Smartphones can be used to detect blood oxygen saturation levels down to 70%, according to new research from the University of Washington and University of California San Diego.
The study shows the potential of using smartphone applications with deep-learning algorithms to measure blood oxygen levels. According to the U.S. Food and Drug Administration, 70% is the lowest value that pulse oximeters should be able to measure.
To measure blood oxygen saturation levels, researchers used an Android smartphone camera with a custom video capture application. The researchers had the study participants wear a standard pulse oximeter on one finger and then place another finger on the same hand over a smartphone's camera and flash. Edward Wang, assistant professor at UC San Diego's Design Lab, said every time a person’s heart beats, fresh blood flows through the part illuminated by the flash.
"The camera records how much that blood absorbs the light from the flash in each of the three-color channels it measures: red, green and blue," said Wang in a prepared statement. "Then we can feed those intensity measurements into our deep-learning model."
When patients had a low oxygen level, researchers say the smartphone app was accurate 80% of the time compared to the standard pulse oximeter device. The team acquired more than 10,000 blood oxygen level readings between 61% and 100% for the study participants.
The team used data from four participants to train the deep-learning algorithm to pull out the blood oxygen levels. It used the remaining data to validate the method and then test it to see how well it performed on new subjects.
Since the start of COVID-19, health systems are increasingly investing in remote patient technology to track patients’ oxygen levels remotely. University Hospitals in Cleveland used this kind of technology to monitor 4,000 patients remotely, said Dr. Peter Pronovost, chief clinical transformation officer in a previous interview.
The results of the study from University of Washington and UC San Diego were published in npj Digital Medicine.