🤖 AI Summary
This study addresses the challenges of rabies diagnosis in low-resource regions such as Africa and Asia, where reliance on fluorescence microscopy and expert interpretation is hindered by scarce samples and a shortage of trained personnel. The authors propose a deep learning–based automated diagnostic approach and systematically evaluate multiple transfer learning architectures—including EfficientNetB0, EfficientNetB2, VGG16, and Vision Transformer (ViT)—alongside various data augmentation strategies on a small dataset of only 155 fluorescence images. They demonstrate for the first time that TrivialAugmentWide effectively preserves critical fluorescent features while enhancing model generalization. Among the tested configurations, EfficientNetB0 combined with tailored augmentation achieves optimal performance on cropped images. The results indicate that reliable and rapid automated diagnosis is attainable even under extreme data scarcity and class imbalance, and the method has already been deployed as an online tool for real-world use.
📝 Abstract
Rabies remains a major public health concern across many African and Asian countries, where accurate diagnosis is critical for effective epidemiological surveillance. The gold standard diagnostic methods rely heavily on fluorescence microscopy, necessitating skilled laboratory personnel for the accurate interpretation of results. Such expertise is often scarce, particularly in regions with low annual sample volumes. This paper presents an automated, AI-driven diagnostic system designed to address these challenges. We developed a robust pipeline utilizing fluorescent image analysis through transfer learning with four deep learning architectures: EfficientNetB0, EfficientNetB2, VGG16, and Vision Transformer (ViTB16). Three distinct data augmentation strategies were evaluated to enhance model generalization on a dataset of 155 microscopic images (123 positive and 32 negative). Our results demonstrate that TrivialAugmentWide was the most effective augmentation technique, as it preserved critical fluorescent patterns while improving model robustness. The EfficientNetB0 model, utilizing Geometric & Color augmentation and selected through stratified 3fold cross-validation, achieved optimal classification performance on cropped images. Despite constraints posed by class imbalance and a limited dataset size, this work confirms the viability of deep learning for automating rabies diagnosis. The proposed method enables fast and reliable detection with significant potential for further optimization. An online tool was deployed to facilitate practical access, establishing a framework for future medical imaging applications. This research underscores the potential of optimized deep learning models to transform rabies diagnostics and improve public health outcomes.