The diagnostic performances of conventional strain elastography (SE),acoustic radiation force impulse (ARFI) imaging and point shear-wave speed (pSWS) measurement for non-calcified thyroid nodules

Abstract

BACKGROUND: Non-calcified thyroid nodules are relatively difficult to diagnose only relying on features of at conventional US images.

OBJECTIVE: To investigate the diagnostic performances of conventional strain elastography (SE), acoustic radiation force impulse (ARFI) SE and point shear-wave speed (pSWS) measurement for non-calcified thyroid nodules.

METHODS: A total of 201 non-calcified thyroid nodules in 195 patients were studied. They were examined with conventional ultrasound (US), conventional SE, ARFI SE and pSWS measurement. Their diagnostic performances and multivariable models were assessed with receiver operating characteristic (ROC) curve and logistic regression analyses respectively.

RESULTS: There were 156 benign and 45 malignant non-calcified nodules proven by histopathology or cystology. The mean diameters of the nodules were 21.2±10.8 mm. Areas under receiver operating characteristic curve (AUCs) of elastography features (ranged, 0.488–0.745) were all greater than that of US (ranged, 0.111–0.332). At multivariate analysis, there were three predictors of malignancy for non-calcified nodules, including pSWS of nodule (odds ratio [OR], 34.960; 95% CI, 11.582–105.529), marked hypoechogenicity (OR, 16.223; 95% CI, 1.761–149.454) and ARFI SE grade (OR, 10.900; 95% CI, 3.567–33.310). US+SE+pSWS owned the largest AUC (0.936; 95% CI, 0.887–0.985; P < 0.05), followed by US+pSWS (0.889; 95% CI, 0.823–0.955), and the poorest was US (0.727; 95% CI, 0.635–0.819).

CONCLUSIONS: ARFI SE and pSWS measurement had better diagnostic performances than conventional SE and US. When US combined with SE and pSWS measurement, it could achieve an excellent diagnostic performance and might contribute a better decision-making of FNA for non-calcified thyroid nodules.

Keywords

Ultrasound strain elastography acoustic radiation force impulse point shear-wave speed measurement elastography

1 Introduction

Thyroid nodule is a common medical pathology and is detected during neck ultrasound (US) up to 67% of the adult population [25]. Although only a small proportion (5–15%) of thyroid nodules are malignant, the diagnostically challenging for nodules is enormous as they are extremely common [36]. Moreover, worrisome related thyroid carcinoma is also great [5].

At conventional US, the features with the high specificities for thyroid cancers are microcalcification, irregular or microlobulated margin, hypoechogenicity or marked hypoechogenicity, solid and taller-than-wide shape [25]. Several studies [8 , 24] displayed that calcifications are reliable factors in identifying thyroid malignancy; while the performance of hypoechogenicity is with low specificity, and irregular margin and taller-than-wide shape are specific characteristics but with low sensitivities. It implies it is relatively difficult to diagnose non-calcified thyroid nodules only relying at conventional US images. Recently, some reports [29 , 37] stated contrast-enhanced ultrasound (CEUS) enables a dynamic evaluation of the microvascularisation of thyroid tumours and is helpful for differentiation between benign and malignant nodules, and yet it needs injecting contrast medium and the cost is also a concern.

Elastography is the most noteworthy of the new technologies in recent diagnostic US systems and able to reveal the differences in the elastic properties of soft tissues [3]. Elastography includes strain elastography (SE) and shear-wave elastography (SWE). SE consists of conventional SE and acoustic radiation force impulse (ARFI) SE which uses ARFI excitation and generates images related to the corresponding tissue displacement within the focused excitation beam. SWE is a further development and it monitors the propagation of shear-wave in tissue, and comprises point shear-wave speed (pSWS) measurement and SWS imaging (SWI) [2, 30]. Although some studies represented elastography was a promising technique [21, 38], there were many substantial arguments about elastography for thyroid nodules. Moon and colleagues [23] retrospectively studied 703 thyroid nodules in comparison to gray-scale US and held the performance of SE was inferior to that of gray-scale US assessment. Asteria and colleagues found the positive predictive value (PPV) of elastography was only 36% comparable to that of microcalcification [1]. The American Thyroid Association (ATA) in their latest guideline did not recommend universal use or widespread adoption of elastography for thyroid nodules [13].

