Machine Learning and Deep Learning Hybrid Approach Based on Muscle Imaging Features for Diagnosis of Esophageal Cancer
Abstract
1. Introduction
2. Methods and Materials
2.1. Clinical Cohort Establishment
2.2. Imaging Data Collection and Processing
2.3. Radiomics Workflow
2.4. Statistical Analysis and Machine Learning Tools
3. Results
3.1. Clinical Cohort Characteristics
3.2. Differential Diagnostic Models for Squamous Cell Carcinoma and Adenocarcinoma
3.3. Construction of T-Stage Diagnostic Models Based on Radiomics
3.4. Construction of N-Stage Diagnostic Models Based on Radiomics
4. Discussion
5. Conclusions
Supplementary Materials
Author Contributions
Funding
Institutional Review Board Statement
Informed Consent Statement
Data Availability Statement
Conflicts of Interest
Abbreviations
EC | Esophageal cancer |
AI | Artificial intelligence |
CT | Computed tomography |
SVM | Support vector machine |
MLP | Multilayer perceptron |
BMI | Body mass index |
ROC | Receiver Operating Characteristic Curve |
AUC | Area under the curve |
DCA | Decision curve analysis |
CNN | Convolutional Neural Network |
ROI | Region of interest |
ResNet | Residual Network |
DenseNet | Densely Connected Convolutional Network |
VGG | Visual Geometry Group |
RF | Random Forest |
KNN | K-Nearest Neighbors |
XGBoost | eXtreme Gradient Boosting |
LightGBM | Light Gradient Boosting Machine |
PCA | Principal Component Analysis |
LR | Logistic regression |
References
- Bray, F.; Laversanne, M.; Sung, H.; Ferlay, J.; Siegel, R.L.; Soerjomataram, I.; Jemal, A. Global cancer statistics 2022: GLOBOCAN estimates of incidence and mortality worldwide for 36 cancers in 185 countries. CA A Cancer J. Clin. 2024, 74, 229–263. [Google Scholar] [CrossRef]
- An, L.; Zheng, R.; Zeng, H.; Zhang, S.; Chen, R.; Wang, S.; Sun, K.; Li, L.; Wei, W.; He, J. The survival of esophageal cancer by subtype in China with comparison to the United States. Int. J. Cancer 2022, 152, 151–161. [Google Scholar] [CrossRef]
- Deboever, N.; Jones, C.M.; Yamashita, K.; Ajani, J.A.; Hofstetter, W.L. Advances in diagnosis and management of cancer of the esophagus. BMJ 2024, 385, e074962. [Google Scholar] [CrossRef]
- Zhu, H.; Ma, X.; Ye, T.; Wang, H.; Wang, Z.; Liu, Q.; Zhao, K. Esophageal cancer in China: Practice and research in the new era. Int. J. Cancer 2023, 152, 1741–1751. [Google Scholar] [CrossRef]
- Lambin, P.; Rios-Velazquez, E.; Leijenaar, R.; Carvalho, S.; van Stiphout, R.G.P.M.; Granton, P.; Zegers, C.M.L.; Gillies, R.; Boellard, R.; Dekker, A.; et al. Radiomics: Extracting more information from medical images using advanced feature analysis. Eur. J. Cancer 2012, 48, 441–446. [Google Scholar] [CrossRef]
- Aerts, H.J.W.L.; Velazquez, E.R.; Leijenaar, R.T.H.; Parmar, C.; Grossmann, P.; Carvalho, S.; Bussink, J.; Monshouwer, R.; Haibe-Kains, B.; Rietveld, D.; et al. Decoding tumour phenotype by noninvasive imaging using a quantitative radiomics approach. Nat. Commun. 2014, 5, 4006. [Google Scholar] [CrossRef]
- Parmar, C.; Leijenaar, R.T.; Grossmann, P.; Rios Velazquez, E.; Bussink, J.; Rietveld, D.; Rietbergen, M.M.; Haibe-Kains, B.; Lambin, P.; Aerts, H.J. Radiomic feature clusters and prognostic signatures specific for Lung and Head & Neck cancer. Sci. Rep. 2015, 5, 11044. [Google Scholar] [CrossRef]
