Deep Learning for Medical Image Analysis: CNN-based Pneumonia Detection on Chest X-Rays

Summary

Haveri and Vajjhala (2026) tested whether a lightweight EfficientNetV2B0 convolutional neural network, trained with ImageNet transfer learning under CLAIM and TRIPOD-AI reporting standards, can accurately and reliably detect pneumonia in paediatric chest X-rays from the Kermany dataset (624-image test set). The model achieved an AUC of 0.967 (95% CI 0.953–0.979), accuracy of 0.894 and pneumonia recall of 0.98, with lower normal-class recall of 0.752 and moderate calibration (ECE = 0.083, Brier Score = 0.091). The paper is a single-dataset paediatric study and does not claim external validation.

Practical implication. The compact model of about 7.1 million parameters is presented as suited to edge devices, telehealth networks and low-resource healthcare settings; its high sensitivity on one paediatric dataset favours pneumonia triage workflows.

The Problem

Deep learning models for chest X-ray pneumonia detection are often reported without confidence intervals, calibration or reporting-standard compliance, limiting their clinical credibility.

The Methodology

Binary medical image classification with an EfficientNetV2B0 convolutional neural network and ImageNet transfer learning on the Kermany paediatric chest X-ray dataset, following CLAIM and TRIPOD-AI reporting, with bootstrapped confidence intervals, DeLong test, calibration metrics and Grad-CAM.

The Core Finding

A lightweight EfficientNetV2B0 model trained under CLAIM and TRIPOD-AI reporting standards achieved an AUC of 0.967 (95% CI 0.953–0.979) and pneumonia recall of 0.98 on paediatric chest X-rays, missing only about 2% of true cases. Lower normal-class recall (0.752) reflects a false-positive tendency that is a clinically preferable trade-off in triage.

The Citation

Haveri, K., & Vajjhala, N. R. (2026). Deep Learning for Medical Image Analysis: CNN-based Pneumonia Detection on Chest X-Rays. In 2026 6th International Conference on Pervasive Computing and Social Networking (ICPCSN) (pp. 156–161). IEEE. https://doi.org/10.1109/ICPCSN68523.2026.11543623 Copy APA

The summary, key findings, methods, implications and relevance notes on this page are this website’s own description of the paper, written from the published abstract and text. The official abstract and citation details are given further down.

The study in detail

What question does this paper answer?

Can a lightweight EfficientNetV2B0 convolutional neural network, trained with transfer learning under strict reproducibility protocols, accurately and reliably detect pneumonia in paediatric chest X-rays?

What did the study find, in detail?

On the Kermany paediatric chest X-ray dataset (624-image test set), the EfficientNetV2B0 model achieved an AUC of 0.967 (95% CI 0.953–0.979; DeLong p < 0.001), accuracy of 0.894, F1-score of 0.920 and pneumonia recall of 0.98, missing only about 2% of true pneumonia cases. Calibration was moderate (ECE = 0.083, Brier Score = 0.091), and normal-class recall was lower at 0.752.

Why does it matter?

The results show that a compact model of about 7.1 million parameters can deliver clinically meaningful pneumonia detection without the heavy computation of larger networks, making it suited to edge devices, telehealth networks and low-resource healthcare settings. Its high sensitivity favours triage workflows, where missing pneumonia is costlier than an unnecessary follow-up, and the authors outline integration with PACS and CMMS platforms.

