Mycteoma is an endemic tropical disease. It is caused by fungi (Eumycetoma) or true bacteria (Actinomycetoma). The cure rate of Mycetoma in Sudan is only 43.3%, and the mean duration of treatment is around 18 months.

Retrospective data was collected from 348 Actinmoycetoma patients seen at the Mycetoma Research Center between 2012 and 2015. Of the 348 patients, 77 patients were cured, and their data was selected to predict the duration of treatment using machine learning algorithms (linear regression, decision trees, and random forests).

Random forests was the best model to predict the duration of treatment, with a good MSE in training, validation, and test set.