Several factors could affect the results of elastography in thyroid, including the nodule calcifications, cystic components, experience of the operator, and motion artifacts [6, 17]. Kwak et al. [20] suggested that elastography should be selectively used in thyroid nodules without calcifications and cystic changes, but absent of further conclusions. To our knowledge, there were few reports aimed at analyzing the value of elastography for non-calcified thyroid nodules. In this study, we investigated the performance of conventional SE, ARFI SE and pSWS measurement for discriminating malignancy in non-calcified thyroid nodules.

2 Materials and methods

2.1 Patients

The institutional review board approved this retrospective study, and the requirement to obtain informed consent was waived. This study was carried out during the 15-month period from August 2014 to October 2015 in the hospital. A total of 1843 consecutive patients were perfectly imaged with conventional US, convetional SE, ARFI SE and pSWS measurement. Among these patients, 263 patients with 280 non-calcified thyroid nodules met the following inclusion criteria: i, the size of nodules was not less than 10 mm (≥10 mm) and was less than 40 mm; ii, they had fine needle aspiration (FNA) cytology and/or surgery pathology within a month after examinations of all the US, SE and point SWS measurement in thyroid; iii, nodules were without calcifications (including microcalcification and macrocalcification). Then, 79 noduels in 68 patients were excluded because the following criteria: i, 47 nodules in 38 patients were with indeterminate or inadequate cytologic results and without surgery pathology; ii, 6 nodules in 5 patients were diagnosed as “suspicious for papillary thyroid carcinoma” only at cytologic examination but did not undergo surgery pathology; iii, 21 nodules in 20 patients nodule was completely cystic or almost cystic (>50%) appearance on US images; iv, 5 nodules in 5 patients had previous history of surgery in thyroid. Finally, 201 non-calcified nodules in 195 patients were studied (Fig. 1). Among the 195 patients, 190 had one thyroid nodule studied, 4 had two nodules, and 1 had three nodules.

2.2 Imaging procedures

In this study, all the examinations of US, convetional SE, ARFI SE and pSWS measurement for thyroid were performed with the same S2000 US instrument (Siemens Medical Solutions, Mountain View, CA, USA). A 9L4-linear transducer (frequency range, 4–9 MHz) was used for thyroid examination. The patients were placed in a supine position with dorsal flexion of the neck. The gain, focus position and depth of instrument were adjusted appropriately to ensure that the nodules were displayed completely and conspicuously on the screen.

Conventional US examinations were all performed under tissue harmonic mode and images were obtained for each nodule in two orthogonal planes on conventional US. For color Doppler flow imaging (CDFI) examination, the color box was adjusted to include the targeted lesion with some surrounding tissue. The transducer was gently scanned on the body surface so that the small vessels in the lesion were well displayed.

Conventional SE was performed following conventional US by the same investigator. The transducer was placed on the body surface, and a slight manual compression and decompression was applied. Continuous scanning to 3–6 s and quality factor value >60 were obtained, and then the image was frozen to ensure high-quality elastography image. The conventional SE images were coded in color and were displayed in a split-screen mode with conventional US images. Blue color indicates hard tissue, whereas red color indicates soft tissue.

ARFI SE and pSWS measurement were required by virtual touch tissue imaging system and virtual touch tissue quantification system (Siemens Medical Solutions, Mountain View, Calif, USA), respectively. When performing ARFI SE of thyroid nodule, the focused acoustic radiation force pushing pulses were used to deform the tissue, the resulting tissue displacement was measured within the focal region of each push within a specified sampling box, and the distribution of displacement or its normalized values within the sampling box were displayed. The sampling box was set to include the lesion and some surrounding thyroid tissues, and the result was represented as gray-scale image over the conventional B-mode image. The gray-scale value in ARFI SE was classified into two values (black or white) through the comparison between the nodule and the surrounding thyroid tissue. When measuring pSWS of nodule and surrounding tissue, the focused acoustic radiation force pushing pulses of short duration are used to generate shear waves within an organ of interest, and the speed of the shear waves propagating away from the pushing location is measured. The information of shear-wave propagating speed can be measured within a region of interest (ROI) by the virtual touch tissue quantification system, which is obtained by measuring the time to peak displacement at each lateral location. The pSWS was expressed as meter per second (m/s) and did not display color-coded images for elastography. When performing ARFI SE and pSWS measurement, the transducer was gently applied to the body surface with pressure as slightly as possible on the thyroid in the longitudinal plane and the patients were asked to hold their breath, because even slight probe pressure can significantly increase tissue stiffness. The ROI with fixed dimension of 5 mm × 6 mm was placed inside the nodule to obtain pSWS, and cystic areas were avoided. The pSWS of nodule was measured in the nodule for seven times repeatedly. After that, the ROI was moved to the surrounding thyroid tissue at the same depth and the procedure was repeated for 7 times, and surrounding pSWS wasobtained.

