Go to the content (press return)

Implementation of personalized medicine in malignant melanoma patients aided by artificial intelligence

Type of activity
Competitive project
Funding entity
Funding entity code
718/C/2019 La Marató TV3
100.000,00 €
Start date
End date
Artificial Intelligence, Deep Learning, Melanoma, Precision Medicine, Survival Analysis
Research Question: Can we improve the management of patients with melanoma using precision medicine
aided by artificial intelligence?
Background: At present, the management of patients with melanoma is guided by a staging system (AJCC)
using limited clinical and pathological information of the tumor. With this system, stratification of risk and
clinical management (follow-up, diagnostic tests, treatments) are established. However, relevant prognostic
information of the tumor and from the patient is not considered. This fact explains the limited accuracy of the
staging system to predict the risk of the patient and to guide clinical management.
We propose six linked work packages to build out a precision medicine-oriented method called AID-MM
(Artificial Intelligence-Driven Melanoma model) for better guiding clinical management of patients with
Methods: 1. Identify clinical, environmental, socio-economic, behavioral, phenotypical, biochemical,
pharmacological, genetic and medical imaging data which correlate with the patient’s risk to relapse,
metastasize or die on a collaborative database which contains information from 23 different hospitals in
Catalonia (“Xarxa de Centres de Melanoma”) with more than 600000 entries from at approximately 14000
melanoma patients, 2. Create a survival prediction multimodal Deep Learning algorithm to aggregate the
information from a patient into a set of scores to predict disease-free survival and melanoma survival
associated risks, and validation of AIM-MM output on European melanoma patients from data registries from
two International Societies of cancer (EORTC and EADO), 3. Use the model’s outputs to create a new
statistically coherent stratification for patients with melanoma and predict the patient’s outcome under different
treatment options, 5. Upload the AID-MM to an online platform for physicians around the world to predict the
risk of metastasis, relapse, and death of their patients.
Anticipated impact: The “Artificial Intelligence-Driven Melanoma Model” (“AID-MM”) will improve melanoma
patients’ care by the implementation of precision medicine.
Call year
Funcding program
Fundació La Marato de TV3
Funding call
Ajuts a projectes de recerca en l'àmbit de salut MARATÓ DE TV3
Grant institution
Fundació La Marató De Tv3