A new research work from the Vall de Hebron Research Institute (VHIR) and the Universitat Politècnica de Catalunya (UPC) won the award for the best communication at the XXV National Congress of the Sociedad Española de Enfermedades Infecciosas y Microbiología Clínica (SEIMC). The project is the development of a new automatic diagnostic method for malaria based on artificial intelligence, and the integration of the technology in a mobile App. The study was presented by Carles Rubio, predoctoral student of the microbiology group (VHIR), under the direction of Dr. Joan Joseph, principal researcher of the same group. During the 3 days of the congress, 251 oral presentations were made, of which only 6 obtained this distinction. As part of the award received, the researcher was interviewed by the congress tv channel.

This is a multicenter and multidisciplinary project in which, apart of the mentioned VHIR group, are involved the Center for Communicable Diseases and International Health Drassanes - Vall d'Hebron; the groups of the UPC: Computational Biology and Complex Systems, Database Technologies and Information Systems and Image Processing. The Probitas Foundation also participated and co-funded the research.

In 2020, 241 million cases of malaria were diagnosed in 85 countries, according to the World Health Organization (WHO). Most cases are reported in Sub-Saharan Africa, where in many cases, there are not enough human and technical resources to make an effective diagnosis with current techniques. The aim of the project is to create neural network models for the detection of malaria parasites in digital images of large-drop blood samples. The UPC has developed a universal model with 3D parts adaptable to any conventional optical microscope, which allows for automating the movements of the slide and the autofocus of the sample. On the other hand, this new technology is integrated into a mobile device application that controls the movements of the microscope and makes it possible to detect and identify malaria parasites by analyzing the digital images it captures using artificial intelligence tools. To train the neural network, 2225 images were captured and labeled from patients diagnosed with malaria at the Drassanes - Vall de Hebron International Health Unit, samples from collections and from hospitals in Africa. At the end of the process, more than 90% accuracy was achieved.

This project has the technical support of the WHO and in the future seeks to implement this new technology for the diagnosis of malaria in low-and-middle-income countries. At the same time, the researchers propose to adapt the technology and application to the diagnosis of some Neglected Tropical Diseases (NTD) and other infectious diseases such as tuberculosis.