2.3 Images interpretation

All the US, SE and pSWS were analyzed in the same setting and in a blind manner by two thyroid radiologists with 8 years and 11 years of experience in thyroid US, 6 years of experience in thyroid SE, and 2 years of experience in thyroid SWE. The identification of the patients, clinical results and pathology results were anonymous to the investigators. In cases of discordance in the evaluation between the two investigators, a third investigator (B.D.C., with 16 years of experience in thyroid US, 7 years of experience in thyroid SE, and 5 years of experience in thyroid SWE reviewed the images to make the final decision.

In this study, calcified nodules were excluded based on conventional gray-scale US. The size of nodule was measured in longitudinal and transverse planes, and the largest diameter was received as evaluating the size of nodule. The other interpreted US features of the nodule included: echogenicity (hyper-, iso-, hypo-echogenicity, or marked hypoechogenicity), composition (completely solid, mixed with <25% cystic portion, or mixed with 25–50% cystic portion), shape (taller-than-wide or wider-than-tall) and margin (irregular, microlobulated, or well-defined). According to the degree and predominance of vascular flow in the peripheral and central portions [16], the CDFI patterns of the nodules were classified into four types: rare flow (only a few vascular spots), peripheral flow (vascular signal in the peripheral portion was higher than that in the central portion), central flow (vascular signal in the central portion was higher than that in the peripheral portion) and mixed flow (vascular signal in the peripheral portion was similar to that in the central portion).

The conventional SE of nodules was classified with five-score system as following: score = 1, the entire nodule is soft; score = 2, part of nodule is hard; score = 3, only margin of nodule is soft; score = 4, the entire nodule is hard; and score = 5, the entire nodule and surrounding area are hard, according to the recent guideline of elastography by the World Federation for Ultrasound in Medicine and Biology (WFUMB) [30]. ARFI SE for the thyroid nodules was divided into ARFI SE grade 1 to 6: grade = 1, predominantly white; grade = 2, predominantly white with few black portions; grade = 3, black and white portion equally; grade = 4, predominantly black with a few white spots; grade = 5, almost completely black, and grade = 6, completely dark, referenced to the Xu’s ARFI imaging grading system [38].

Among 7 measurements of pSWS for nodule and surrounding tissue, the maximal and the minimal values were eliminated, and the averages of remaining 5 measurements were served as the pSWS of nodule and the pSWS of surrounding tissue. Referenced to the relevant reports [10, 31], five measurements were sufficient to assess thyroid stiffness. The pSWS ranged from 0.44 m/s to 8.13 m/s in this study. According to the manufacturer’s suggestion and related reports [22 , 39], the measurement result of “X.XX m/sec” was replaced by 0 m/s or 9.0 m/s, with 0 m/s corresponding to the cystic portion and 9.0 m/s corresponding to the solid portion. Sometimes, SE imaging was added to determine the value of “X.XX m/sec,” with 0 m/s allocated when SE images showed more white, which indicated soft tissue, whereas 9.0 m/s was allocated when SE images showed more black, which indicated hard tissue. The pSWS ratio was average pSWS of nodule divided by the average pSWS of surrounding tissue.

2.4 Reference standard

Surgery histopathology is the golden standard for diagnosing thyroid nodules. Referred to the Bethesda system for reporting thyroid FNA cytology in our study [4 , 11], a classification scheme of 6 cytologic diagnostic categories is recommended for the cytologic report in our study: non-diagnostic, benign, atypia/follicular lesion of undetermined significance, follicular neoplasm/suspicion for a follicular neoplasm, suspicious for malignancy, or malignant. The benign (n = 124) or malignant (n = 6) cases at cytologic examination were included. While the other non-diagnostic, atypia/follicular lesion of undetermined significance, follicular neoplasm/suspicion for a follicular neoplasm and suspicious for malignancy cases without surgery histopathology wereexcluded.

