Machine Learning-Based Predictor for Treatment Outcomes of Patients With Salivary Gland Cancer After Operation
0
조회
0
다운로드
0
인용
0
참고문헌
키워드
초록
Background and Objectives The purpose of this study was to analyze the survival data of salivary gland cancer (SGCs) patients to construct machine learning and deep learning mod- els that can predict survival and use them to stratify SGC patients according to risk estimate. Subjects and Method We retrospectively analyzed the clinicopathologic data from 460 pa- tients with SGCs from 2006 to 2018. Results In Cox proportional hazard (CPH) model, pM, stage, lymphovascular invasion, lymph node ratio, and age exhibited significant correlation with patient’s survival. In the CPH model, the c-index value for the training set was 0.85, and that for the test set was 0.81. In the Random Survival Forest model, the c-index value for the training set was 0.86, and that for the test set was 0.82. Stage and age exhibited high importance in both the Random Survival For- est and CPH models. In the deep learning-based model, the c-index value was 0.72 for the training set and 0.72 for the test set. Among the three models mentioned above, the Random Survival Forest model exhibited the highest performance in predicting the survival of SGC patients. Conclusion A survival prediction model using machine learning techniques showed accept- able performance in predicting the survival of SGC patients. Although large-scale clinical and multicenter studies should be conducted to establish more powerful predictive model, we ex- pect that individualized treatment can be realized according to risk stratification made by the machine learning model.
Abstract
Background and Objectives The purpose of this study was to analyze the survival data of salivary gland cancer (SGCs) patients to construct machine learning and deep learning mod- els that can predict survival and use them to stratify SGC patients according to risk estimate. Subjects and Method We retrospectively analyzed the clinicopathologic data from 460 pa- tients with SGCs from 2006 to 2018. Results In Cox proportional hazard (CPH) model, pM, stage, lymphovascular invasion, lymph node ratio, and age exhibited significant correlation with patient’s survival. In the CPH model, the c-index value for the training set was 0.85, and that for the test set was 0.81. In the Random Survival Forest model, the c-index value for the training set was 0.86, and that for the test set was 0.82. Stage and age exhibited high importance in both the Random Survival For- est and CPH models. In the deep learning-based model, the c-index value was 0.72 for the training set and 0.72 for the test set. Among the three models mentioned above, the Random Survival Forest model exhibited the highest performance in predicting the survival of SGC patients. Conclusion A survival prediction model using machine learning techniques showed accept- able performance in predicting the survival of SGC patients. Although large-scale clinical and multicenter studies should be conducted to establish more powerful predictive model, we ex- pect that individualized treatment can be realized according to risk stratification made by the machine learning model.
참고문헌 (19)
- Lewis AG, "Diagnosis and management of malignant salivary gland tumors of the parotid gland", Otolaryngol Clin North Am, 49(2), pp.343~380 (2016)
- Ali S, "Postoperative nomograms predictive of survival after surgical management of malignant tumors of the major salivary glands", Ann Surg Oncol, 21(2), pp.637~642 (2014)
- Ali S, "Distant metastases in patients with carcinoma of the major salivary glands", Ann Surg Oncol, 22(12), pp.4014~4019 (2015)
- Ettl T, "Predictors of cervical lymph node metastasis in salivary gland cancer", Head Neck, 36(4), pp.517~523 (2014)
- Saintigny P, "Gene expression profiling predicts the development of oral cancer", Cancer Prev Res(Phila), 4(2), pp.218~229 (2011)
- Kann BH, "Pretreatment identification of head and neck cancer nodal metastasis and extranodal extension using deep learning neural networks", Sci Rep, 8(1), pp.14036~ (2018)
- Harrell FE Jr, "Evaluating the yield of medical tests", JAMA, 247(18), pp.2543~2546 (1982)
- Katzman JL, "DeepSurv : Personalized treatment recommender system using a Cox proportional hazards deep neural network", BMC Med Res Methodol, 18(1), pp.24~ (2018)
- Ishwaran H, "Random survival forests for competing risks", Biostatistics, 15(4), pp.757~773 (2014)
- Su X, "Interaction trees with censored survival data", Int J Biostat, 4(1) (2008)
- Lombardi D, "Surgical treatment of salivary malignant tumors", Oral Oncol, 65, pp.102~113 (2017)
- Godballe C, "Parotid carcinoma : Impact of clinical factors on prognosis in a histologically revised series", Laryngoscop, 113(8), pp.1411~1417 (2003)
- Vander Poorten VL, "The development of a prognostic score for patients with parotid carcinoma", Cancer, 85(9), pp.2057~2067 (1999)
- Hay A, "Validation of nomograms for overall survival, cancer-specific survival, and recurrence in carcinoma of the major salivary glands", Head Neck, 40(5), pp.1008~1015 (2018)
- Mannelli G, "Nomograms predictive for oncological outcomes in malignant parotid tumours : Recurrence and mortality rates of 228patients from a single institution", Eur Arch Otorhinolaryngol (2019)
- Balachandran VP, "Nomograms in oncology : More than meets the eye", Lancet Oncol, 16(4), pp.e173~e180 (2015)
- Chen F, "MRI-based radiomics of rectal cancer : Assessment of the local recurrence at the site of anastomosis", Acad Radiol, 28(Suppl 1), pp.S87~S94 (2021)
- Zhang H, "Prediction of cervical lymph node metastasis using MRI radiomics approach in papillary thyroid carcinoma : A feasibility study", Technol Cancer Res Treat, 19, pp.1533033820969451~ (2020)
- Zhang M, "Performance of 18F-FDG PET", CT radiomics for predicting EGFR mutation status in patients with non-small cell lung cancer, 2020, pp.10, 568857~