Application of Artificial Intelligence in the prediction of breast cancer survival in Mexican women
In Mexico, the leading cause of death caused by malignant tumors in women is breast cancer and the general survival of five years treated in facilities of the Public Health System is between 75 and 80%. There are applications that determine the survival of patients with breast cancer, based on the use of drugs that are not prescribed in Mexico, so cancer specialists cannot consider the information offered by these programs for decision-making with patients Mexican. This article describes the development of an expert system that, by applying artificial intelligence techniques, allows the evaluation and prediction of patient survival, based on a model generated with data mining techniques. Rules were obtained from the patterns obtained with data collected from patients with breast cancer since 2006. The development of the system is governed by the Knowledge Discovery from Databases (KDD) methodology, supported by the WEKA tool for modeling data mining techniques. There is a data warehouse of 4,773 women with breast cancer provided by two tertiary hospitals in Mexico City: an INCan cohort of 4,300 patients diagnosed from 2006 to 2013 with a median follow-up of 40.5 months of survival and an INCMSZprovided cohort of 473 patients from 2011 to 2018 with a median of 39 months.