AI-Powered Video System Shows Promise for Non-Invasive Blood Pressure and Diabetes Screening

AI-based video system may offer quick, no-contact screenings for high blood pressure and diabetes

American Heart AssociationNov 11 2024 A new system that combines high-speed video and an artificial intelligence (AI)-powered algorithm may offer quick, no-contact screenings for high blood pressure and Type 1 or Type 2 diabetes without needing blood tests, blood pressure cuffs or expensive wearable devices, according to a preliminary study to be presented at the American Heart Association's Scientific Sessions 2024. The meeting, Nov. 16-18, 2024, in Chicago, is a premier global exchange of the latest scientific advancements, research and evidence-based clinical practice updates in cardiovascular science. This method may someday allow people to monitor their own health at home and could lead to early detection and treatment of high blood pressure and diabetes in people who avoid medical exams and blood tests." Ryoko Uchida, B.Sc. (Pharm.), study author, project researcher in the department of advanced cardiology, University of Tokyo, Japan Blood pressure and diabetes subtly alter blood flow in the face and hands. The researchers tested the effectiveness of a high-speed video camera in capturing face and palm recordings at a rate of 150 images per second. Using wavelength data to detect pulse waves, the research team used an AI algorithm to detect high blood pressure and diabetes from blood flow features in the skin captured in video images. The analysis found: Compared with using the blood pressure values measured by the continuous blood pressure monitor at the same time during video recording, the video imaging/algorithm combo was 94% accurate in detecting stage 1 hypertension according to the American Heart Association's guidelines for blood pressure 130/80 mm Hg or higher. Compared with using the measurements from a continuous blood pressure monitor, a 30-second video imaging/algorithm combo in a subset was 86% accurate in detecting if blood pressure was above normal, while a 5-second video/algorithm combo was 81% accurate. (Normal home blood pressure was based on the Japanese Society of Hypertension guidelines (2018): systolic blood pressure below 115 mm Hg and diastolic blood pressure below 75 mm Hg.) Compared with using hemoglobin A1c blood test results to screen for diabetes, the video/algorithm combo was 75% accurate in identifying people with diabetes. The A1c test measures the average blood sugar level over the past 1-2 months. "I was surprised about the applicability of the blood flow algorithm to detect diabetes. However, some of the major complications of diabetes are peripheral neuropathy -; weakness, pain and numbness, usually in the hands and feet -; and other diseases related to blood vessel damage. It makes sense that changes in blood flow would be a hallmark of diabetes," Uchida said. Several steps must be taken before video images/algorithm combinations are ready for use outside of a research setting. "To detect high blood pressure, we need to incorporate an algorithm that considers arrhythmias or irregular heartbeats. In the future, the prototype camera we used to develop the algorithm could be substituted with an affordable sensor that uses only the essential wavelengths and requires just a few seconds to gather data. Once it reaches that stage, it may be added to smartphones (or even hung on a mirror where someone sits for a few moments), may be mass-produced and inexpensive," Uchida said. She noted that once the accuracy of diabetes detection is improved, they hope to seek approval from the U.S. Food and Drug Administration for an at-home device to detect diabetes. "Currently, the only way to confirm the diagnosis of diabetes is invasive blood tests, however, if it were to require only a non-invasive photo or video, that could be a game-changer," Uchida said. "It is really exciting to see more research that identifies ways to diagnose high blood pressure and diabetes non-invasively, two major risk factors for cardiovascular disease," said Eugene Yang, M.D., M.S., clinical professor of medicine in the division of cardiology and the Carl and Renée Behnke Endowed Chair for Asian Health at the University of Washington School of Medicine in Seattle. "While the results are promising, it is important to recognize that validation of these technologies is lacking. The referenced blood pressure monitor device used in this study, while cleared by the FDA, has not gone through appropriate validation protocols to ensure accuracy. Until we have approved validation protocols for these technologies, including wearable devices like smartwatches, we must use validated devices for measuring blood pressure and glucose levels," said Yang, who is also co-director of the University's Cardiovascular Wellness and Prevention Program. The study has several limitations. It is a preliminary study in early development. The results from this study of predominantly Japanese and Asian adults may not be generalizable to people in other population groups. The camera and algorithm used inside a hospital in this study may yield different results in darker or lighter settings. While participants were instructed not to move their hands or faces during data collection, results may differ if there was movement. Study details, background and design: Participants included 215 adults (average age of 64 years; 36% female; self-identified as mostly Japanese and "other" Asians). Of the total participants, 62 people had been diagnosed with high blood pressure (blood pressure of 130/80 mm Hg or higher); 88 had normal readings according to Japanese standards (less than 115/75 mm Hg); and 65 had readings that fell between these two ranges. Additionally, 44 people had either been diagnosed with diabetes or had an HbA1c level of 6.5% or higher. The data was collected between August 2022 and May 2024 in the cardiology wards and outpatient booths at the University of Tokyo Hospital. Participants sat still while the video was taken of 22 regions of the face and 8 sections of the palm of the hands. Diabetes was defined as a glycosylated hemoglobin (Hba1c) level of 6.5% or higher, the standard cut-off for a measurement that reflects a person's average blood sugar control over the past few months. Source: American Heart Association

