High blood pressure and diabetes, the diagnosis will come with artificial intelligence

by time news

Measure blood pressure? Artificial intelligence will take care of it with an ultra-modern diagnosis. The gesture that everyone has seen done at least once seems destined to go into retirement: the cloth sleeve tightened around the patient’s arm, the tube connected ‍to ‍a pump​ that inflates it, the hand that begins to move on a sort of clock, while‍ a person‍ with a stethoscope on ⁤his ears reads the values. In the not too distant future, the ‘ritual’ could⁣ be archived, together with the iconic device, known in technical jargon as a sphygmomanometer.

Artificial intelligence diagnosis: how ​it ‌will work

Because AI will tell us if we have high blood pressure. In fact, a⁤ team of Japanese ​scientists seems to have ​opened this path, developing a new ‘contactless’ diagnosis system that could​ also work for diabetes.

The system – which combines a patented algorithm‌ based on artificial intelligence with a high-speed video (duration 5-30 seconds) of⁢ the ‍skin of the face⁤ and the palm of the hand – in a preliminary study managed to detect whether a subject suffered from pressure high, exactly like a blood pressure monitor.

The system is still in the early stages​ of development in Japan, and – as the authors explain – it also accurately⁢ detected type 1 or type 2 ‍diabetes. With the modifications necessary for practical use, it could in the future offer rapid and contactless for hypertension and for the ‘sweet blood‍ disease’,⁢ without the need for blood tests, blood pressure monitors ⁤or expensive⁤ wearable devices, and help monitor‌ response to treatment. The study is among the abstracts presented in the American Heart Association’s 2024 Scientific Sessions (November 16-18, Chicago) and obviously has not yet been peer-reviewed.

“This method‌ could one day allow people to monitor their health at‍ home and⁣ could ⁢lead to early diagnosis and treatment” of the two conditions⁢ “in people who shy away from doctor visits and⁢ blood tests,” says study author Ryoko Uchida, a project researcher‌ in the Department of ⁤Advanced Cardiology at ​the University of Tokyo. How does the system work? Blood pressure and diabetes subtly alter blood flow in⁤ the face and hands. The researchers tested the camera’s effectiveness 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 hypertension and diabetes from blood ⁤flow characteristics in the ⁤skin ‍captured in video images.

Precise measurement

What the analysis allowed us to detect is for example that, compared to using blood pressure values ​​simultaneously measured by the continuous blood pressure monitor, ⁣the combination⁢ of video images/algorithm had an accuracy ‌of 94% in detecting⁤ the stage 1 hypertension according ⁢to the American​ Heart‍ Association guidelines‍ (equal to or greater than 130/80‌ mmHg). And again, it was found​ that, compared to the use of the glycated hemoglobin test (which measures the average ⁢blood sugar‌ level over the last‌ 1-2 ​months) for diabetes screening, the video/algorithm combination​ was 75% accurate % in identifying people with diabetes.

“I was surprised by the‌ applicability of the blood flow algorithm to detect diabetes. Although in reality some of the main⁣ complications ​of diabetes are peripheral neuropathy (weakness, pain and numbness, usually‍ in the hands and feet) and other ‌related diseases to blood vessel damage. It therefore makes sense that changes in blood flow are a hallmark ‍of the disease,” reflects Uchida.

The next steps

Several steps are now ⁢required before the system is ready for use ⁤outside a research context, the experts point out. “To detect hypertension, we need to incorporate an algorithm that considers arrhythmias or irregular heartbeats – Uchida proposes – In the future, the⁤ prototype camera we used to develop the algorithm could be replaced with a sensor that uses only essential wavelengths‌ and takes only a few seconds to collect data. Once it reaches this⁢ stage, it could⁢ be added to smartphones ⁤(or even hung on a mirror to sit in⁢ front of for a few moments), ⁢and it ⁢could be mass-produced and cheap.”

