Abstract

Abstract Overview

Background: There is a significant global concern about the increased screen time associated with smartphones, tablets, and computers. Research shows that prolonged exposure to blue light emitted from digital screens may contribute to cancer. Leveraging machine learning (ML) techniques has enhanced the prediction of cancer susceptibility, recurrence, and survival.

Objective

To develop an ML model to predict breast cancer risk in women, focusing on various exposure parameters to ionizing and non-ionizing radiation.

Material and Methods

In this analytical study, three ML models—Random Forest (RF), Support Vector Machine (SVM), and Multi-Layer Perceptron Neural Network (MLPNN)—were deployed to analyze data from 603 individuals, comprising 309 breast cancer cases and 294 controls matched by gender and age. Data collection was done via standard face-to-face interviews using a questionnaire.

Results

• The RF, SVM, and MLPNN models achieved high performance in classifying breast cancer cases and healthy controls with mean sensitivity over 97.2%, specificity above 96.4%, and overall accuracy greater than 97.1%.

Conclusion

Machine learning models are effective in predicting breast cancer risk based on exposure history to both ionizing and non-ionizing radiation, including parameters like blue light and screen time. While promising, further research is necessary to verify the automatic diagnosis capabilities of these machine learning techniques for high-accuracy breast cancer detection.