- Fu, N.; Fu, W.; Chen, H.; Chai, W.; Qian, X.; Wang, W.; Jiang, Y.; Shen, B. A deep-learning radiomics-based lymph node metastasis predictive model for pancreatic cancer: A diagnostic study. Int. J. Surg. 2023, 109, 2196–2203. [Google Scholar] [CrossRef]
- Arthur, A.; Orton, M.R.; Emsley, R.; Vit, S.; Kelly-Morland, C.; Strauss, D.; Lunn, J.; Doran, S.; Lmalem, H.; Nzokirantevye, A.; et al. A CT-based radiomics classification model for the prediction of histological type and tumour grade in retroperitoneal sarcoma (RADSARC-R): A retrospective multicohort analysis. Lancet Oncol. 2023, 24, 1277–1286. [Google Scholar] [CrossRef]
- Du, K.P.; Huang, W.P.; Liu, S.Y.; Chen, Y.J.; Li, L.M.; Liu, X.N.; Han, Y.J.; Zhou, Y.; Liu, C.C.; Gao, J.B. Application of computed tomography-based radiomics in differential diagnosis of adenocarcinoma and squamous cell carcinoma at the esophagogastric junction. World J. Gastroenterol. 2022, 28, 4363–4375. [Google Scholar] [CrossRef]
- Yuan, P.; Huang, Z.-H.; Yang, Y.-H.; Bao, F.-C.; Sun, K.; Chao, F.-F.; Liu, T.-T.; Zhang, J.-J.; Xu, J.-M.; Li, X.-N.; et al. A 18F-FDG PET/CT-based deep learning-radiomics-clinical model for prediction of cervical lymph node metastasis in esophageal squamous cell carcinoma. Cancer Imaging 2024, 24, 153. [Google Scholar] [CrossRef]
- Prado, C.M.; Lieffers, J.R.; McCargar, L.J.; Reiman, T.; Sawyer, M.B.; Martin, L.; Baracos, V.E. Prevalence and clinical implications of sarcopenic obesity in patients with solid tumours of the respiratory and gastrointestinal tracts: A population-based study. Lancet Oncol. 2008, 9, 629–635. [Google Scholar] [CrossRef]
- Martin, L.; Birdsell, L.; Macdonald, N.; Reiman, T.; Clandinin, M.T.; McCargar, L.J.; Murphy, R.; Ghosh, S.; Sawyer, M.B.; Baracos, V.E. Cancer cachexia in the age of obesity: Skeletal muscle depletion is a powerful prognostic factor, independent of body mass index. J. Clin. Oncol. 2013, 31, 1539–1547. [Google Scholar] [CrossRef]
- Mourtzakis, M.; Prado, C.M.; Lieffers, J.R.; Reiman, T.; McCargar, L.J.; Baracos, V.E. A practical and precise approach to quantification of body composition in cancer patients using computed tomography images acquired during routine care. Appl. Physiol. Nutr. Metab. 2008, 33, 997–1006. [Google Scholar] [CrossRef]
- Zhou, Y.; Zhou, J.; Cai, X.; Ge, S.; Sang, S.; Yang, Y.; Zhang, B.; Deng, S. Integrating (18)F-FDG PET/CT radiomics and body composition for enhanced prognostic assessment in patients with esophageal cancer. BMC Cancer 2024, 24, 1402. [Google Scholar] [CrossRef]
- Gu, L.; Cai, X.C. Fusing 2D and 3D convolutional neural networks for the segmentation of aorta and coronary arteries from CT images. Artif. Intell. Med. 2021, 121, 102189. [Google Scholar] [CrossRef]
- Granata, V.; Fusco, R.; Brunese, M.C.; Di Mauro, A.; Avallone, A.; Ottaiano, A.; Izzo, F.; Normanno, N.; Petrillo, A. Machine learning-based radiomics analysis in predicting RAS mutational status using magnetic resonance imaging. Radiol. Medica 2024, 129, 420–428. [Google Scholar] [CrossRef]
- Van Griethuysen, J.J.M.; Fedorov, A.; Parmar, C.; Hosny, A.; Aucoin, N.; Narayan, V.; Beets-Tan, R.G.H.; Fillion-Robin, J.C.; Pieper, S.; Aerts, H. Computational Radiomics System to Decode the Radiographic Phenotype. Cancer Res. 2017, 77, e104–e107. [Google Scholar] [CrossRef]