Methods and data

Standards, datasets and methods named in this study

• Kermany paediatric chest X-ray dataset Dataset
  Source of the paediatric chest X-ray dataset on which the model was trained and tested.
  Kermany, D. S., Goldbaum, M., Cai, W., Valentim, C. C. S., Liang, H., Baxter, S. L., et al. (2018). Identifying medical diagnoses and treatable diseases by image-based deep learning. Cell, 172(5), 1122–1131.e9. https://doi.org/10.1016/j.cell.2018.02.010
• EfficientNetV2 Model
  The EfficientNetV2B0 architecture is the model evaluated.
  Tan, M., & Le, Q. V. (2021). EfficientNetV2: Smaller models and faster training. Proceedings of the 38th International Conference on Machine Learning (PMLR 139).
• ImageNet Dataset
  The model is initialised by transfer learning from ImageNet.
  Deng, J., Dong, W., Socher, R., Li, L.-J., Li, K., & Fei-Fei, L. (2009). ImageNet: A large-scale hierarchical image database. 2009 IEEE Conference on Computer Vision and Pattern Recognition, 248–255. https://doi.org/10.1109/CVPR.2009.5206848
• CLAIM — Checklist for Artificial Intelligence in Medical Imaging Reporting guideline
  A reporting standard the study follows.
  Mongan, J., Moy, L., & Kahn, C. E., Jr. (2020). Checklist for Artificial Intelligence in Medical Imaging (CLAIM): A guide for authors and reviewers. Radiology: Artificial Intelligence, 2(2), e200029. https://doi.org/10.1148/ryai.2020200029
• TRIPOD+AI statement Reporting guideline
  A reporting standard the study follows; the paper names it TRIPOD-AI.
  Collins, G. S., Moons, K. G. M., Dhiman, P., Riley, R. D., Beam, A. L., Van Calster, B., et al. (2024). TRIPOD+AI statement: Updated guidance for reporting clinical prediction models that use regression or machine learning methods. BMJ, 385, e078378. https://doi.org/10.1136/bmj-2023-078378

Definitions and key terms

Transfer learning

Adapting a model pretrained on a large dataset (here ImageNet) to a specific task such as chest X-ray classification, useful when annotated clinical data are limited. — as used in Haveri & Vajjhala (2026)

Grad-CAM (Gradient-weighted Class Activation Mapping)

A technique that produces heatmaps highlighting the image regions most influential in a model’s classification decision. — as used in Haveri & Vajjhala (2026)

Expected Calibration Error (ECE)

The weighted average absolute difference between a model’s predicted confidence and its observed accuracy across probability bins. — as used in Haveri & Vajjhala (2026)

All key terms across the publications →

Abstract

Pneumonia remains one of the leading causes of mortality worldwide, particularly among paediatric populations, and timely radiographic diagnosis is essential for effective clinical intervention. The proliferation of pervasive computing and communication technologies has enabled the deployment of deep learning models within distributed healthcare systems, facilitating rapid and automated medical image analysis at scale. This study presents a convolutional neural network framework based on EfficientNetV2B0 with transfer learning for binary pneumonia classification from paediatric chest X-ray images. The model employs a two-phase training strategy comprising frozen-backbone feature extraction followed by full fine-tuning, augmented by deterministic data preprocessing, class-weighted loss computation, and Gradient-weighted Class Activation Mapping for interpretability. The input to the model consists of 224×224 pixel RGB chest radiographs, and the output is a sigmoid-activated probability score indicating the likelihood of pneumonia. Evaluated on the benchmark Kermany dataset of 5,863 radiographs (1,583 normal and 4,280 pneumonia cases across training, validation, and test splits), the proposed framework achieves an area under the receiver operating characteristic curve (AUC) of 0.967, confirmed as statistically significant via DeLong testing (p < 0.001), and an overall accuracy of 0.894, with a pneumonia recall of 0.98. The model further yields an Expected Calibration Error (ECE) of 0.083 and a Brier Score of 0.091. Bootstrapped confidence intervals and calibration analysis confirm the robustness and reliability of the predictions. The results demonstrate that lightweight EfficientNetV2 architectures, when combined with rigorous reproducibility protocols compliant with CLAIM and TRIPOD-AI standards, can deliver clinically meaningful diagnostic performance suitable for integration into pervasive healthcare information and communication technology platforms.

Abstract as published in 2026 6th International Conference on Pervasive Computing and Social Networking (ICPCSN).

Keywords: deep learning; convolutional neural networks; pneumonia detection; chest X-ray classification; EfficientNetV2; transfer learning; medical image analysis; pervasive computing

Limitations

• The Kermany dataset contains only paediatric chest radiographs from a single medical centre, limiting demographic and scanner diversity and external validity; cross-dataset evaluation (e.g., NIH ChestX-ray14, RSNA, CheXpert) is needed.
• The model was trained on a relatively small dataset, so performance consistency on larger, more heterogeneous datasets remains to be verified.
• Calibration analysis shows systematic under-confidence in low-probability regions (ECE = 0.083), so probabilities may need post-hoc calibration before clinical risk stratification.
• Grad-CAM captures only high-level activation patterns and does not guarantee clinically valid explanations.