⁣ A ​similar system takes on even more value if you consider that today “the ‌only way ⁣to confirm the⁣ diagnosis of diabetes⁣ is through invasive blood testsso if only a⁢ non-invasive photo or video were needed, it could⁢ be a game-changer,” Uchida points out.⁢ Once the accuracy of diabetes detection has improved, the team’s hope is to obtain approval from the US FDA for a home diabetes device.

Interview Between Time.news Editor and AI Health Diagnostics ‌Expert

Time.news Editor: Welcome, Dr. Ryoko Uchida! We’re excited to have you here ‍to discuss your groundbreaking work in AI and health diagnostics. Your recent study ⁣seems to suggest a ‍future where traditional methods of⁣ diagnosis, like blood pressure measuring, may become obsolete. Can you walk us through how your AI system works?

Dr. Uchida: Thank you for having me!⁣ Absolutely. Our ‍system uses a combination of high-speed video recording and a patented AI algorithm to analyze subtle⁤ changes in ​blood flow that occur due to conditions like hypertension and ‍diabetes.‌ By capturing video of the skin on‌ the⁣ face and palms at an impressive rate of 150 ‌images per second, we can infer health metrics without any direct​ contact or invasive procedures.

Time.news‌ Editor: Fascinating! So instead of the traditional sphygmomanometer and stethoscope, we could just use a camera? How does the algorithm interpret the video data to make such accurate​ health assessments?

Dr. Uchida: That’s⁣ correct! The AI analyzes specific characteristics of blood flow captured in the video footage. It detects pulse ⁢waves and measures variations in blood circulation that are indicative of conditions like high blood pressure. In our preliminary studies, the AI demonstrated ‌a 94% ⁢accuracy rate in​ detecting stage 1 hypertension when compared to⁢ continuous blood pressure monitors.

Time.news Editor: That’s impressive! It sounds like this technology could also‌ ease the burden on those who typically avoid regular health check-ups. How do you envision this technology being utilized in⁢ everyday life?

Dr. Uchida: Exactly! Our goal is⁣ to empower individuals to monitor their ​health at home. This could be particularly beneficial for people who are hesitant to visit healthcare facilities or undergo traditional testing procedures. By providing contactless and rapid diagnostics, we hope to enable early detection and treatment of hypertension and diabetes, ultimately improving⁢ health outcomes.

Time.news Editor: What about the accuracy of this system? You ⁤mentioned that the approach was 75% accurate for diagnosing diabetes as well. How‌ do you ensure its reliability as it moves toward practical application?

Dr. ⁢Uchida: Reliability is critical, of ⁢course. While our current accuracy is promising, we’re committed to refining the system through more extensive trials and modifications. The technology‌ is still in its early stages, and further peer-reviewed studies will help ​confirm‌ and improve its effectiveness before widespread adoption.

Time.news Editor: And speaking of trials, your study has been presented at the American Heart Association’s 2024​ Scientific ‌Sessions. What kind of feedback have you received from the scientific community?

Dr. Uchida: The response has been enthusiastic!⁣ Many⁢ researchers and healthcare professionals recognize the potential of AI ‌to transform diagnostics. There’s a keen interest in exploring how this technology can integrate with existing healthcare systems and improve access​ to medical care.

Time.news Editor: As we look to the future, what excites you most about the role of artificial intelligence⁣ in ​medicine?

Dr. Uchida: I am particularly excited about the potential for AI to personalize‌ healthcare. The​ ability to continuously monitor individuals’ health ‌data​ remotely can help physicians tailor treatments to each ⁢patient’s unique needs. This could ‌lead to more effective management of chronic diseases and ultimately reduce​ the overall healthcare burden.

Time.news Editor: ‌ It sounds like we’re on the brink of quite a revolution in health diagnostics! Thank you for sharing your insights with us, Dr. Uchida. We look forward to seeing how this technology ⁤develops and changes the landscape of healthcare.

Dr. Uchida: Thank you for having ⁢me! It’s an exciting ⁤journey, ‌and I ‍appreciate the opportunity to discuss our⁢ work with your audience.

You may also like

Leave a Comment