Among malignant eye cancers, uveal melanoma is the most common in adults. As the eyes do not have a lymphatic system, it is highly likely that tumour cells can spread through the bloodstream and metastasise to other parts of the body. Therefore, early and effective detection of cancer is essential for the treatment, quality of life and life expectancy of oncology patients. The Institute Curie reports that in recent decades, progress in this area has not been significant, given the great variability and complexity of the disease. This study proposes a methodology for the diagnosis and detection of uveal melanoma cancer. Thirteen models have been generated from data from the New York Eye Cancer Center and the Chinese Academy of Sciences. The models are based on mathematical and computational intelligence techniques, as well as secondary developments including the use of the Gabor filter as a data augmentation technique and the optimisation of automatic iris segmentation algorithms. The results have been supported by the publication of eight scientific articles in international journals, some of which belong to Q1 and Q2. This is one of the first proposals for the development of a computer-aided diagnosis tool for uveal melanoma.