2.5 Statistical analysis

Statistical analysis was performed with SPSS software (version 22.0; SPSS, Chicago, Ill) and MedCalc software (version 13.0, Mariakerke, Belgium). Quantitative values were expressed as means±standard deviations (SD) and ranges. Nonparametric variables were analyzed by Chi-square test, and continuous variables by independent t test. The diagnostic performance of independent features of US, conventional SE, ARFI SE and pSWS measurement were expressed as the area under the receiver operating characteristic (ROC) curve (AUC). The optimal cut-off values of SE and pSWS for distinguishing malignant nodules from benign nodules were obtained yileding the maximum Youden index [28]. The sensitivity, specificity, PPV and negative predictive value (NPV) were calculated according to the diagnostic test 2×2 contingency tables. A total of 8 malignancy predictors in US, SE and pSWS measurement were analyzed by univariate logistic regression analysis with odds ratio (OR) and 95% confidence interval (CI), and then 7 valid predictors were further investigated by multivariate regression analysis. Multivariable diagnostic models were constructed and were evaluated by regression analysis and ROC curve [18], including US, US+conventional SE, US+ARFI SE, US+conventional SE+ARFI SE,US+pSWS measurement and US+SE+pSWS measurement models, and the differences between models were assessed by Z test, referred to the related statistics [18]. Significance was assigned for two-tailed P value <0.05.

3 Results

3.1 Pathologic diagnosis

Among 201 nodules, there were 77.6% (156/201) benign nodules and 22.3% (45/201) malignant nodules. Of 156 benign nodules, 32 were diagnosed by surgery pathology and 124 by FNA cytology. Of 45 malignant nodules, 39 were certified by surgery pathology and 6 by FNA cytology (Fig. 1). At the surgery pathology results, there were 37 cases with papillary thyroid carcinomas and 2 cases with follicular thyroid carcinomas, and there were 26 cases with nodular goiters, 4 cases with follicular adenomas and 2 cases with Hashimoto’s nodules.

3.2 Basic characteristics of patients and US images

The mean age of patients was 49.9±13.2 years (means±SD), and ranged 18–80 years. Males and females accounted for 20.5% and 79.5% of patients,respectively. 77.6% of nodules were benign, and 22.4% of noudles were malignant. The diameters of the nodules ranged from 10.0–62.5 mm, and mean diameters were 21.2±10.8 mm (means±SD). There were no significant differences between benign and malignant nodules on the characteristics of patient age, patient sex, distribution of nodules, size of nodules, and composition of nodule (P > 0.05), while echogenicity, margin, shape and vascular flow at the US images were significantly different (P < 0.05) (Table 1).

3.3 Features of conventional SE, ARFI SE and pSWS between benign and malignant nodules

The pSWS of all nodules were 2.59±1.69 m/s (means±SD), ranged 0.44–8.13 m/s, and the pSWS for surrounding tissues were 2.18±1.01 m/s (means±SD), ranged 0.57–7.46 m/s. The mean pSWS ratio for all nodules was 1.36±1.47 (means±SD), ranged 0.57–7.46. As shown in Table 2, the features of conventional SE, ARFI SE, pSWS measurement of nodule and pSWS ratio were significantly different between benign and malignant nodules (P < 0.001). The pSWS of malignant nodules was obviously higher than that of benign nodules (means±SD, 4.19±2.30 m/s vs 2.03±1.11 m/s), and same to the pSWS ratio (means±SD, 2.32±2.21 vs 1.08±1.04) (P < 0.001) (Figs. 2 and 3).

3.4 Diagnostic performance of US, SE and SWE for non-calcified thyroid nodules

Yielding the maximum Youden index, the optimal cutoff values of conventional SE, ARFI SE, pSWS measurement of nodule and pSWS ratio for diagnosing malignancy were conventional SE score >3, ARFI SE grade >3, pSWS of nodule >2.49 m/s and pSWS ratio >1.22.

The pSWS of nodule had the highest Youden index with 0.745 for diagnostic performance, followed by pSWS ratio with 0.573 and ARFI SE grade with 0.567. Youden indices of hypoechogenicity, marked hypoechogenicity, irregular or microlobulated margin, taller-than-wide shape, and conventional SE score were lower and were 0.111, 0.187, 0.332, 0.240 and 0.488, respectively. The pSWS of nodule and marked hypoechogenicity owned preferable specificity (92.3% and 98.7%) and PPV (75.5% and 81.8%), but sensitivity of marked hypoechogenicity was only 20.0% compared with 82.2% of pSWS measurement (p < 0.001) (Table 3).

