to better model all of these differences.”

The researchers said the findings so far represent a good first step towards the development of machine learning-assisted biomedical devices.

“It’s so important to do a study like this,” Wang said. “Traditional medical devices go through rigorous testing, but computer science research is just beginning to tackle the use of machine learning for biomedical device development and we are all still learning. By forcing ourselves to be rigorous, we force ourselves to learn how to do things well.”

The research paper – “Smartphone camera oximetry in an dependent hypoxemia study” – was published in the journal npj Digital Medicine.

Sign up for the E&T News email to get great stories like this delivered to your inbox every day.


Comments are closed.