Citation :

Abhijit Nath, Sheikh Wakie Masood, Shahin Ara Begum, Ravi Kannan, "Hybrid Resampling and Ensemble Learning for Predicting Radiation-Induced Pneumonitis in Head and Neck Cancer," International Journal of Engineering Trends and Technology (IJETT), vol. 74, no. 8, pp. 50-67, 2026. Crossref, https://doi.org/10.14445/22315381/IJETT-V74I8P104

Abstract

Radiation-Induced Pneumonitis (RP) is a clinically important but difficult-to-predict toxicity in Head and Neck Cancer (HNC), particularly when only routinely collected hospital variables are available. This study developed a recall-oriented machine learning framework for RP prediction using clinical and treatment-related data from 2,482 HNC patients treated with radiotherapy between 2018 and 2024. Twenty-five baseline variables and four engineered interaction terms were analysed using a stratified 70/30 train–test split. Three phase-wise experiments were performed: Phase I on the original imbalanced training data, Phase II with SMOTETomek-based class balancing, and Phase III as an extension of Phase II with Gaussian-noise data augmentation. XGBoost, LightGBM, CatBoost, and SVM were evaluated as individual classifiers, while the proposed stacking classifier combined XGBoost, LightGBM, and CatBoost with Logistic Regression as the meta-learner. Hyperparameters were optimized by Bayesian search with stratified cross-validation, and thresholds were selected from training-only out-of-fold probabilities using F2-score as the primary criterion. In Phase I, the stacking classifier achieved the highest F2-score 0.57 and recall 0.86. In Phase II, the stacking classifier achieved the highest F2-score 0.58 and ROC-AUC 0.67. In Phase III, it again performed best overall, with a precision 0.24, recall 0.92, F2-score 0.59, and ROC-AUC 0.65. These findings support stacking-based prediction as a screening-oriented aid for identifying high-risk RP cases from routinely available hospital records.

Keywords

Head and Neck Cancer, Radiotherapy, Radiation-Induced Pneumonitis, Data augmentation, Ensemble learning.

References

[1] Suzanne Tanya Nethan, Shalini Gupta, and Saman Warnakulasuriya, Risk Factors for Oral Squamous Cell Carcinoma in the Indian Population, Microbes and Oral Squamous Cell Carcinoma, Springer, Singapore, pp. 9-40, 2022.
[CrossRef] [Google Scholar] [Publisher Link]

[2] Debojyoti Roy et al., “Profiling of Oral Squamous Cell Carcinoma of Lower Assam Region – A Hospital-based Retrospective Study,” Contemporary Clinical Dentistry, vol. 15, no. 4, pp. 240-243, 2024.
[CrossRef] [Google Scholar] [Publisher Link]

[3] Feihu Chen et al., “Re-Evaluating the Risk Factors for Radiation Pneumonitis in the Era of Immunotherapy,” Journal of Translational Medicine, vol. 21, no. 1, pp. 1-15, 2023.
[CrossRef] [Google Scholar] [Publisher Link]

[4] Marisol Arroyo-Hernández et al., “Radiation-Induced Lung Injury: Current Evidence,” BMC Pulmonary Medicine, vol. 21, no. 1, pp. 1-12, 2021.
[CrossRef] [Google Scholar] [Publisher Link]

[5] Fengsong Ye et al., “Predicting Radiation Pneumonitis in Lung Cancer: A EUD-based Machine Learning Approach for VMAT Patients,” Frontiers in Oncology, vol. 14, 2024.
[CrossRef] [Google Scholar] [Publisher Link]

[6] Zhi Chen et al., “Predicting Radiation Pneumonitis in Lung Cancer using Machine Learning and Multimodal Features: A Systematic Review and Meta-Analysis of Diagnostic Accuracy,” BMC Cancer, vol. 24, no. 1, pp. 1-18, 2024.
[CrossRef] [Google Scholar] [Publisher Link]

[7] Yoshiyuki Katsuta et al., “Prediction of Radiation Pneumonitis with Machine Learning using 4D-CT-based Dose-Function Features,” Journal of Radiation Research, vol.  63, no. 1, pp. 71-79, 2022.
[CrossRef] [Google Scholar] [Publisher Link]

[8] N.V. Chawla et al., “SMOTE: Synthetic Minority Over-Sampling Technique,” Journal of Artificial Intelligence Research, vol. 16, pp. 321-357, 2002.
[CrossRef] [Google Scholar] [Publisher Link]

[9] Haibo He, and Edwardo A. Garcia, “Learning from Imbalanced Data,” IEEE Transactions on Knowledge and Data Engineering, vol. 21, no. 9, pp. 1263-1284, 2009.
[CrossRef] [Google Scholar] [Publisher Link]