At univariate regression analysis, only three US features (marked hypoechogenicity, irregular or microlobulated margin, taller-than-wide shape) and all the elastography features were significant predictors in association with thyroid cancers (p < 0.001). Multivariate analysis showed that marked hypoechogenicity, ARFI SE grade >3 and SWS of nodule >2.49 m/s were significant factors in predicting malignancy, and their ORs were 16.223 (95% CI, 1.761–149.454; p = 0.014), 10.900 (95% CI, 3.567–33.310; p < 0.001), 34.960 (95% CI, 11.582–105.529; p < 0.001) respectively (Table 4).

3.5 Multivariable models of US, SE and pSWS measurement

Multivariable models about US, conventional SE, ARFI SE and pSWS were shown in Table 5. The AUC of US+SE+pSWS measurement was largest with 0.936 (95% CI, 0.887–0.985; P < 0.05), orderly followed by US+pSWS measurement (AUC, 0.889; 95% CI, 0.823–0.955), US+ARFI SE (AUC, 0.844; 95% CI, 0.769–0.918), US+conventional SE (AUC, 0.814; 95% CI, 0.735-0.893) and US (AUC, 0.727; 95% CI, 0.635–0.819). All the AUCs of US+pSWS measurement, US+ARFI SE, US+conventional SE and US+SE were larger than that of US (p, 0.001–0.024). Yet, the differences among the AUCs of US+pSWS measurement, US+ARFI SE, US+conventional SE and US+SE were not significant (p, 0.092–1).In addition, the AUC of conventional SE+ARFI SE+pSWS measurement (ie. compound elastography multivariable model) was 0.922 (95% CI, 0.876–0.955).

4 Discussion

Our study showed that conventional US images were not sufficient to discriminate malignancy in non-calcified throid nodules. Youden indices of hypoechogenicity, marked hypoechogenicity, irregular or microlobulated margin and taller-than-wide shape were 0.111, 0.187, 0.332 and 0.240 respectively, and sensitivities of marked hypoechogenicity and taller-than-wide shape were only 20.0% and 31.1%. Reportedly the incidence of thyroid cancer in non-calcified nodules is 7.8–11.3% and lower than that in calcified nodules [15 , 32], however the number of cancers in non-calcified nodules is nearly to the number in calcified nodules since non-calcified nodules account for more than 60% of the whole thyroid nodules [15]. The importance of US in thyroid nodules is to evaluate the risk of malignancy and aid decision-making about whether FNA is indicated [13]. The ATA in their latest guideline considered the hypoechoic solid thyroid nodule without microcalification as intermediate suspicion, and recommended FNA biopsy to refute malignancy when these nodules >1 cm in the size. This signifies there would be over much FNA cases for non-calcified thyroid nodules only depending on conventional US images. CEUS was reported as a reliable diagnostic tool for thyroid with a high diagnostic accuracy [29 , 37]. Further, CEUS can indentify the microvascularization area in thyroid nodule, so it is helpful in enhancing the FNA success ratio [14]. Thyroid scintigraphy is not necessary for diagnosis in most cases; however, it may be used to detect functional autonomy nodule and devoid of suspicious US features can be excluded from FNA [11].

In our study, conventional SE, ARFI SE and pSWS measurement demonstrated preferable diagnostic performances for non-calcified nodules than US. Among all image features, SWS of nodule had the highest Youden index with 0.745 for diagnostic performance, followed by SWS ratio with 0.573 and ARFI SE grade with 0.567. All the AUCs of elastography (range, 0.488–0.745) were higher than that of US features (range, 0.111–0.332). Multivariate analysis revealed the predicting malignancy of pSWS was better than marked hypoechogenicity, with the OR of 34.960 (95% CI, 11.582–105.529; p < 0.001) in comparison with 16.223 (95% CI, 1.761–149.454; p = 0.014). A meta-analysis [34] included 5,942 thyroid nodules in 38 studies, and the results showed the pooled sensitivity, specificity, PPV, NPV and diagnostic accuracy with elastography were 87.0%, 80.6%, 48.9%, 96.7%, 81.7%, respectively. Another meta-analysis [21] included 1,525 patients in 131 studies, displayed the pooled sensitivity, specificity, and AUC of pSWS measurement for detecting malignant thyroid nodules were 0.73, 0.77 and 0.76, respectively. In our study, the specificities of ARFI SE and pSWS measurement were 92.3% and 87.8%, the AUC of pSWS measurement was 0.873(95% CI, 0.803–0.942), and these results were better than that in above-mentioned reports. These results might be associatied with the effect of calcifications on elastography. Rago et al revealed coarse or peripheral rim calcifications cause an incorrect SE result and proposed nodules with a calcified shell should be excluded from elastography evaluation [27, 33]. Veyrieres et al reported calcifications in thyroid nodules increased the pSWS values and caused six false positive cases in 297 nodules. A previous report about breast lesions in vivo displayed highly dense clusters of micro-calcifications and single macro-calcifications were able to create the appearance of high SWS [12]. Our results implied SE and pSWS measurement may contribute better diagnostic performances for stratifying the malignant risk of nodule without the effect of calcification, and were in favor of the decision-making about whether FNA would be necessary.

