Skip to content
DraconiticResearch
Search Sources API
223,238 studies31 sourcesUpdated 41 min ago
Peer-reviewedCC BY-NC-ND 4.0Radiation Oncology (London, England)

Combining handcrafted radiomics features and deep learning features within different regions of interest in planning CT images for the prediction of symptomatic radiation pneumonitis in locally advanced non-small cell lung cancer

  • Xuan Li1,2
  • Hao Hang2,3
  • Yan Kong2
  • Jinmeng Zhang2,4
  • Mingming Su1,2
  • Zongqiong Sun5
  • Yan Zhu2
  • Fei Yang6
  • Qing Zhao7
  • Jianjiao Ni4
  • Jianfeng Huang2
7 affiliations
  1. 1Department of Medical Oncology, Wuxi Huishan District People’ s Hospital, Wuxi, Jiangsu 214187 China
  2. 2Department of Radiation Oncology, Affiliated Hospital of Jiangnan University, Wuxi, Jiangsu 214122 China
  3. 3Wuxi School of Medicine, Jiangnan University, Wuxi, Jiangsu 214122 China
  4. 4Department of Radiation Oncology, Fudan University Shanghai Cancer Center, Shanghai, 200032 China
  5. 5Department of Radiology, Affiliated Hospital of Jiangnan University, Wuxi, Jiangsu 214062 China
  6. 6Department of Radiation Oncology, University of Miami, Miami, FL 33136 USA
  7. 7Department of Pharmacy, Affiliated Hospital of Jiangnan University, Wuxi, Jiangsu 214122 China
Read full textDOICite

Abstract

Purpose: To predict symptomatic radiation pneumonitis (RP) in locally advanced non-small cell lung cancer (LA-NSCLC) patients by handcrafted radiomics (HCR) and deep learning (DL) features of three different regions of interest (ROIs) in normal lung tissue.

Methods: Ninety LA-NSCLC patients from Center A were collected as the development cohort, and an additional 57 patients from Center B were collected as the external validation set. Clinical data and lung dose-volume parameters were collected from the patients, and univariate analysis was conducted. Three ROIs in lung were delineated in pre-radiotherapy CT images: Lung-PTV, Lung-GTV, and PTV-GTV. For each ROI, 1046 HCR features and 512 DL features were respectively extracted, and then correlated. For each ROI, three models (HCR, DLR, and a hybrid model) were developed with feature selection via Spearman correlation, mRMR, the LASSO method, and SVM-based modeling. And finally, their performance was evaluated and compared using AUC values from ROC curves. The optimal model was selected by comparing the best AUC through internal validation results, and then external validation was conducted. Calibration and decision curve analyses were used to evaluate the model.

Results: In the development cohort, univariate analysis revealed no statistically significant associations with RP. The correlation analysis revealed a weak correlation between HCR and DL features. In internal validation, for Lung-PTV, Lung-GTV and PTV-GTV, HCR models achieved AUCs of 0.658 (95%CI: 0.645–0.671), 0.659 (95%CI: 0.646–0.672), and 0.890 (95%CI: 0.881–0.899); DLR models obtained AUCs of 0.832 (95%CI: 0.824–0.840), 0.900 (95%CI: 0.893–0.907), and 0.777 (95%CI: 0.766–0.788); and the hybrid model showed AUCs of 0.838 (95%CI: 0.831–0.846), 0.901 (95%CI: 0.894–0.908), and 0.919 (95%CI: 0.912–0.926). The results showed that the DLR model outperformed the HCR model for Lung-PTV and Lung-GTV, and the hybrid model surpassed both for all three ROIs. The hybrid model of PTV-GTV was found to be the best, and its external validation was conducted. The results showed that the AUC for the external validation set was 0.828 (95%CI: 0.696–0.960). The model was well calibrated (p > 0.05), and decision curve analysis indicated positive net benefit across relevant threshold probabilities.

Conclusions: Hybrid models combining HCR and DL features showed the highest predictive performance in internal comparisons and the best-performing PTV-GTV hybrid model was externally validated, highlighting the potential of combined features for improved prediction.

