• Ayush
• Ashish Verma

Pneumonia detection through chest X-ray (CXR) imaging is a crucial diagnostic process that requires expert evaluation. However, manual interpretation is time-consuming, subjective, and prone to errors. This research presents a novel hybrid deep learning model that integrates ResNet34 and MaxViT architectures along with an attention-based feature fusion mechanism to improve pneumonia classification accuracy. The ResNet34 model extracts local features, while MaxViT captures global dependencies, providing a robust representation of lung abnormalities. The attention mechanism optimally fuses these features, ensuring the model emphasizes the most critical pneumonia-related patterns. Performance evaluation on a large CXR dataset shows that our approach significantly outperforms previous models, achieving 99.04% accuracy, 0.9750 Kappa score, 99.48% sensitivity, and 97.78% specificity. The results suggest that this hybrid model can assist radiologists in automating pneumonia diagnosis with high reliability, particularly in resource-constrained settings.

Pneumonia Detection, Deep Learning, ResNet34, MaxViT, Attention Mechanism, Medical Imaging, Chest X-ray.

"Pneumonia Detection Using Hybrid Deep Learning: ResNet34 and MaxViT with Attention Mechanism", International Journal of Emerging Technologies and Innovative Research (www.jetir.org), ISSN:2349-5162, Vol.12, Issue 2, page no.e89-e92, February-2025, Available :http://www.jetir.org/papers/JETIR2502409.pdf

"Pneumonia Detection Using Hybrid Deep Learning: ResNet34 and MaxViT with Attention Mechanism", International Journal of Emerging Technologies and Innovative Research (www.jetir.org | UGC and issn Approved), ISSN:2349-5162, Vol.12, Issue 2, page no. ppe89-e92, February-2025, Available at : http://www.jetir.org/papers/JETIR2502409.pdf