Development of a predictive model for breast cancer based on risk factors, a significant advance in improving early detection of the disease

Abstract

Breast cancer is one of the most common cancers among women, representing a major cause of morbidity and mortality worldwide. In the Democratic Republic of Congo, the lack of systematic screening exacerbates the difficulties in identifying high-risk patients at an early stage. This study aims to develop a predictive model based on machine learning, drawing on an analysis of breast cancer risk factors. The main objective is to facilitate early detection and improve patient care by proactively identifying those at increased risk of developing breast cancer.

The methodology of this study is based on the Team Data Science Process (TDSP), which organises and optimises the various stages of predictive model development, from data collection to results evaluation. Clinical data and variables related to family history, health behaviours, and physiological characteristics were incorporated into the model. Machine learning algorithms, including Support Vector Machine (SVM), were used to build the model, offering 97% accuracy in predicting breast cancer risk.

This model was then deployed in a production environment with an intuitive user interface, allowing physicians to easily analyse the results. The results showed that integrating such a tool could significantly improve early detection capabilities in settings where screening resources are limited, while reducing treatment costs associated with late diagnoses.

The findings of this article pave the way for the implementation of a national screening programme based on predictive tools, thereby helping to reduce healthcare disparities and improve the prognosis for at-risk patients. Future improvements to the model could focus on integrating new biomedical data and adjusting models for more diverse clinical settings.

Keywords: model, cancer, early, factors, detection, disease, predictive, breast, significant.

Received Date: October 20, 2024

Accepted Date: November 11, 2025

Published Date: December 01, 2025

Available Online at: https://www.ijsrisjournal.com/index.php/ojsfiles/article/view/514