计算机辅助诊断系统辅助超声诊断甲状腺弥漫性病变合并结节良恶性的应用价值

doi:10.16150/j.1671-2870.2022.03.017

摘要/Abstract

摘要：

目的：探究计算机辅助诊断（computer-aided diagnosis,CAD）系统辅助超声医师诊断甲状腺弥漫性病变合并结节良恶性的效能。方法：收集2017年8月至2017年12月在我院就诊并行超声检查的甲状腺弥漫性病变合并结节患者342例（共533个结节）,以病理检查结果为金标准,比较按常规超声诊断方法（依据成人甲状腺结节与分化型甲状腺癌指南的标准）与CAD系统辅助下超声诊断判断甲状腺弥漫性病变合并结节良恶性的灵敏度、特异度、阳性预测值、阴性预测值及受试者操作特征曲线（receiver operator characteristic curve, ROC曲线）的曲线下面积（erea under the curve, AUC）。结果：常规超声诊断甲状腺弥漫性病变合并结节良恶性的灵敏度为96.6%,特异度为72.5%,AUC为0.846;在CAD系统辅助下超声诊断的灵敏度为96.6%,特异度为80.9%,AUC为0.888。在CAD系统辅助下的诊断特异度和AUC均较高（P均<0.01）。结论：对于甲状腺弥漫性病变合并结节的患者,采用超声联合CAD系统诊断甲状腺结节良恶性时,可在保持灵敏度不变的同时提升诊断特异度,以减少不必要的穿刺活检。

关键词: 计算机辅助诊断, 甲状腺弥漫性病变, 甲状腺结节, 诊断效能

Abstract:

Objective: To investigate the effect of ultrasound by computer aided diagnosis (CAD) system in diagno-sing nodules in patients with diffuse thyroid lesions. Methods: A total of 342 patients with diffuse thyroid lesions and thyroid nodules who underwent thyroid surgery in our hospital during August 2017 to December 2017 were enrolled. Based on guidelines for adult thyroid nodules and differentiated thyroid cancer, ultrasound with and without CAD were performed on 533 nodules from the patients. The findings of ultrasound were compared with pathological results, and sensitivity, specificity, positive predictive value, negative predictive value, and area under the ROC curve(AUC) were calculated to observe difference in efficacy between ultrasound and ultrasound by CAD system. Results: The sensitivity, specificity of conventional ultrasound for diagnosing malignant and benign nodules in diffuse thyroid lesions was 96.6%, 72.5%, with AUC of 0.846. While the diagnostic sensitivity, specificity of ultrasound and ultrasound by CAD system was 96.6%, 80.9%, with AUC of 0.888. The use of CAD system enabled ultrasound achieve a higher specificity and AUC (P<0.01). Conclusions: For patients with diffuse thyroid lesions, ultrasound aided by CAD has a better specificity in diagnosis of thyroid nodules, which may reduce unnecessary puncture biopsies.

Key words: Computer-assisted diagnostics, Diffuse lesions of the thyroid gland, Thyroid nodules, Diagnostic efficiency

中图分类号:

R736.1

赵然, 詹维伟, 侯怡卿. 计算机辅助诊断系统辅助超声诊断甲状腺弥漫性病变合并结节良恶性的应用价值[J]. 诊断学理论与实践, 2022, 21(03): 390-394.

ZHAO Ran, ZHAN Weiwei, HOU Yiqing. The application value of computer-aided ultrasound diagnosis system in differentiating malignant from benign thyroid nodules in diffuse thyroid lesions[J]. Journal of Diagnostics Concepts & Practice, 2022, 21(03): 390-394.