AI-powered tool may offer quick, no-contact blood pressure and diabetes screening

A new system that combines high-speed video and an artificial intelligence (AI)-powered algorithm may offer quick, no-contact screenings for high blood pressure and type 1 or type 2 diabetes without needing blood tests, blood pressure cuffs or expensive wearable devices, according to a preliminary study to be presented at the American Heart Association's Scientific Sessions 2024, held Nov. 16-18, 2024, in Chicago. "This method may someday allow people to monitor their own health at home and could lead to early detection and treatment of high blood pressure and diabetes in people who avoid medical exams and blood tests," said study author Ryoko Uchida, B.Sc. (Pharm.), a project researcher in the department of advanced cardiology at the University of Tokyo in Japan. Blood pressure and diabetes subtly alter blood flow in the face and hands. The researchers tested the effectiveness of a high-speed video camera in capturing face and palm recordings at a rate of 150 images per second. Using wavelength data to detect pulse waves, the research team used an AI algorithm to detect high blood pressure and diabetes from blood flow features in the skin captured in video images. The analysis found: "I was surprised about the applicability of the blood flow algorithm to detect diabetes. However, some of the major complications of diabetes are peripheral neuropathy -- weakness, pain and numbness, usually in the hands and feet -- and other diseases related to blood vessel damage. It makes sense that changes in blood flow would be a hallmark of diabetes," Uchida said. Several steps must be taken before video images/algorithm combinations are ready for use outside of a research setting. "To detect high blood pressure, we need to incorporate an algorithm that considers arrhythmias or irregular heartbeats," Uchida said. "In the future, the prototype camera we used to develop the algorithm could be substituted with an affordable sensor that uses only the essential wavelengths and requires just a few seconds to gather data. Once it reaches that stage, it may be added to smartphones (or even hung on a mirror where someone sits for a few moments), and may be mass-produced and inexpensive." She noted that once the accuracy of diabetes detection is improved, they hope to seek approval from the U.S. Food and Drug Administration for an at-home device to detect diabetes. "Currently, the only way to confirm the diagnosis of diabetes is invasive blood tests. However, if it were to require only a non-invasive photo or video, that could be a game-changer," Uchida said. "It is really exciting to see more research that identifies ways to diagnose high blood pressure and diabetes non-invasively, two major risk factors for cardiovascular disease," said Eugene Yang, M.D., M.S., clinical professor of medicine in the division of cardiology and the Carl and Renée Behnke Endowed Chair for Asian Health at the University of Washington School of Medicine in Seattle. "While the results are promising, it is important to recognize that validation of these technologies is lacking. The referenced blood pressure monitor device used in this study, while cleared by the FDA, has not gone through appropriate validation protocols to ensure accuracy. Until we have approved validation protocols for these technologies, including wearable devices like smartwatches, we must use validated devices for measuring blood pressure and glucose levels," said Yang, who is also co-director of the University's Cardiovascular Wellness and Prevention Program. The study has several limitations. It is a preliminary study in early development. The results from this study of predominantly Japanese and Asian adults may not be generalizable to people in other population groups. The camera and algorithm used inside a hospital in this study may yield different results in darker or lighter settings. While participants were instructed not to move their hands or faces during data collection, results may differ if there was movement. Study details, background and design: