They are developing an algorithm for detecting tropical diseases with 95% accuracy

Polytechnic University of Madrid (UPM), together with Spotlab, the National Center of Microbiology of the Carlos III Institute of Health (ISCIII) and the areas of Bioengineering, Biomaterials and Nanomedicine (CIBERBBN) and Infectious Diseases (CIBERINFEC) of the Network Center for Biomedical Research (CIBER). ), developed an artificial intelligence algorithm for diagnosing filariasis using a mobile phone connected to a microscope.

Filariasis is parasitic disease caused by several species of nematodes (roundworms) of the genus Filarioidea. These parasites are transmitted to humans through the bites of infected mosquitoes. exist three main types filariasis affecting humans: lymphatic filariasis, onchocerciasis (river blindness) and loiasis. Diagnosis of filariasis may include identification of microfilariae in the blood or skin, serological tests and, in some cases, imaging tests to detect adult worms.

As part of the study, the authors of the work developed algorithms artificial intelligence to detect microfilariae in the blood, infectious larvae that can transmit filariasis. most common parasites in Africa and Southeast Asia using a mobile phone camera connected to an optical microscope using a 3D printed adapter.

The main authors of the work are an engineer specializing in artificial intelligence, Linlin And Elena Dakal who works in a clinical team, both from the UOM, under the direction of the principal investigators Miguel Angel Luengo (Spotlab), Jose Miguel Rubio (CNM-ISCIII, CIBERINFEC) and Maria Jesus Ledesma (UPM, SIBERBBN). To create this system, the researchers used 115 clinical cases and tested the system in the clinical environment of the National Center for Microbiology ISCIII. The system has an accuracy of about 95%.

“The HuggingSpot app represents a paradigm shift in the fight against filariasis.”

Application HuggingSpot represents a paradigm shift in the fight against filariasis, a disease that affects more than one billion people worldwide. Matching the target product profile for lymphatic filariasis As defined by the World Health Organization (WHO), this technology offers an innovative and affordable solution for improve detection and disease monitoring.

With this application it is possible digitize clinical data and images. While the user views the image on the mobile phone screen, the AI ​​analyzes in real time depending on the magnification used (10x or 40x), generating predictions and delimiting detected parasites into rectangles. When you take a photo, both the images and predictions are saved and the photo count increases. parasite detection. If the user finds parasites not detected by the AI, he can increase their number manually. Once the analysis is complete, the information is loaded into telemedicine platform view and adjust forecasts and share information.

It is acceptable to use a 3D printed adapter. integrate smartphone into a regular microscope. The AI ​​works in real time as the analyst moves the sample and analyzes it using AI.

By allowing healthcare professionals to download and use artificial intelligence models on their mobile devices, the HuggingSpot app democratizes access to advanced diagnostic tools, empowering healthcare professionals and facilitating early detection and requires filariasis.