摘要:目的:采用Meta分析的方法评价甲状腺球蛋白在甲状腺良性疾病和甲状腺癌中的临床意义。方法:通过检索MEDLINE、EMBASE、The Cochrane Library, 中国生物医学文献数据库、中国学术期刊全文数据库和其他方式广泛收集文献。根据QUADAS质量评价标准评价纳入文献的质量,用MetaDisc软件对其敏感度、特异度、阳性似然比、阴性似然比等进行合并分析,并进行异质性检验,绘制综合受试者工作特征曲线(summary receiver operator characteristic curve,SROC)。结果:最终纳入5篇文献。合并敏感度0.60,合并特异度0.83,合并比值比2.68, SROC下面积(AUC)=0.645 4。结论:现有研究证实:甲状腺球蛋白在甲状腺癌中的阳性率是甲状腺良性疾病中的2.68倍,有统计学差异,但敏感度不高。尚需更多设计严谨、科学的临床试验进一步证实。Abstract: Objective: To evaluate the quality of the current studies involving the value of serum thyroglobulin in the diagnosis of thyroid benign diseases and thyroid carcinoma. Methods: We comprehensive collected current studies about serum thyroglobulin in thyroid benign diseases and thyroid carcinoma by computer and manual searches. QUADAS items were used for quality assessment in our systematic review. Metadisc software was used to analyze pooled sensitivity, pooled specificity,pooled positive likelihood ratio and pooled negative likelihood ratio,pooled diagnostic test odds ratio and heterogeneity test,and draw summary receiver operator characteristic curve (SROC). Results: Totally 5 studies were included. To identify thyroid benign diseases and thyroid carcinoma, pooled sensitivity was 0.60, pooled specificity was 0.83,pooled odds ratio was 2.68, the area under curve (AUC) was 0.645 4.Conclusion: The results of statistic alanalysis showed that the positive rate of thyroglobulin in thyroid carcinoma is 2.68 times more than in benign thyroid diseases. There was significant difference. But sensitivity was not high and reporting quality of the studies was relatively poor. The conclusion still need more clinical trials to confirm.
Objective To explore the value of virtual touch tissue quantification(VTQ) of acoustic radiation force impulse (ARFI) in the differential diagnosis between benign and malignant thyroid nodules. Methods The ultrasound (US), elastography imaging(EI), and VTQ of ARFI were performed to determine the characteristics and features of 63 thyroid nodules. The pathological diagnosis was the gold standard. According to the receiver operating characteristic curve (ROC) of US, EI, and VTQ, the critical points and diagnostic values of US, EI, and VTQ in diag-nosis of malignant thyroid nodules were achieved. Results Of the 63 nodules, 45 were benign and 18 were malignant. The area under curves of US, EI, and shear wave velocity(SWV) were 0.837(95% CI:0.712-0.962), 0.863(95% CI:0.751-0.974), and 0.900 (95% CI:0.810-0.990) respectively, and all the 3 kinds of technique had diagnostic value(P=0.001), but there were no significant difference among the 3 kinds of technique on the area under curve(P > 0.05). According to the receiver operating characteristic(ROC) curve, the critical point of US in distinguishing benign nodules with malignant nodules was 3 conventional ultrasonography, which displayed a sensitivity of 83.3%, a specificity of 86.7%, and a accuracy of 85.7%. The critical point of EI grades in distinguishing benign nodules with malignant nodules was gradeⅣ, which displayed a sensitivity of 94.1%, a specificity of 82.6%, and a accuracy of 87.3%. The critical value of SWV in distinguishing benign nodules with malignant nodules was 3.39 m/s, which displayed a sensiti-vity of 88.9%, a specificity of 91.1%, and a accuracy of 90.5%. Conclusion US, EI, and VTQ techniques all have diagnostic values in the differential diagnosis between benign and malignant thyroid nodules, and we should make combination with all of the 3 kinds of technique when performing differential diagnosis.