- Yun, K.; He, T.; Zhen, S.; Quan, M.; Yang, X.; Man, D.; Zhang, S.; Wang, W.; Han, X. Development and validation of explainable machine-learning models for carotid atherosclerosis early screening. J. Transl. Med. 2023, 21, 353. [Google Scholar] [CrossRef]
- Boldini, D.; Grisoni, F.; Kuhn, D.; Friedrich, L.; Sieber, S.A. Practical guidelines for the use of gradient boosting for molecular property prediction. J. Cheminform. 2023, 15, 73. [Google Scholar] [CrossRef]
- Yang, Y.; Cheng, J.; Peng, Z.; Yi, L.; Lin, Z.; He, A.; Jin, M.; Cui, C.; Liu, Y.; Zhong, Q.; et al. Development and Validation of Contrast-Enhanced CT-Based Deep Transfer Learning and Combined Clinical-Radiomics Model to Discriminate Thymomas and Thymic Cysts: A Multicenter Study. Acad. Radiol. 2024, 31, 1615–1628. [Google Scholar] [CrossRef]
- Zahari, R.; Cox, J.; Obara, B. Uncertainty-aware image classification on 3D CT lung. Comput. Biol. Med. 2024, 172, 108324. [Google Scholar] [CrossRef]
- Sun, T.; Ding, S.; Guo, L. Low-degree term first in ResNet, its variants and the whole neural network family. Neural Netw. 2022, 148, 155–165. [Google Scholar] [CrossRef]
- Chen, P.; Gao, L.; Shi, X.; Allen, K.; Yang, L. Fully automatic knee osteoarthritis severity grading using deep neural networks with a novel ordinal loss. Comput. Med. Imaging Graph. 2019, 75, 84–92. [Google Scholar] [CrossRef]
- Guo, Y.; Zheng, Y.; Tan, M.; Chen, Q.; Li, Z.; Chen, J.; Zhao, P.; Huang, J. Towards Accurate and Compact Architectures via Neural Architecture Transformer. IEEE Trans. Pattern Anal. Mach. Intell. 2022, 44, 6501–6516. [Google Scholar] [CrossRef]
- Larue, R.; Klaassen, R.; Jochems, A.; Leijenaar, R.T.H.; Hulshof, M.; van Berge Henegouwen, M.I.; Schreurs, W.M.J.; Sosef, M.N.; van Elmpt, W.; van Laarhoven, H.W.M.; et al. Pre-treatment CT radiomics to predict 3-year overall survival following chemoradiotherapy of esophageal cancer. Acta Oncol. 2018, 57, 1475–1481. [Google Scholar] [CrossRef]
- Lei, X.; Cao, Z.; Wu, Y.; Lin, J.; Zhang, Z.; Jin, J.; Ai, Y.; Zhang, J.; Du, D.; Tian, Z.; et al. Preoperative prediction of clinical and pathological stages for patients with esophageal cancer using PET/CT radiomics. Insights Imaging 2023, 14, 174. [Google Scholar] [CrossRef]
- Yang, M.; Hu, P.; Li, M.; Ding, R.; Wang, Y.; Pan, S.; Kang, M.; Kong, W.; Du, D.; Wang, F. Computed Tomography-Based Radiomics in Predicting T Stage and Length of Esophageal Squamous Cell Carcinoma. Front. Oncol. 2021, 11, 722961. [Google Scholar] [CrossRef]
- Jannatdoust, P.; Valizadeh, P.; Pahlevan-Fallahy, M.T.; Hassankhani, A.; Amoukhteh, M.; Behrouzieh, S.; Ghadimi, D.J.; Bilgin, C.; Gholamrezanezhad, A. Diagnostic accuracy of CT-based radiomics and deep learning for predicting lymph node metastasis in esophageal cancer. Clin. Imaging 2024, 113, 110225. [Google Scholar] [CrossRef]
- O’Shea, R.; Withey, S.J.; Owczarczyk, K.; Rookyard, C.; Gossage, J.; Godfrey, E.; Jobling, C.; Parsons, S.L.; Skipworth, R.J.E.; Goh, V.; et al. Multicentre validation of CT grey-level co-occurrence matrix features for overall survival in primary oesophageal adenocarcinoma. Eur. Radiol. 2024, 34, 6919–6928. [Google Scholar] [CrossRef]