When US combined with SE and pSWS measurement, it could obviously enhance the diagnostic performance of US. The multivariable models of US+SE+pSWS measurement possessed the most excellent AUC (0.936; 95% CI, 0.887–0.985), followed with US+pSWS measurement (0.889; 95% CI, 0.823–0.955), and the poorest was US (0.727; 95% CI, 0.635–0.819) (P≤0.001). These stated US united SE and pSWS measurement was the more effective diagnostic tool for non-calcified thyroid nodules and was better to aid decision-making about whether FNA or not. In our study, the differences of between US+conventional SE and US+ARFI SE were not significant (P = 0.449), and same to that between US+SE (including conventional SE and ARFI SE) and US+ARFI SE (p = 1). It indicated the role of US+ARFI SE was same to US+ARFI SE+conventional SE (i.e., US+SE). So, we considered ARFI SE should be sufficient for SE, instead of both conventional SE and ARFI SE.

This study had several limitations. Firstly, selection bias existed. The thyroid nodules included in our study were scheduled for FNA with suspicious US features, completely cystic or almost cystic (>50%) nodules were excluded, and the size not less than 10 mm, which may have affected the diagnostic performance on US, SE and pSWS. The diagnostic value of elastography was inferior associated with the cystic nodule,tumor cysts and nodule less than 10 mm in size [40]. Secondly, there was no compared study for the elastography between non-calcified and calcified thyroid nodules. Further research about the impact of calcifications in nodules on pSWS maybe needed. Thirdly, our study did not include 2-D shear-wave imaging. Further study may analyze the performance of shear-wave imaging in diagnosing non-calcified thyroid nodules. Fourthly, being limited to the scope of quantification in measuring pSWS, our study displayed some “X.XX m/sec”, while the true stiffness of the nodules was uncertain; thus, future refinement of this application is needed. Fourth, the depth of measurement target can impact the pSWS of nodule [7]. Fifthly, not all of the nodules were certified by surgery histopathology. In addition, it was a single-center retrospective study, and further multicenter prospective studies are necessary.

5 Conclusion

In summary, it is relatively hard to stratify the risk of malignancy for non-calcified thyroid nodules only relying on conventional US images, while ARFI SE and pSWS measurement had better diagnostic performances than conventional SE and US. ARFI SE should be sufficient for SE instead of both conventional SE and ARFI SE. When US combined with SE and pSWS measurement, it could achieve an excellent diagnostic performance and contribute a better decision-making of FNA for non-calcified thyroid nodules.

Footnotes

Acknowledgments

This work was supported in part by the Shanghai Hospital Development Center (Grant SHDC12014229), the Science and Technology Commission of Shanghai Municipality (Grants 14441900900 and 15411969000), the Shanghai Municipal Human Resources and Social Security Bureau (Grant 2012045) and the National Natural Scientific Foundation of China (Grants 81401417 and 81501475).

References

Asteria

, Giovanardi

, Pizzocaro

, Cozzaglio

, Morabito

, Somalvico

, et al., US-elastography in the differential diagnosis of benign and malignant thyroid nodules, Thyroid: Official Journal of the American Thyroid Association 18(5) (2008), 523–531.

Bamber

, Cosgrove

, Dietrich

C.F.

, Fromageau

, Bojunga

, Calliada

, et al., EFSUMB guidelines and recommendations on the clinical use of ultrasound elastography. Part 1: Basic principles and technology, Ultraschall in der Medizin (Stuttgart, Germany: 1980) 34(2) (2013), 169–184.

