TRANSFER LEARNING FOR BINARY PNEUMONIA CLASSIFICATION FROM CHEST X-RAY IMAGES: A COMPARATIVE STUDY OF DENSENET121, DENSENET169, AND DENSENET201

Arabboev M. ., Begmatov S. ., Bobojanov S. . & Nishanov A. .


Annotatsiya

This paper presents a comparative study of three DenseNet-based transfer learning models for binary pneumonia classification from chest X-ray images. DenseNet121, DenseNet169, and DenseNet201 were evaluated under a unified framework using ImageNet-pretrained backbones, a lightweight classification head, online augmentation, two-stage optimization, and validation AUC-based early stopping. All images were resized to 224 × 224 pixels and divided into training, validation, and test subsets. The test set contained 587 radiographs, including 159 NORMAL and 428 PNEUMONIA cases. DenseNet169 achieved the best overall performance, with 93.70% accuracy, 0.9796 AUC, and 96.88% precision. DenseNet121 showed similarly strong results with slightly higher sensitivity, while DenseNet201 achieved the highest sensitivity (98.36%) but substantially lower specificity. These findings indicate that DenseNet169 provides the most balanced performance for binary pneumonia screening, whereas DenseNet201 may be more suitable when minimizing missed pneumonia cases is the primary priority.


Adabiyotlar

  1. D. S. Kermany et al., “Identifying Medical Diagnoses and Treatable Diseases by Image-Based Deep Learning,” Cell, vol. 172, no. 5, pp. 1122-1131.e9, 2018.
  2. P. Rajpurkar et al., “CheXNet: Radiologist-Level Pneumonia Detection on Chest X-Rays with Deep Learning,” 2017, arXiv preprint arXiv:1711.05225
  3. G. Huang, Z. Liu, L. van der Maaten, and K. Q. Weinberger, “Densely Connected Convolutional Networks,” in IEEE Conference on Computer Vision and Pattern Recognition (CVPR 2017), 2017, pp. 4700–4708.
  4. S. J. Pan and Q. Yang, “A Survey on Transfer Learning,” IEEE Trans. Knowl. Data Eng., vol. 22, no. 10, pp. 1345–1359, Oct. 2010.
  5. X. Wang, Y. Peng, L. Lu, Z. Lu, M. Bagheri, and R. M. Summers, “ChestX-ray8: Hospital-scale Chest X-ray Database and Benchmarks on Weakly-Supervised Classification and Localization of Common Thorax Diseases,” in IEEE Conference on Computer Vision and Pattern Recognition (CVPR 2017), 2017, pp. 3462–3471.
  6. C. Ortiz-Toro, A. García-Pedrero, M. Lillo-Saavedra, and C. Gonzalo-Martín, “Automatic detection of pneumonia in chest X-ray images using textural features,” Computers in Biology and Medicine, vol. 145, 105466, 2022, doi: 10.1016/j.compbiomed.2022.105466
  7. F. Alshanketi, A. Alharbi, M. Kuruvilla, V. Mahzoon, S. T. Siddiqui, N. Rana, and A. Tahir, “Pneumonia Detection from Chest X-Ray Images Using Deep Learning and Transfer Learning for Imbalanced Datasets,” Journal of Imaging Informatics in Medicine, vol. 38, pp. 2021–2040, 2025, doi: 10.1007/s10278-024-01334-0
  8. S. Sharma and K. Guleria, “A Deep Learning based model for the Detection of Pneumonia from Chest X-Ray Images using VGG-16 and Neural Networks,” Procedia Computer Science, vol. 218, pp. 357–366, 2023, doi: 10.1016/j.procs.2023.01.018
  9. Y. Hedhoud, T. Mekhaznia, and M. Amroune, “An improvement of the CNN-XGboost model for pneumonia disease classification,” Polish Journal of Radiology, vol. 88, pp. e483–e493, 2023, doi: 10.5114/pjr.2023.132533