[10] Heesoon Sheen et al., “Radiomics-based Hybrid Model for Predicting Radiation Pneumonitis: A Systematic Review and Meta-Analysis,” Physica Medica, vol. 123, pp. 1-9, 2024.
[CrossRef] [Google Scholar] [Publisher Link]

[11] David H. Wolpert, “Stacked Generalization,” Neural Networks, vol. 5, no. 2, pp. 241-259, 1992.
[CrossRef] [Google Scholar] [Publisher Link]

[12] Tianqi Chen, and Carlos Guestrin, “XGBoost: A Scalable Tree Boosting System,” Proceedings of the 22^nd ACM SIGKDD International Conference on Knowledge Discovery and Data Mining, Association for Computing Machinery, New York, NY, United States, pp. 785-794, 2016.
[CrossRef] [Google Scholar] [Publisher Link]

[13] Guolin Ke et al., “LightGBM: A Highly Efficient Gradient Boosting Decision Tree,” Advances in Neural Information Processing Systems, vol. 30, 2017.
[Google Scholar] [Publisher Link]

[14] Corinna Cortes, and Vladimir Vapnik, “Support-Vector Networks,” Machine Learning, vol. 20, no. 3, pp. 273-297, 1995.
[CrossRef] [Google Scholar] [Publisher Link]

[15] Peter I. Frazier, “A Tutorial on Bayesian Optimization,” arXiv preprint, pp. 1-22, 2018.
[CrossRef] [Google Scholar] [Publisher Link]

[16] Alaa Tharwat, “Classification Assessment Methods,” Applied Computing and Informatics, vol. 17, no. 1, pp. 168-192, 2021.
[CrossRef] [Google Scholar] [Publisher Link]

[17] Tom Fawcett, “An Introduction to ROC Analysis,” Pattern Recognition Letters, vol. 27, no. 8, pp. 861-874, 2006.
[CrossRef] [Google Scholar] [Publisher Link]

[18] Jingli Tang et al., “Radiation Pneumonitis Prediction using Dual-Modal Data Fusion based on Med3D Transfer Network,” Journal of Digital Imaging, vol. 38, no. 4, pp. 2085-2101, 2024.
[CrossRef] [Google Scholar] [Publisher Link]

[19] Kyu Hye Choi et al., “Using Deep Learning to Predict Radiation Pneumonitis in Patients Treated with Stereotactic Body Radiotherapy (SBRT) for Pulmonary Nodules: Preliminary Results,” Journal of the Korean Physical Society, vol. 81, no. 5, pp. 460-470, 2022.
[CrossRef] [Google Scholar] [Publisher Link]

[20] Yan Kong et al., “Enhancing the Prediction of Symptomatic Radiation Pneumonitis for Locally Advanced Non-Small-Cell Lung Cancer by Combining 3D Deep Learning-Derived Imaging Features with Dose–Volume Metrics: A Two-Center Study,” Radiation Oncology and Oncology, vol. 201, no. 3, pp. 274-282, 2024.
[CrossRef] [Google Scholar] [Publisher Link]

[21] Jang Hyung Lee et al., “Deep-Learning Model Prediction of Radiation Pneumonitis using Pretreatment Chest CT and Clinical Factors,” Technology in Cancer Research and Treatment, vol. 23, pp. 1-11, 2024.
[CrossRef] [Google Scholar] [Publisher Link]

[22] Liming Sheng et al., “Radiation Pneumonia Predictive Model for Radiotherapy in Oesophageal Carcinoma Patients,” BMC Cancer, vol. 23, no. 1, pp. 1-11, 2023.
[CrossRef] [Google Scholar] [Publisher Link]

[23] Rutger H. Stoffers et al., “Rising Incidence of Radiation Pneumonitis After Adjuvant Durvalumab in NSCLC Patients Treated with Concurrent Chemoradiotherapy,” Acta Oncologica, vol. 64, pp. 267-275, 2025.
[CrossRef] [Google Scholar] [Publisher Link]

[24] M. Athiyaman et al., “A Physicist Description of Indigenous Telecobalt Machine Bhabhatron-II TAW,” Global Journal of Science Frontier Research, vol. 15, no. 3, pp. 105-110, 2015.
[Google Scholar]

[25] National Cancer Institute, Common Terminology Criteria for Adverse Events (CTCAE) Version 5.0, U.S. Department of Health and Human Services, NIH, NCI, 2017. [Online]. Available: https://dctd.cancer.gov/research/ctep-trials/for-sites/adverse-events/ctcae-v5-5x7.pdf