Researchers have developed a “digital biomarker” that would use a smartphone’s built-in camera to detect diabetes. The tool could assist in identifying people at higher risk of having diabetes, ultimately helping to decrease the prevalence of undiagnosed diabetes.
Screening tools that can be deployed easily using technology already contained in smartphones could rapidly increase the ability to detect diabetes including populations out of reach of traditional medical care. To date, noninvasive and widely scalable tools to detect diabetes have been lacking, motivating development of the algorithm.
In developing the biomarker, the researchers hypothesized that a smartphone camera could be used to detect vascular damage due to diabetes by measuring signals called photoplethysmography (PPG), which most mobile devices including smartwatches and fitness trackers are capable of acquiring. The researchers used the phone flashlight and camera to measure PPGs by capturing color changes in the fingertip corresponding with each heartbeat.
The team obtained nearly 3 million PPG recordings from 53,870 patients who used the Azumio Instant Heart Rate app on the iPhone and reported having been diagnosed with diabetes by a healthcare provider. This data was used to both develop and validate a deep-learning algorithm to detect the presence of diabetes using smartphone-measured PPG signals.
Overall, the algorithm correctly identified the presence of diabetes in up to 81 percent of patients in two separate datasets. When the algorithm was tested in an additional dataset of patients enrolled from in-person clinics, it correctly identified 82 percent of patients with diabetes. Among the patients that the algorithm predicted did not have diabetes, 92 to 97 percent did not have the disease across the validation data-sets. When this PPG-derived prediction was combined with other easily obtainable patient information — such as age, gender, body mass index, and race/ethnicity — predictive performance improved further.
At this level of predictive performance, the algorithm could serve a similar role to other widespread disease screening tools to reach a much broader group of people, followed by a physician’s confirmation of the diabetes diagnosis and a treatment plan. The algorithm’s performance is comparable to other commonly used tests, such as mammography for breast cancer or cervical cytology for cervical cancer, and its painlessness makes it attractive for repeated testing.
For more information, contact Scott Maier at