- Kocak, B.; Baessler, B.; Bakas, S.; Cuocolo, R.; Fedorov, A.; Maier-Hein, L.; Mercaldo, N.; Müller, H.; Orlhac, F.; Pinto Dos Santos, D.; et al. CheckList for EvaluAtion of Radiomics research (CLEAR): A step-by-step reporting guideline for authors and reviewers endorsed by ESR and EuSoMII. Insights Imaging 2023, 14, 75. [Google Scholar] [CrossRef]
- Kocak, B.; D’Antonoli, T.A.; Mercaldo, N.; Alberich-Bayarri, A.; Baessler, B.; Ambrosini, I.; Andreychenko, A.E.; Bakas, S.; Beets-Tan, R.G.H.; Bressem, K.; et al. METhodological RadiomICs Score (METRICS): A quality scoring tool for radiomics research endorsed by EuSoMII. Insights Imaging 2024, 15, 8. [Google Scholar] [CrossRef]
Machine Learning | 2D Deep Learning | 3D Deep Learning | Algorithm | |
---|---|---|---|---|
Pathological classification | Esophageal features Esophageal plus stomach features Muscle feature | Esophageal_densenet201 Esophageal plus stomach_densenet201 Muscle_densenet201 | - | LR |
T staging | Esophageal features | Esophageal_resnet152 Esophageal plus stomach_densenet201 Muscle_resnet152 | Esophageal_resnet152 Muscle_resnet152 | SVM |
N staging | Esophageal plus stomach features | Esophageal plus stomach_resnet50 Muscle_resnet50 | Esophageal plus stomach_resnet152 | MLP |
Characteristics | All | Mean + SD/n (%) |
---|---|---|
Age | 1066 | 66.58 ± 8.08 |
BMI | 1066 | 22.33 ± 3.08 |
Sex | 1066 | |
Male | 871 (81.71%) | |
Female | 195 (18.29%) | |
Smoking status | 1066 | |
No | 681 (63.88%) | |
Yes | 385 (36.12%) | |
Drinking status | 1066 | |
No | 749 (70.26%) | |
Yes | 317 (29.74%) | |
Pathological classification | 1066 | |
Squamous carcinoma | 956 (89.68%) | |
Adenocarcinoma | 110 (10.32%) | |
T staging | 1066 | |
Tis | 30 (2.81%) | |
T1 | 210 (19.70%) | |
T2 | 260 (24.39%) | |
T3 | 552 (51.78%) | |
T4 | 14 (1.31%) | |
N staging | 1066 | |
N0 | 610 (57.22%) | |
N1 | 263 (24.67%) | |
N2 | 149 (13.98%) | |
N3 | 44 (4.13%) |
Model Classification | Esophageal Phase | Muscle Phase | p-Value |
---|---|---|---|
Training Set | |||
ACC | 0.63 ± 0.21 | 0.61 ± 0.25 | 0.72 |
AUC | 0.78 ± 0.09 | 0.67 ± 0.14 | 0.06 |
Test Set | |||
ACC | 0.63 ± 0.21 | 0.55 ± 0.21 | 0.42 |
AUC | 0.66 ± 0.05 | 0.62 ± 0.12 | 0.84 |
Model Classification | Esophageal Phase | Esophagus Plus Stomach Phase | p-Value |
---|---|---|---|
Training Set | |||
ACC | 0.63 ± 0.21 | 0.59 ± 0.25 | 0.92 |
AUC | 0.78 ± 0.09 | 0.70 ± 0.21 | 0.26 |
Test Set | |||
ACC | 0.63 ± 0.21 | 0.68 ± 0.15 | 0.53 |
AUC | 0.66 ± 0.05 | 0.60 ± 0.10 | 0.12 |
Model Classification | Esophagus Plus Stomach Phase | Muscle Phase | p-Value |
---|---|---|---|
Training Set | |||
ACC | 0.59 ± 0.25 | 0.61 ± 0.25 | 0.89 |
AUC | 0.70 ± 0.21 | 0.67 ± 0.14 | 0.77 |
Test Set | |||
ACC | 0.68 ± 0.15 | 0.55 ± 0.21 | 0.14 |
AUC | 0.60 ± 0.10 | 0.62 ± 0.12 | 0.31 |
Model Classification | Esophageal Phase | Esophagus Plus Stomach Phase | p-Value |
---|---|---|---|
T Staging | |||
Train ACC | 0.71 ± 0.19 | 0.64 ± 0.19 | 0.31 |
Test ACC | 0.56 ± 0.06 | 0.53 ± 0.02 | 0.12 |
AUC | |||
T0 | 0.48 ± 0.19 | 0.45 ± 0.13 | 0.65 |
T1 | 0.78 ± 0.08 | 0.58 ± 0.05 | <0.01 |
T2 | 0.61 ± 0.07 | 0.62 ± 0.05 | 0.79 |
T3 | 0.64 ± 0.03 | 0.63 ± 0.07 | 0.89 |
T4 | 0.32 ± 0.28 | 0.40 ± 0.35 | 0.56 |
