AI System MEL-IA Automates Classification of Skin Lesions in Hospital Workflows

AI System Tackles Global Skin Cancer Challenge

The University of Alicante and Sant Joan d'Alacant University Hospital have launched MEL-IA (MobilE skin Lesion dIAgnosis), an AI system designed to automate the classification of skin lesions1

1

2

2

. With nearly 1.5 million new skin cancer cases diagnosed in 2024 according to World Health Organization data, this AI tool for skin lesions addresses a pressing global health challenge where early detection remains vital to improving clinical outcomes.

Comprehensive Technology Package Integrates Into Hospital Workflows

MEL-IA delivers more than just an algorithm. The system incorporates a mobile app to capture images of skin lesions such as moles or spots and record clinical data, an AI model to perform classification, and complex integration models that securely connect all data with hospital systems1

1

. This comprehensive technology package embeds directly into a hospital's IT network, linking image capture, AI analysis, secure storage, and confidential exchange of clinical information. The Bio-inspired Engineering and Health Informatics (IBIS) research group at the University of Alicante, which has spent over a decade researching clinical decision-support systems for diagnosing skin lesions, developed the system alongside IT specialists from Sant Joan d'Alacant University Hospital2

2

.

MEL-IA Classifies Five Major Skin Lesion Types

Unlike tools designed solely to distinguish between malignant and benign lesions, MEL-IA provides differential categorization across five major skin lesion types: melanoma, naevus, basal cell carcinoma, actinic keratosis, and benign keratosis1

1

. Researchers integrated over 15,000 dermatoscopic images alongside clinical patient data, including age, sex, and anatomical lesion location, to enhance classification accuracy during training and validation. This multimodal approach allows the clinical decision-support tool to consider contextual patient information beyond visual analysis alone.

System Achieves 86% Accuracy in Live Hospital Deployment

Results indicate that MEL-IA achieves an overall accuracy of 86%. For melanoma detection, the system achieved a sensitivity of 88%, while reaching 92% for basal cell carcinoma1

1

. The highest overall scores corresponded to naevi and basal cell carcinomas. The technology moved beyond the experimental stage to deployment at Sant Joan d'Alacant University Hospital to evaluate performance in a live healthcare environment. The team processed 980 dermatological studies with response times under one second2

2

, demonstrating the system's capability to support healthcare staff in evaluating lesions without disrupting clinical workflows.

Longitudinal Tracking and Future Expansion Plans

The system maintains a longitudinal record of lesions, embedding images, diagnoses, and medical observations over time1

1

. This feature enables clinicians to track changes in skin lesions across multiple patient visits, potentially improving monitoring of suspicious lesions. Researchers emphasize that MEL-IA serves as a clinical decision-support tool rather than an autonomous diagnostic system. Key next steps include conducting prospective studies with healthcare professionals, testing direct image acquisition via smartphones, and expanding the range of lesion types the system can classify2

2

. The findings were published in the Journal of Medical Systems by Alberto de Ramón, Daniel Ruiz, Marcelo Saval, and Pablo Candela from the University of Alicante, along with José María Salinas and Diego Guijarro from the hospital's IT Service. As AI-driven healthcare applications continue advancing, watch for expanded deployment of MEL-IA across additional hospitals and potential integration with smartphone cameras to enable broader access to skin lesion screening.