Abstract

Abstract Summary

Overview

Electromagnetic fields (EMFs), primarily from cell phones and mobile towers, affect the brain morphology of humans and other organisms. Prolonged exposure may lead to significant neurological and morphological changes.

Methodology

This preliminary study introduces an innovative, machine learning-based method for detecting EMF effects on brain morphology. Utilizing automatic image processing, geometric features from segmented brain regions of drosophila were analyzed. A genetic algorithm aided the selection of an optimal feature set for machine learning classifiers.

Findings

• Optimal classifier performance was achieved using neural networks with a selected subset of geometric features.
• A statistical test confirmed the significance of improved classifier performance post-feature selection.
• Clear discrimination was observed between the brain images of EMF-exposed and non-exposed drosophila.

Conclusion

This study supports the hypothesis that electromagnetic fields can adversely affect brain morphology, highlighting the critical need for further investigation and validation in broader settings.