Machine learning‑based multiparametric traditional multislice computed tomography radiomics for improving the discrimination of parotid neoplasms

Xu,Zhifeng; Jin,Yabin; Wu,Wenxiu; Wu,Jinmian; Luo,Bing; Zeng,Chenglong; Guo,Xiuqin; Gao,Mingcong; Guo,Shiqin; Pan,Aizhen

doi:10.3892/mco.2021.2407

Machine learning‑based multiparametric traditional multislice computed tomography radiomics for improving the discrimination of parotid neoplasms

Authors:
- Zhifeng Xu
- Yabin Jin
- Wenxiu Wu
- Jinmian Wu
- Bing Luo
- Chenglong Zeng
- Xiuqin Guo
- Mingcong Gao
- Shiqin Guo
- Aizhen Pan
View Affiliations

Affiliations: Department of Radiology, The First People's Hospital of Foshan, Foshan, Guangdong 528000, P.R. China, Clinical Research Institute, The First People's Hospital of Foshan, Foshan, Guangdong 528000, P.R. China, Department of Radiology, The Sanshui People's Hospital, Foshan, Guangdong 528100, P.R. China, Department of Radiology, The Foshan Chancheng District Central Hospital, Foshan, Guangdong 528000, P.R. China, Department of Radiology, The Foshan Medicine Hospital, Guangdong 528000, P.R. China, Department of Radiology, The Second People's Hospital of Nanhai District, Foshan, Guangdong 528000, P.R. China
Published online on: September 24, 2021 https://doi.org/10.3892/mco.2021.2407
Article Number: 245

Metrics: Total Views: 0 (Spandidos Publications: | PMC Statistics: )

Metrics: Total PDF Downloads: 0 (Spandidos Publications: | PMC Statistics: )

Cited By (CrossRef): 0 citations Loading Articles...

Abstract

Characterization of parotid tumors is important for treatment planning and prognosis, and parotid tumor discrimination has recently been developed at the molecular level. The aim of the present study was to establish a machine learning (ML) predictive model based on multiparametric traditional multislice CT (MSCT) radiomic and clinical data analysis to improve the accuracy of differentiation among pleomorphic adenoma (PA), Warthin tumor (WT) and parotid carcinoma (PCa). A total of 345 patients (200 with WT, 91 with PA and 54 with PCa) with pathologically confirmed parotid tumors were retrospectively enrolled from five independent institutions between January 2010 and May 2019. A total of 273 patients recruited from institutions 1, 2 and 3 were randomly assigned to the training model; the independent validation set consisted of 72 patients treated at institutions 1, 4 and 5. Data were investigated using a linear discriminant analysis‑based ML classifier. Feature selection and dimension reduction were conducted using reproducibility testing and a wrapper method. The diagnostic accuracy of the predictive model was compared with histopathological findings as reference results. This classifier achieved a satisfactory performance for the discrimination of PA, WT and PCa, with a total accuracy of 82.1% in the training cohort and 80.5% in the validation cohort. In conclusion, ML‑based multiparametric traditional MSCT radiomics can improve the accuracy of differentiation among PA, WT and PCa. The findings of the present study should be validated by multicenter prospective studies using completely independent external data.

