There has been a steady increase in the use of healthcare apps over the past few years, thanks to new sensors and machine learning algorithms. While these health apps help us keep tabs on how our body (and mind) is doing, how much we walked, how we slept, or how many calories we burned, and whether they could help professionals as well. health care to diagnose illnesses and diseases more accurately and faster than ever?
Researchers at University College London have partnered with the Africa Health Research Institute and developed AI technology that could help healthcare professionals interpret HIV test results and provide a diagnosis faster than never.
The two teams created the AI ââalgorithm following a study in which a team of more than 60 trained field workers from the Africa Health Research Institute built a comprehensive library of images of HIV tests taken in KwaZulu-Natal, South Africa, using a health app and image capture protocol created by University College London.
The library contained over 11,000 images, all of which were fed into the machine learning algorithm, and then the results were compared for accuracy with users who only interpreted them with the naked eye.
The AI ââreduced errors and correctly classified the images and by extension provided an accurate diagnosis with an accuracy of 98.9%, while the usual human interpretation and precision remained constant at 92.1%.
This is great news because it also means that this algorithm could support readings for malaria, syphilis, tuberculosis, influenza, and other illnesses.
What’s even better is that almost anyone can use the mobile app: A pilot study focused on five users who had varying degrees of experience in the field, which included nurses and health workers. trained community health workers, all of whom used an app to record their interpretation of 40 HIV test results as well as transmit the test information to the AI. The small study proved that all participants can use the app without any prior training.
Dr Valerian Turbe of University College London, Center for Nanotechnology said: âHaving spent some time in KwaZulu-Natal with field workers organizing data collection, I have seen how difficult it is for people to access basic health services. If these tools can help train people to interpret the images, you can make a big difference in detecting HIV at a very early stage, which means better access to health care or avoiding incorrect diagnosis. This could have massive implications for people’s lives, especially since HIV is transmissible. “
A digital system to link the laboratory and the health care management structure has also been created. This will aid in the rapid deployment and delivery of diagnostic tests.
Such a tool would be invaluable due to its ability to report rapid diagnostic test results in real time, which could, in the worst case scenario, help healthcare workers manage an outbreak by highlighting hot spots. in various areas where the number of positive tests is high.
Professor Deenan Pillay of University College London, Infection & Immunity said of the project that since digital health research is becoming more mainstream âThere remain serious concerns that the most needy populations around the world will not benefit as much as those in higher income communities. Our work shows how, with the right partnerships and engagement, we can demonstrate utility and benefits for low- and middle-income people.