Intelligent System for the Detection and Prediction of Endometriosis at Maria Auxiliadora Hospital in Lima, Perú

Bautista A.; Tardillo J.; Alvarado C. · ADMIT 2025 - Conference Proceedings: 2025 4th International Conference on Algorithms, Data Mining, and Information Technology (2025) · DOI: 10.1109/admit67050.2025.11337114

Resumen

This study presents the design and validation of an intelligent system based on machine learning techniques, specifically the Random Forest algorithm, aimed at the early detection and prediction of endometriosis at the 'Hospital María Auxiliadora' in Lima, Peru. Endometriosis is a chronic gynecological condition characterized by nonspecific symptoms, which complicate timely diagnosis and typically leads to delays ranging from 6 to 12 years, adversely affecting patients' quality of life. In response to this issue, a web-based solution was developed to enable healthcare professionals to input structured clinical data-such as age, menstrual cycle duration, presence of pain, and infertility-in order to obtain an accurate prediction of endometriosis risk (achieving 99% precision and 98% recall). The system was validated through surveys administered by specialists in the field of gynecology, who rated its usability, reliability, and maintainability positively. The results confirm that the tool is functional, accurate, and accessible, even in clinical environments with limited infrastructure. The integration of machine learning models such as Random Forest provides a non-invasive, efficient, and adaptable diagnostic alternative with the potential to improve diagnostic timelines, enable more timely and personalized medical care, and enhance the quality of life for women affected by this disease in Peru.

Instituciones: Universidad Peruana de Ciencias Aplicadas