View Figures

View References

1	Eissa L, Seif SA, Desooky SE, Eid M and Koraitim T: Accuracy assessment of combined diffusion weighed and dynamic gadolinium MR sequences in characterization of salivary gland tumors. Egypt J Radiol Nucl Med. 47:127–139. 2016.
2	Razek AAK, Samir S and Ashmalla GA: Characterization of parotid tumors with dynamic susceptibility contrast perfusion weighted magnetic resonance imaging and diffusion-weighted MR imaging. J Comput Assist Tomogr. 41:131–136. 2017.PubMed/NCBI View Article : Google Scholar
3	Xu Z, Zheng S, Pan A, Cheng X and Gao M: A multiparametric analysis based on DCE-MRI to improve the accuracy of parotid tumor discrimination. Eur J Nucl Med Mol Imaging. 46:2228–22344. 2019.PubMed/NCBI View Article : Google Scholar
4	Razek AA: Characterization of salivary gland tumours with diffusion tensor imaging. Dentomaxillofac Radiol. 47(20170343)2018.PubMed/NCBI View Article : Google Scholar
5	Xu ZF, Yong F, Yu T, Chen YY, Gao MY, Zhou T, Pan AZ and Wu RH: Different histological subtypes of parotid gland tumors: CT findings and diagnostic strategy. World J Radiol. 5:313–320. 2013.PubMed/NCBI View Article : Google Scholar
6	Fan Y, Chen C, Zhao F, Tian Z, Wang J, Ma X and Xu J: Radiomics-based machine learning technology enables better differentiation between glioblastoma and anaplastic oligodendroglioma. Front Oncol. 9(1164)2019.PubMed/NCBI View Article : Google Scholar
7	Hua X, Huang X, Huang ZZ, Song CG, Deng JP, Long ZQ, Lin HX and Yuan ZY: Establishment of prognostic nomograms based on skeletal muscle index and serum biomarker in breast cancer patients receiving radiotherapy. Clin Transl Med. 10(e115)2020.PubMed/NCBI View Article : Google Scholar
8	Jin GQ, Su DK, Xie D, Zhao W, Liu LD and Zhu XN: Distinguishing benign from malignant parotid gland tumours: Low-dose multiphasic CT protocol with 5-minute delay. Eur Radiol. 21:1692–1698. 2011.PubMed/NCBI View Article : Google Scholar
9	Xu Z, Rong F, Yu T, Chen YY, Gao Q, Zhou T and Pan AZ: Pleomorphic adenoma versus Warthin tumor of the parotid gland: Diagnostic value of CT perfusion imaging and its pathologic explanation:CT Perfusion of pleomorphic adenoma versus Warthin. J Tumor. 4:419–425. 2016.
10	Yuan Y, Tang W and Tao X: Parotid gland lesions: Separate and combined diagnostic value of conventional MRI, diffusion-weighted imaging, and dynamic contrast-enhanced MRI. Br J Radiol. 89(20150912)2016.PubMed/NCBI View Article : Google Scholar
11	Matsusue E, Fujihara Y, Matsuda E, Tokuyasu Y, Nakamoto S, Nakamura K and Ogawa T: Vanishing parotid tumors on MR imaging. Yonago Acta Med. 61:33–39. 2018.PubMed/NCBI View Article : Google Scholar
12	Kato H, Kanematsu M, Watanabe H, Mizuta K and Aoki M: Salivary gland tumors of the parotid gland: CT and MR imaging findings with emphasis on intratumoral cystic components. Neuroradiology. 56:789–795. 2014.PubMed/NCBI View Article : Google Scholar
13	Ji GW, Zhu FP, Xu Q, Wang K, Wu MY, Tang WW, Li XC and Wang XH: Machine-learning analysis of contrast-enhanced CT radiomics predicts recurrence of hepatocellular carcinoma after resection: A multi-institutional study. EBioMedicine. 50:156–165. 2019.PubMed/NCBI View Article : Google Scholar
14	Colakoglu B, Alis D and Yergin M: Diagnostic value of machine learning-based quantitative texture analysis in differentiating benign and malignant thyroid nodules. J Oncol. 31(6328329)2019.PubMed/NCBI View Article : Google Scholar
15	Ding J, Xing Z, Jiang Z, Chen J, Pan L, Qiu J and Xing W: CT-based radiomic model predicts high grade of clear cell renal cell carcinoma. Eur J Radiol. 103:51–56. 2018.PubMed/NCBI View Article : Google Scholar
16	Gessert N, Bengs M, Wittig L, Drömann D, Keck T, Schlaefer A and Ellebrecht DB: Deep transfer learning methods for colon cancer classification in confocal laser microscopy images Int J Comput Assist Radiol. Surg. 14:1837–1845. 2019.PubMed/NCBI View Article : Google Scholar
17	van Griethuysen JJ, Fedorov A, Parmar C, Hosny A, Aucoin N, Narayan V, Beets-Tan RG, Fillion-Robin JC, Pieper S and Aerts HJ: Computational radiomics system to decode the radiographic phenotype. Cancer Res. 77:e104–e107. 2017.PubMed/NCBI View Article : Google Scholar
18	Bellingegni AD, Gruppioni E, Colazzo G, Davalli A, Sacchetti R, Guglielmelli E and Zollo L: NLR, MLP, SVM, and LDA: A comparative analysis on EMG data from people with trans-radial amputation. J Neuroeng Rehabil. 15(82)2017.PubMed/NCBI View Article : Google Scholar
19	Xiao qing D and Hong cai C: Feature extraction using a fusion method based on sub-pattern row-column two-dimensional linear discriminant analysis. Journal of Computer Applications. 12:3593–3598. 2014.
20	Zhu Y and Ding Y: Auto-optimized paralleled sinogram noise reduction method based on relative quality assessment for low-dose X-ray computed tomography. J Med Imaging Health Informatics. 7:278–282. 2017.