Practical implications

• For hospitals, the authors present EfficientNetV2B0, with approximately 7.1 million parameters, as suitable for edge deployment and integration with PACS and CMMS platforms via a containerized inference pipeline.
• On the 624-image test set, recall for normal chest X-rays was 0.752, indicating a tendency toward false-positive pneumonia predictions that the authors regard as a clinically preferable trade-off in triage.
• Healthcare settings may need post-hoc calibration of the EfficientNetV2B0 model’s probabilities before clinical risk stratification, because calibration analysis shows systematic under-confidence in low-probability regions (ECE = 0.083).
• The Kermany dataset contains only paediatric chest radiographs from a single medical centre, limiting demographic and scanner diversity and external validity; cross-dataset evaluation is needed.

Drawn from Haveri & Vajjhala (2026), 2026 6th International Conference on Pervasive Computing and Social Networking (ICPCSN), pp. 156–161, IEEE.

When this research may be relevant

This article may be relevant to literature on deep learning for pneumonia detection and chest radiograph analysis, lightweight CNNs for edge or low-resource deployment, reproducibility and calibration reporting in medical AI, and explainability with Grad-CAM. It is a single-dataset paediatric study and does not claim external validation.

Research topics addressed: deep learning; convolutional neural networks; pneumonia detection; chest X-ray; medical image analysis; EfficientNetV2B0; transfer learning; paediatric radiology; Grad-CAM; explainable AI; model calibration; reproducibility; healthcare informatics; edge computing

How to cite

APA 7

Haveri, K., & Vajjhala, N. R. (2026). Deep Learning for Medical Image Analysis: CNN-based Pneumonia Detection on Chest X-Rays. In 2026 6th International Conference on Pervasive Computing and Social Networking (ICPCSN) (pp. 156–161). IEEE. https://doi.org/10.1109/ICPCSN68523.2026.11543623Copy APA citation

MLA 9

Haveri, Katia, and Narasimha Rao Vajjhala. “Deep Learning for Medical Image Analysis: CNN-based Pneumonia Detection on Chest X-Rays.” 2026 6th International Conference on Pervasive Computing and Social Networking (ICPCSN), IEEE, 2026, pp. 156–61. https://doi.org/10.1109/ICPCSN68523.2026.11543623.Copy MLA citation

IEEE

K. Haveri and N. R. Vajjhala, “Deep Learning for Medical Image Analysis: CNN-based Pneumonia Detection on Chest X-Rays,” in 2026 6th International Conference on Pervasive Computing and Social Networking (ICPCSN), 2026, pp. 156–161, doi: 10.1109/ICPCSN68523.2026.11543623.Copy IEEE citation

@inproceedings{haveri2026cnn,
  title = {Deep Learning for Medical Image Analysis: CNN-based Pneumonia Detection on Chest X-Rays},
  author = {Haveri, Katia and Vajjhala, Narasimha Rao},
  booktitle = {2026 6th International Conference on Pervasive Computing and Social Networking (ICPCSN)},
  pages = {156--161},
  year = {2026},
  publisher = {IEEE},
  doi = {10.1109/ICPCSN68523.2026.11543623},
  url = {https://doi.org/10.1109/ICPCSN68523.2026.11543623}
}

TY  - CPAPER
TI  - Deep Learning for Medical Image Analysis: CNN-based Pneumonia Detection on Chest X-Rays
AU  - Haveri, Katia
AU  - Vajjhala, Narasimha Rao
T2  - 2026 6th International Conference on Pervasive Computing and Social Networking (ICPCSN)
PY  - 2026
DA  - 2026/05/06
SP  - 156
EP  - 161
PB  - IEEE
SN  - 979-8-3315-7236-5
DO  - 10.1109/ICPCSN68523.2026.11543623
UR  - https://www.narasimharao.net/research/cnn-pneumonia-detection-chest-xray-efficientnetv2/
LA  - en
ER  - 

Links and files

• DOI: https://doi.org/10.1109/ICPCSN68523.2026.11543623
• This page as text: https://www.narasimharao.net/research/cnn-pneumonia-detection-chest-xray-efficientnetv2/paper.md
• Permanent address of this page: https://www.narasimharao.net/research/cnn-pneumonia-detection-chest-xray-efficientnetv2/