Artificial Intelligence System Streamlines Skin Cancer Detection in Hospitals
The University of Alicante and Sant Joan d'Alacant University Hospital launched MEL-IA, an integrated AI system that automatically classifies skin lesions to support early cancer detection. The platform combines a mobile application for image capture, an AI classification model, and secure hospital network integration to streamline clinical workflows. The technology represents over a decade of research into clinical decision-support systems for dermatological diagnosis.
The MEL-IA system addresses a significant clinical need by automating the initial screening phase of skin lesion evaluation. Rather than simply flagging suspicious growths, the platform distinguishes among five distinct lesion categories, allowing clinicians to prioritize cases and allocate resources more efficiently. The developers trained the AI using over 15,000 dermatoscopic images combined with patient metadata such as age and lesion location, enabling the system to recognize patterns that correlate with diagnostic outcomes.
The integration into hospital infrastructure represents a critical distinction from standalone diagnostic applications. By connecting image capture, analysis, secure data storage, and clinical record systems into a unified workflow, MEL-IA reduces friction in clinical decision-making. Testing at Sant Joan d'Alacant University Hospital demonstrated rapid processing—completing nearly 1,000 analyses in under one second each—while maintaining historical records of lesion progression for individual patients.
AI-assisted dermatological screening could potentially reduce diagnostic delays and improve early cancer detection rates, particularly in regions with limited specialist availability. Healthcare systems might benefit from accelerated triage workflows and more consistent preliminary assessments across providers. However, outcomes depend heavily on clinical adoption, user training, and whether the 86% accuracy translates effectively to diverse patient populations. The technology's explicit positioning as a decision-support tool rather than autonomous diagnostic system may influence how clinicians integrate it into existing practices and medico-legal frameworks.