Objective: To investigate the diagnostic value of contrast-enhanced CT texture analysis in differentiating thymoma and high-density thymic cysts. Methods: The complete data of 123 patients with thymic tumors confirmed by surgery and pathology were retrospectively analyzed, including 88 patients with thymoma and 35 patients with high-density thymic cysts. All patients were randomly divided into training group (n=87) and validation group (n=36) according to the ratio of 7∶3. Using 3D Slicer, two radiologists manually delineated the region of interest (ROI) along the edge of the target lesion in plain, arterial, and venous images, and extracted the texture features. Intraclass correlation coefficient (ICC) was used to evaluate the consistency of texture features extracted by 2 physicians. Least absolute shrinkage and selection operator (LASSO) regression was applied for feature dimensionality reduction and selection, followed by logistic regression modeling. A radiomics score (Rad-score) was calculated by weighting selected features based on their coefficients. Receiver operating characteristic (ROC) curve analysis was used to evaluate diagnostic performance, and the area under curve (AUC) was calculated. Finally, decision curve analysis (DCA) was employed to validate clinical net benefits across threshold probabilities. Results: The agreement of the texture features extracted by 2 physicians was good (ICC＞0.75). The AUC for the differential diagnosis of thymoma and high-density thymic cysts in the training set and validation set was 0.87 and 0.90. The AUC of plain, arterial, venous, plain＋arterial, plain＋venous, arterial＋venous, and three-phases were 0.90, 0.91, 0.95, 0.94, 0.98, 0.95 and 0.93 in the training set, while in the validation set, the AUC were 0.94, 0.95, 0.97, 0.94, 0.96, 0.97 and 0.96. The results of the DCA analysis showed that the plain＋venous phase model had a good clinical benefit. Conclusion: Based on the combined model of plain＋venous stage in contrast-enhanced CT texture analysis has relatively high value in the differential diagnosis of thymoma and high-density thymic cyst.
ZHANG Yu
,
XIE Fengzhi
,
ZHOU Changqing
. Differential diagnosis of thymoma and high-density thymic cyst based on contrast-enhanced CT texture analysis[J]. Journal of Baotou Medical College, 2025
, 41(11)
: 75
-81
.
DOI: 10.16833/j.cnki.jbmc.2025.11.014
[1] Fang WT, Chen WH, Chen G, et al. Surgical management of thymic epithelial tumors: a retrospective review of 204 cases[J]. Ann Thorac Surg, 2005, 80(6): 2002-2007.
[2] Kent MS, Wang T, Gangadharan SP, et al. What is the prevalence of a "nontherapeutic" thymectomy[J]. Ann Thorac Surg, 2014, 97(1): 276-282; discussion82.
[3] 孔丹, 张建东, 单文莉, 等. CT影像组学模型对甲状腺良恶性结节的鉴别价值[J]. 中华放射学杂志, 2020, 54(3): 187-191.
[4] Miyazawa R, Matsusako M, Nozaki T, et al. Incidental mediastinal masses detected at low-dose CT screening: prevalence and radiological characteristics[J]. Jpn J Radiol, 2020,38(12): 1150-1157.
[5] Wang S, Ao YQ, Jiang JH, et al. How can the rate of nontherapeutic thymectomy be reduced[J]. Interact Cardiovasc Thorac Surg, 2022, 35(3): ivac132.
[6] Brown RA, Frayne R. A comparison of texture quantification techniques based on the Fourier and S transforms[J]. Med Phys, 2008, 35(11): 4998-5008.
[7] Dettori L, Semler L. A comparison of wavelet, ridgelet, and curvelet-based texture classification algorithms in computed tomography[J]. Comput Biol Med, 2007, 37(4): 486-498.
[8] Lubner MG, Smith AD, Sandrasegaran K, et al. CT texture analysis: definitions, applications, biologic correlates, and challenges[J]. Radiographics, 2017, 37(5): 1483-1503.
[9] 王健, 卢璐, 周荣真, 等. 前纵隔胸腺囊肿的CT诊断及定量分析在鉴别诊断中的价值[J]. 实用放射学杂志, 2019, 35(9): 1415-1418.
[10] Yamazaki M, Oyanagi K, Umezu H, et al. Quantitative 3D shape analysis of CT images of thymoma: a comparison with histological types[J]. AJR Am J Roentgenol, 2020, 214(2): 341-347.
[11] 叶勇军, 胡玉敏, 孔春丽, 等. 基于CT平扫的影像组学列线图模型对高密度胸腺囊肿和低危胸腺瘤的鉴别诊断价值[J]. 中国中西医结合影像学杂志, 2022, 20(4): 319-322.
[12] Zhang HW, Graham CM, Elci O, et al. Locally advanced squamous cell carcinoma of the head and neck: CT texture and histogram analysis allow independent prediction of overall survival in patients treated with induction chemotherapy[J]. Radiology, 2013, 269(3): 801-809.
[13] Li H, Zhu YT, Burnside ES, et al. MR imaging radiomics signatures for predicting the risk of breast cancer recurrence as given by research versions of MammaPrint, oncotype DX and PAM50 gene assays[J]. Radiology, 2016, 281(2): 382-391.
[14] Ng F, Ganeshan B, Kozarski R, et al. Assessment of primary colorectal cancer heterogeneity by using whole-tumor texture analysis: contrast-enhanced CT texture as a biomarker of 5-year survival[J]. Radiology, 2013, 266(1): 177-184.