Related

  1. Development and internal validation of CTA-based radiomics models for differentiating symptomatic from asymptomatic unruptured spontaneous intracranial vertebrobasilar artery dissection
    Jia Jiang, Sheng Bi, Siming Gui +3
    Frontiers in NeurologyPeer-reviewedCC BY 4.0Full textMedical imaging & radiologyCrossrefDOAJEurope PMCPMC
  2. An explainable machine learning model combining intratumoral and peritumoral computed tomography radiomics with maximum standard uptake value for predicting International Association for the Study of Lung Cancer grading and prognosis in stage I pure-solid lung adenocarcinoma.
    Hanzhe Wang, Junping Lan, Yimin Chen +11
    Journal of thoracic diseasePeer-reviewedCC BY-NC-NDFull textMedical imaging & radiologyEurope PMCPMC
  3. CT-based multi-station lymph-node radiomics for preoperative prediction of lymph node metastasis in colorectal cancer: a proof-of-concept comparison with primary-tumor radiomics
    Zejin Lin, Jue Wang, Hairong Chen +6
    European radiology experimentalPeer-reviewedCC BYFull textMedical imaging & radiologyEurope PMC
  4. Habitat-based ultrasound radiomics for preoperative prediction of axillary lymph node metastasis in breast cancer: a two-center study
    Zimeng Fang, Xiaoyu Yan, Yunyun Zhan +6
    Gland SurgeryPeer-reviewedCC BY-NC-NDFull textGeneral medicineEurope PMC
  5. Nomogram models integrating multiregion radiomics, quantitative computed tomography, and clinical indicators for distinguishing connective tissue disease-associated interstitial lung disease from idiopathic pulmonary fibrosis and improving the prediction of gender-age-physiology stage
    Quantitative imaging in medicine and surgeryPeer-reviewedCC BY-NC-NDFull textMedical imaging & radiologyEurope PMC

Cite

BibTeXRIS
BibTeX
@article{li2026combining,
  title = {{Combining handcrafted radiomics features and deep learning features within different regions of interest in planning CT images for the prediction of symptomatic radiation pneumonitis in locally advanced non-small cell lung cancer}},
  author = {Xuan Li and Hao Hang and Yan Kong and Jinmeng Zhang and Mingming Su and Zongqiong Sun and Yan Zhu and Fei Yang and Qing Zhao and Jianjiao Ni and Jianfeng Huang},
  journal = {Radiation Oncology (London, England)},
  year = {2026},
  month = jul,
  doi = {10.1186/s13014-026-02896-9},
  url = {https://doi.org/10.1186/s13014-026-02896-9},
  note = {CC BY-NC-ND 4.0},
}
RIS
TY  - JOUR
TI  - Combining handcrafted radiomics features and deep learning features within different regions of interest in planning CT images for the prediction of symptomatic radiation pneumonitis in locally advanced non-small cell lung cancer
AU  - Xuan Li
AU  - Hao Hang
AU  - Yan Kong
AU  - Jinmeng Zhang
AU  - Mingming Su
AU  - Zongqiong Sun
AU  - Yan Zhu
AU  - Fei Yang
AU  - Qing Zhao
AU  - Jianjiao Ni
AU  - Jianfeng Huang
PY  - 2026
DA  - 2026/07/31
JO  - Radiation Oncology (London, England)
DO  - 10.1186/s13014-026-02896-9
UR  - https://doi.org/10.1186/s13014-026-02896-9
AB  - Purpose: To predict symptomatic radiation pneumonitis (RP) in locally advanced non-small cell lung cancer (LA-NSCLC) patients by handcrafted radiomics (HCR) and deep learning (DL) features of three different regions of interest (ROIs) in normal lung tissue.  Methods: Ninety LA-NSCLC patients from Center A were collected as the development cohort, and an additional 57 patients from Center B were collected as the external validation set. Clinical data and lung dose-volume parameters were collected from the patients, and univariate analysis was conducted. Three ROIs in lung were delineated in pre-radiotherapy CT images: Lung-PTV, Lung-GTV, and PTV-GTV. For each ROI, 1046 HCR features and 512 DL features were respectively extracted, and then correlated. For each ROI, three models (HCR, DLR, and a hybrid model) were developed with feature selection via Spearman correlation, mRMR, the LASSO method, and SVM-based modeling. And finally, their performance was evaluated and compared using AUC values from ROC curves. The optimal model was selected by comparing the best AUC through internal validation results, and then external validation was conducted. Calibration and decision curve analyses were used to evaluate the model.  Results: In the development cohort, univariate analysis revealed no statistically significant associations with RP. The correlation analysis revealed a weak correlation between HCR and DL features. In internal validation, for Lung-PTV, Lung-GTV and PTV-GTV, HCR models achieved AUCs of 0.658 (95%CI: 0.645–0.671), 0.659 (95%CI: 0.646–0.672), and 0.890 (95%CI: 0.881–0.899); DLR models obtained AUCs of 0.832 (95%CI: 0.824–0.840), 0.900 (95%CI: 0.893–0.907), and 0.777 (95%CI: 0.766–0.788); and the hybrid model showed AUCs of 0.838 (95%CI: 0.831–0.846), 0.901 (95%CI: 0.894–0.908), and 0.919 (95%CI: 0.912–0.926). The results showed that the DLR model outperformed the HCR model for Lung-PTV and Lung-GTV, and the hybrid model surpassed both for all three ROIs. The hybrid model of PTV-GTV was found to be the best, and its external validation was conducted. The results showed that the AUC for the external validation set was 0.828 (95%CI: 0.696–0.960). The model was well calibrated (p > 0.05), and decision curve analysis indicated positive net benefit across relevant threshold probabilities.  Conclusions: Hybrid models combining HCR and DL features showed the highest predictive performance in internal comparisons and the best-performing PTV-GTV hybrid model was externally validated, highlighting the potential of combined features for improved prediction.
SN  - 1748-717X
ER  -