MACNeXt-Based Bacteria Species Detection.

Area of Science:

• Microbiology
• Computer Science
• Bioinformatics

Background:

• Accurate bacterial identification is crucial for human health, environmental monitoring, and industrial applications.
• Traditional methods like culture and microscopy are time-consuming, costly, and prone to errors.
• Deep learning offers a promising avenue for rapid and objective bacterial classification.

Purpose of the Study:

• To develop a novel, compact deep learning architecture for bacterial species classification.
• To achieve high accuracy and efficiency for potential routine clinical application.
• To introduce the Multiple Activation Network (MACNeXt) for bacterial image analysis.

Main Methods:

• A curated dataset of 18,221 bacterial microscopic images from 24 species was utilized.
• A novel Convolutional Neural Network (CNN) architecture, MACNeXt, was developed.
• MACNeXt features a multi-branch design incorporating GELU and ReLU activation functions for enhanced feature representation.

Main Results:

• MACNeXt achieved 90.97% accuracy, 89.63% precision, 88.64% recall, and 88.99% F1-score on the test set.
• The model demonstrated balanced and stable performance across all bacterial species.
• MACNeXt is a lightweight model with approximately 4.4 million learnable parameters, indicating low computational cost.

Conclusions:

• The developed MACNeXt model is a compact, lightweight, and highly accurate CNN for bacterial species classification.
• Its efficient design makes it suitable for routine clinical use, improving diagnostic speed and reliability.
• This deep learning approach advances bacterial identification, supporting responsible antibiotic stewardship and diagnostics.