图/表 3

参考文献 15

[1]	Haugen BR, Alexander EK, Bible KC, et al. 2015 Ame-rican Thyroid Association Management Guidelines for adult patients with thyroid nodules and differentiated thyroid cancer: the American Thyroid Association Guidelines task force on thyroid nodules and differentiated thyroid cancer[J]. Thyroid, 2016, 26(1):1-133. doi: 10.1089/thy.2015.0020 pmid: 26462967
[2]	Miller KD, Siegel RL, Lin CC, et al. Cancer treatment and survivorship statistics, 2016[J]. CA Cancer J Clin, 2016, 66(4):271-289. doi: 10.3322/caac.21349 URL
[3]	Bray F, Ferlay J, Soerjomataram I, et al. Global cancer statistics 2018: GLOBOCAN estimates of incidence and mortality worldwide for 36 cancers in 185 countries[J]. CA Cancer J Clin, 2018, 68(6):394-424. doi: 10.3322/caac.21492 URL
[4]	Chen W, Zheng R, Baade PD, et al. Cancer statistics in China, 2015[J]. CA Cancer J Clin, 2016, 66(2):115-132. doi: 10.3322/caac.21338 URL
[5]	Morris LG, Tuttle RM, Davies L. Changing trends in the incidence of thyroid cancer in the United States[J]. JAMA Otolaryngol Head Neck Surg, 2016, 142(7):709-711. doi: 10.1001/jamaoto.2016.0230 URL
[6]	Baser H, Ozdemir D, Cuhaci N, et al. Hashimoto′s thyroiditis does not affect ultrasonographical, cytological, and histopathological features in patients with papillary thyroid carcinoma[J]. Endocr Pathol, 2015, 26(4):356-364. doi: 10.1007/s12022-015-9401-8 URL
[7]	Gul K, Dirikoc A, Kiyak G, et al. The association between thyroid carcinoma and Hashimoto′s thyroiditis: the ultrasonographic and histopathologic characteristics of malignant nodules[J]. Thyroid, 2010, 20(8):873-878. doi: 10.1089/thy.2009.0118 URL
[8]	Zhang Y, Wu Q, Chen Y, et al. A clinical assessment of an ultrasound computer-aided diagnosis system in diffe-rentiating thyroid nodules with radiologists of different diagnostic experience[J]. Front Oncol, 2020, 10:557169. doi: 10.3389/fonc.2020.557169 URL
[9]	Choi YJ, Baek JH, Park HS, et al. A computer-aided diagnosis system using artificial intelligence for the diagnosis and characterization of thyroid nodules on ultrasound: initial clinical assessment[J]. Thyroid, 2017, 27(4):546-552. doi: 10.1089/thy.2016.0372 URL
[10]	Li X, Zhang S, Zhang Q, et al. Diagnosis of thyroid cancer using deep convolutional neural network models applied to sonographic images: a retrospective, multicohort, diagnostic study[J]. Lancet Oncol, 2019, 20(2):193-201. doi: 10.1016/S1470-2045(18)30762-9 URL
[11]	Hou Y, Chen C, Zhang L, et al. Using Deep Neural Network to diagnose thyroid nodules on ultrasound in patients with Hashimoto′s thyroiditis[J]. Front Oncol, 2021, 11:614172. doi: 10.3389/fonc.2021.614172 URL
[12]	Lai X, Xia Y, Zhang B, et al. A meta-analysis of Hashimoto′s thyroiditis and papillary thyroid carcinoma risk[J]. Oncotarget, 2017, 8(37):62414-62424. doi: 10.18632/oncotarget.18620 URL
[13]	Park M, Park SH, Kim EK, et al. Heterogeneous echogenicity of the underlying thyroid parenchyma: how does this affect the analysis of a thyroid nodule?[J]. BMC Cancer, 2013, 13:550. doi: 10.1186/1471-2407-13-550 URL
[14]	Faust O, Acharya UR, Tamura T. Formal design methods for reliable computer-aided diagnosis: a review[J]. IEEE Rev Biomed Eng, 2012, 5:15-28. doi: 10.1109/RBME.2012.2184750 URL
[15]	Lam J, Ying M, Cheung SY, et al. A comparison of the diagnostic accuracy and reliability of subjective grading and computer-aided assessment of intranodal vascularity in differentiating metastatic and reactive cervical lymphadenopathy[J]. Ultraschall Med, 2016, 37(1):63-67. doi: 10.1055/s-0034-1384939 pmid: 25140495

分组	灵敏度（%）	特异度（%）	阳性预测值	阴性预测值	AUC
超声	96.6% (86/89）	72.5% (322/444)	0.413 (86/208)	0.991 (322/325)	0.846
超声+CAD	96.6% (86/89)	80.9% (359/444)	0.503 (86/171)	0.992 (359/362)	0.888
P值	>0.990	0.003	/	/	<0.001