Barr

R.G.

, Nakashima

, Amy

, Cosgrove

, Farrokh

, Schafer

, et al., WFUMB guidelines and recommendations for clinical use of ultrasound elastography: Part 2: Breast, Ultrasound in medicine & biology 41(5) (2015), 1148–1160.

Bongiovanni

, Spitale

, Faquin

W.C.

, Mazzucchelli

and Baloch

Z.W.

, The bethesda system for reporting thyroid cytopathology: A meta-analysis, Acta Cytologica 56(4) (2012), 333–339.

Bresner

, Banach

, Rodin

, Thabane

, Ezzat

and Sawka

A.M.

, Cancer-related worry in Canadian thyroid cancer survivors, The Journal of Clinical Endocrinology and Metabolism 100(3) (2015), 977–985.

Calvete

A.C.

, Rodriguez

J.M.

, de Dios Berna-Mestre

, Rios

, Abellan-Rivero

and Reus

, Interobserver agreement for thyroid elastography: Value of the quality factor, Journal of Ultrasound in Medicine: Official Journal of the American Institute of Ultrasound in Medicine 32(3) (2013), 495–504.

Carlsen

J.F.

, Pedersen

M.R.

, Ewertsen

, Saftoiu

, Lonn

, Rafaelsen

S.R.

, et al., A comparative study of strain and shear-wave elastography in an elasticity phantom, AJR American Journal of Roentgenology 204(3) (2015), W236–W242.

Choi

W.J.

, Park

J.S.

, Kim

K.G.

, Kim

S.Y.

, Koo

H.R.

and Lee

Y.J.

, Computerized analysis of calcification of thyroid nodules as visualized by ultrasonography, European Journal of Radiology 84(10) (2015), 1949–1953.

Cibas

E.S.

and Ali

S.Z.

, The bethesda system for reporting thyroid cytopathology, American Journal of Clinical Pathology 132(5) (2009), 658–665.

10.

Fukuhara

, Matsuda

, Endo

, Donishi

, Izawa

, Fujiwara

, et al., Impact of fibrotic tissue on shear wave velocity in thyroid: An ex vivo study with fresh thyroid specimens, BioMed Research International 2015 (2015), 569367.

11.

Gharib

, Papini

, Garber

J.R.

, Duick

D.S.

, Harrell

R.M.

, Hegedus

, et al., American association of clinical endocrinologists, American college of endocrinology, and associazione medici endocrinologi medical guidelines for clinical practice for the diagnosis and management of thyroid nodules–update, Endocrine Practice: Official Journal of the American College of Endocrinology and the American Association of Clinical Endocrinologists 22(5) (2016), 622–639.

12.

Gregory

, Mehrmohammadi

, Denis

, Bayat

, Stan

D.L.

, Fatemi

, et al., Effect of calcifications on breast ultrasound shear wave elastography: an investigational study, PloS One 10(9) (2015), e0137898.

13.

Haugen

B.R.

, Alexander

E.K.

, Bible

K.C.

, Doherty

G.M.

, Mandel

S.J.

, Nikiforov

Y.E.

, et al., American 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, Thyroid: Official Journal of the American Thyroid Association 26(1) (2016), 1–133.

14.

Hornung

, Jung

E.M.

, Georgieva

, Schlitt

H.J.

, Stroszczynski

and Agha

, Detection of microvascularization of thyroid carcinomas using linear high resolution contrast-enhanced ultrasonography (CEUS), Clinical Hemorheology and Microcirculation 52(2-4) (2012), 197–203.

15.

Kim

B.K.

, Choi

Y.S.

, Kwon

H.J.

, Lee

J.S.

, Heo

J.J.

, Han

Y.J.

, et al., Relationship between patterns of calcification in thyroid nodules and histopathologic findings, Endocrine Journal 60(2) (2013), 155–160.

16.

Kim

D.W.

, In

H.S.

, Choo

H.J.

, Jung

S.J.

and Kim

Y.W.

, Solid and isoechoic thyroid nodules without malignant sonographic features: Comparison of malignancy rate according to nodule size, shape and color Doppler pattern, Ultrasound in Medicine & Biology 39(2) (2013), 269–274.

17.

Kim

J.K.

, Baek

J.H.

, Lee

J.H.

, Kim

J.L.

