Researchers from Fraunhofer Portugal AICOS (FhP-AICOS) have developed a digital solution with the aim of making the process of referencing skin lesions more efficient. The technology that results from the Derm.AI project consists of two components: a mobile application and an Artificial Intelligence (AI) software.
The mobile application developed helps the general and family doctor in the acquisition of dermatological images. Using Computer Vision and Machine Learning techniques, the application allows the automatic acquisition of macroscopic and anatomical images with the necessary quality for the specialist to make the diagnosis. A pilot has been implemented within the Derm.AI in which the application has been tested, for over 10 months, and had the collaboration of six general and family doctors and a dermatologist from the Unidade de Saúde Local da Guarda (ULS Guarda). The “ease of image acquisition” and “image quality” were the unanimous considerations resulting from this pilot.
Regarding the software, the team of researchers developed an Artificial Intelligence algorithm that allows prioritizing cases referred by family doctors to dermatology services in hospitals. By using a system of colours (green, yellow and red), the algorithm sorts the priority cases, according to the degree of urgency (normal, priority and high priority) to help the dermatologist responsible for the categorisation. “The software does not make a decision. It simply provides a pre-selection based on probability. The actual examination and diagnosis are still in the hands of dermatologists or skin-cancer specialists”, explains FhP-AICOS senior researcher Maria Vasconcelos. For the construction of this algorithm, the integration of the clinical knowledge of the project partners was essential.
“In recent years, clinicians at the National Health Service (SNS) have been increasingly concerned with spotting skin cancer early. However, in Portugal, with the reduced number of dermatology specialists in the SNS, it is important to improve the referral process, ensuring that the information collected is of good quality and helping to prioritize the referred cases”, adds the researcher.
In what scientific results is concerned, over the three years of the project, Derm.AI led to nine scientific publications, some of which were presented at international conferences, five master theses and two awards, with HINNT standing out in the Clinical Outcomes category.
The Derm.AI project was coordinated by Fraunhofer Portugal AICOS and developed in partnership with the Serviços Partilhados do Ministério da Saúde (SPMS), the IPO Coimbra, the Centro Hospitalar Universitário do Porto (CHUPorto), and the Unidade de Saúde Local da Guarda (ULS Guarda).
The Project “DERM.AI – Usage of Artificial Intelligence to Power Teledermatological Screening”, with reference DSAIPA/AI/0031/2018, is supported by national funds through ‘FCT – Foundation for Science and Technology’ in the scope of “Projects of Scientific Research and Technological Development in Data Science and Artificial Intelligence in Public Administration 2018”.