ObjectiveTo assess the effect of the size of thyroid nodules on the diagnostic rate of ultrasound guided aspiration cytology (US-FNAB). MethodsThe data of 1 142 (performed by two doctors, 571 each) thyroid nodules between March 2011 and April 2014 in our hospital were retrospectively analyzed. Yields of US-FNAB were divided into two levels of adequacy and inadequacy according to the classification standard of the Bethesda system. The thyroid nodules were classified into five groups according to the largest diameter:≤5 mm group, 5-10 mm group,10-20 mm group, 20-30 mm group, and <30 mm group. According to the grouping of the nodules and the efficiency of US-FNAB drawed curve, the adequacy rates of alone and total of two examiners in each group were analyzed, respectively. ResultsThe adequacy rates of US-FNAB of alone and total of two examiners in≤5 mm group, 5-10 mm group,10-20 mm group, 20-30 mm group, and <30 mm group was 68.42%, 83.72%, 86.08%, 84.62%, and 73.53% (examiner 1); 68.75%, 70.53%, 81.05%, 86.15%, and 73.91% (examiner 2); 68.59%, 77.53%, 83.59%, 85.47%, and 73.75% (total of two examiners), respectively. The total adequacy rate of US-FNAB of two examiners in≤5 mm group was lower than that in 10-20 mm group (P<0.001) and 20-30 mm group (P=0.001). The adequacy rate of US-FNAB of examiner 1 in 5-10 mm group was higher than that examiner 2 (P=0.001). ConclusionsThe size of thyroid nodules significantly influences the adequate diagnostic rate of US-FNAB. The adequacy rates of US-FNAB of the largest diameter≤5 mm or <3mm were lower. The low adequacy rate of US-FNAB may be associated with cystic degeneration in the larger nodules.
Ultrasound is the best way to diagnose thyroid nodules. To discriminate benign and malignant nodules, calcification is an important characteristic. However, calcification in ultrasonic images cannot be extracted accurately because of capsule wall and other internal tissue. In this paper, deep learning was first proposed to extract calcification, and two improved methods were proposed on the basis of Alexnet convolutional neural network. First, adding the corresponding anti-pooling (unpooling) and deconvolution layers (deconv2D) made the network to be trained for the required features and finally extract the calcification feature. Second, modifying the number of convolution templates and full connection layer nodes made feature extraction more refined. The final network was the combination of two improved methods above. To verify the method presented in this article, we got 8 416 images with calcification, and 10 844 without calcification. The result showed that the accuracy of the calcification extraction was 86% by using the improved Alexnet convolutional neural network. Compared with traditional methods, it has been improved greatly, which provides effective means for the identification of benign and malignant thyroid nodules.
目的 采用回顾性研究的方法，分析B型超声引导下的细针穿刺活检能否提高对甲状腺结节诊断的准确性。方法 54例患者行B型超声引导下的细针穿刺活检，另76例患者采用触诊法行细针穿刺活检，并将54例B超引导下的细针穿刺活检结果与手术结果进行比较。结果 超声引导下细针穿刺活检对可触及的甲状腺结节诊断的敏感性为63%，特异性为71%，准确性为65%，阳性预告值为100%，阴性预告值为70%。标本取材不当率为18%； 触诊法细针穿刺活检的灵敏性为47%，特异性为51%，准确性为43%，阳性预告值是54%，取材不当率为28%。B超引导下的细针穿刺活检准确性显著高于触诊法细针穿刺活检。结论 B超引导下的细针穿刺活检能提高手术前甲状腺癌的诊断。
ObjectiveTo investigate the clinical value of real-time elastosonography in diagnosis of thyroid nodules. MethodsThe clinical date of 168 patients with thyroid nodules, admitted to our hospital from June 2009 to December, were retrospectively analyzed with pathological findings as the reference standard． The findings on elastography were classified into five grades: grade 0 to Ⅳ. The nodules with grade less than Ⅱ were diagnosed as benign, otherwise more than Ⅲ were diagnosed as malignancy. ResultsThere were 208 nodules in 168 patients: nodular goiter was in 125 cases, thyroid adenoma in 36 cases, thyroid malignant tumor in 41 cases including 39 cases for thyroid papillary carcinoma, 1 case for thyroid medullary carcinoma, 1 case for thyroid non-Hodgkin lymphoma. Other nodules were in 6 cases, including diffuse toxic goiter in 1 case, local lymphocytic thyroiditis in 3 cases, and subacute granulomatous thyroiditis in 2 cases. There were 148 thyroid nodules in grade 0-Ⅱ and 60 nodules in grade Ⅲ-Ⅳ. According to the pathological findings, 97.3%(144/148) thyroid nodules of grade 0-Ⅱ were benign, however, 38.3% (23/60) of grade Ⅲ-Ⅳ were benign. Furthermore, 2.7% (4/148) thyroid nodules of grade 0-Ⅱ were malignancy, while 61.7% (37/60) of grade Ⅲ-Ⅳ were malignancy. The sensitivity， specificity, and accuracy of the real-time elastosonography in diagnosis of thyroid nodules was 90.2%(37/41), 86.2%(144/167), and 87.0%(181/208), respectively. ConclusionReal-time elastosonography can reflect the relative hardness of thyroid nodules and provide useful information for diagnosis, which is helpful in differentiating benign and malignant thyroid nodules.