Model Classification | Esophageal Phase | Muscle Phase | p-Value |
---|---|---|---|
N Staging | |||
Train ACC | 0.69 ± 0.17 | 0.68 ± 0.17 | 0.7 |
Test ACC | 0.59 ± 0.03 | 0.56 ± 0.04 | 0.53 |
AUC | |||
N0 | 0.66 ± 0.09 | 0.54 ± 0.08 | 0.005 |
N1 | 0.46 ± 0.12 | 0.51 ± 0.07 | 0.231 |
N2 | 0.59 ± 0.12 | 0.63 ± 0.11 | 0.380 |
N3 | 0.51 ± 0.15 | 0.39 ± 0.13 | 0.08 |
Model Classification | Esophageal Phase | Esophagus Plus Stomach Phase | p-Value |
---|---|---|---|
N Staging | |||
Train ACC | 0.69 ± 0.17 | 0.68 ± 0.17 | 0.64 |
Test ACC | 0.59 ± 0.03 | 0.58 ± 0.02 | 0.29 |
AUC | |||
N0 | 0.66 ± 0.09 | 0.59 ± 0.09 | 0.09 |
N1 | 0.46 ± 0.12 | 0.59 ± 0.13 | 0.02 |
N2 | 0.59 ± 0.12 | 0.52 ± 0.11 | 0.20 |
N3 | 0.51 ± 0.15 | 0.31 ± 0.17 | 0.01 |
Model Classification | Esophagus Plus Stomach Phase | Muscle Phase | p-Value |
---|---|---|---|
N Staging | |||
Train ACC | 0.68 ± 0.17 | 0.68 ± 0.17 | 0.96 |
Test ACC | 0.59 ± 0.03 | 0.56 ± 0.04 | 0.09 |
AUC | |||
N0 | 0.59 ± 0.09 | 0.54 ± 0.08 | 0.20 |
N1 | 0.59 ± 0.13 | 0.51 ± 0.07 | 0.10 |
N2 | 0.52 ± 0.11 | 0.63 ± 0.11 | 0.03 |
N3 | 0.31 ± 0.17 | 0.39 ± 0.13 | 0.26 |
Research | Number of Population | Target | AUC | Accuracy | Sensitivity | Specificity |
---|---|---|---|---|---|---|
Du, K.P. et al. [10] | 260 | Pathological Subtype | 0.904 | 0.841 | 0.802 | 0.879 |
Lei, X. et al. [27] | 100 | T-stage | 0.850 | - | - | |
Yang, M. et al. [28] | 116 | T-stage | 0.860 | - | 0.77 | 0.87 |
Jannatdoust, P. et al. [29] | - | N-stage | 0.870 | - | 0.787 | 0.818 |
This study | 1066 | Pathological Subtype | 0.980 | 0.900 | - | - |
T-stage | 0.800 | 0.900 | - | - | ||
N-stage | 0.920 | - | - | - |
Disclaimer/Publisher’s Note: The statements, opinions and data contained in all publications are solely those of the individual author(s) and contributor(s) and not of MDPI and/or the editor(s). MDPI and/or the editor(s) disclaim responsibility for any injury to people or property resulting from any ideas, methods, instructions or products referred to in the content. |
© 2025 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (https://creativecommons.org/licenses/by/4.0/).
Share and Cite
Hong, Y.; Wang, H.; Zhang, Q.; Zhang, P.; Cheng, K.; Cao, G.; Zhang, R.; Chen, B. Machine Learning and Deep Learning Hybrid Approach Based on Muscle Imaging Features for Diagnosis of Esophageal Cancer. Diagnostics 2025, 15, 1730. https://doi.org/10.3390/diagnostics15141730
Hong Y, Wang H, Zhang Q, Zhang P, Cheng K, Cao G, Zhang R, Chen B. Machine Learning and Deep Learning Hybrid Approach Based on Muscle Imaging Features for Diagnosis of Esophageal Cancer. Diagnostics. 2025; 15(14):1730. https://doi.org/10.3390/diagnostics15141730
Chicago/Turabian StyleHong, Yuan, Hanlin Wang, Qi Zhang, Peng Zhang, Kang Cheng, Guodong Cao, Renquan Zhang, and Bo Chen. 2025. "Machine Learning and Deep Learning Hybrid Approach Based on Muscle Imaging Features for Diagnosis of Esophageal Cancer" Diagnostics 15, no. 14: 1730. https://doi.org/10.3390/diagnostics15141730
APA StyleHong, Y., Wang, H., Zhang, Q., Zhang, P., Cheng, K., Cao, G., Zhang, R., & Chen, B. (2025). Machine Learning and Deep Learning Hybrid Approach Based on Muscle Imaging Features for Diagnosis of Esophageal Cancer. Diagnostics, 15(14), 1730. https://doi.org/10.3390/diagnostics15141730