, Ha

E.J.

, Kim

T.Y.

, et al., Ultrasound elastography for thyroid nodules: A reliable study? Ultrasound in Medicine & Biology 38(9) (2012), 1508–1513.

18.

Knottnerus

J.A.

, Application of logistic regression to the analysis of diagnostic data: Exact modeling of a probability tree of multiple binary variables, Medical Decision Making: An International Journal of the Society for Medical Decision Making 12(2) (1992), 93–108.

19.

Kwak

J.Y.

, Han

K.H.

, Yoon

J.H.

, Moon

H.J.

, Son

E.J.

, Park

S.H.

, et al., Thyroid imaging reporting and data system for US features of nodules: A step in establishing better stratification of cancer risk, Radiology 260(3) (2011), 892–899.

20.

Kwak

J.Y.

and Kim

E.K.

, Ultrasound elastography for thyroid nodules: Recent advances, Radiology 33(2) (2014), 75–82.

21.

Lin

, Chen

, Liu

, Wang

and Li

, Diagnostic performance of shear wave elastography in the identification of malignant thyroid nodules: A meta-analysis, European Radiology 24(11) (2014), 2729–2738.

22.

Meng

, Zhang

, Wu

, Song

and Lu

, Preliminary results of acoustic radiation force impulse (ARFI) ultrasound imaging of breast lesions, Ultrasound in Medicine & Biology 37(9) (2011), 1436–1443.

23.

Moon

H.J.

, Sung

J.M.

, Kim

E.K.

, Yoon

J.H.

, Youk

J.H.

and Kwak

J.Y.

, Diagnostic performance of gray-scale US and elastography in solid thyroid nodules, Radiology 262(3) (2012), 1002–1013.

24.

Moon

H.J.

, Sung

J.M.

, Kim

E.K.

, Yoon

J.H.

, Youk

J.H.

and Kwak

J.Y.

, Diagnostic performance of gray-scale US and elastography in solid thyroid nodules, Radiology 262 (2012), 12.

25.

Moon

W.J.

, Jung

S.L.

, Lee

J.H.

, Na

D.G.

, Baek

J.H.

, Lee

Y.H.

, et al., Benign and malignant thyroid nodules: US differentiation–multicenter retrospective study, Radiology 247(3) (2008), 762–770.

26.

D.G.

, Baek

J.H.

, Sung

J.Y.

, Kim

J.H.

, Kim

J.K.

, Choi

Y.J.

, et al., Thyroid imaging reporting and data system risk stratification of thyroid nodules: Categorization based on solidity and echogenicity, Thyroid: Official Journal of the American Thyroid Association 26(4) (2016), 562–572.

27.

Rago

, Scutari

, Santini

, Loiacono

, Piaggi

, Di Coscio

, et al., Real-time elastosonography: Useful tool for refining the presurgical diagnosis in thyroid nodules with indeterminate or nondiagnostic cytology, The Journal of Clinical Endocrinology and Metabolism 95(12) (2010), 5274–5280.

28.

Schisterman

E.F.

, Perkins

N.J.

, Liu

and Bondell

, Optimal cut-point and its corresponding youden index to discriminate individuals using pooled blood samples, Epidemiology 16(1) (2005), 73–81.

29.

Schleder

, Janke

, Agha

, Schacherer

, Hornung

, Schlitt

H.J.

, et al., Preoperative differentiation of thyroid adenomas and thyroid carcinomas using high resolution contrast-enhanced ultrasound (CEUS), Clinical Hemorheology and Microcirculation 61(1) (2015), 13–22.

30.

Shiina

, Nightingale

K.R.

, Palmeri

M.L.

, Hall

T.J.

, Bamber

J.C.

, Barr

R.G.

, et al., WFUMB guidelines and recommendations for clinical use of ultrasound elastography: Part 1: Basic principles and terminology, Ultrasound in Medicine & Biology 41(5) (2015), 1126–1147.

31.

Sporea

, Sirli

, Bota

, Vlad

, Popescu

and Zosin

, ARFI elastography for the evaluation of diffuse thyroid gland pathology: Preliminary results, World Journal of Radiology 4(4) (2012), 174–178.

32.

Stoian

, Timar

, Derban

, Pantea

, Varcus

, Craina

, et al., Thyroid Imaging Reporting and Data System (TI-RADS): The impact of Quantitative Strain Elastography for better stratification of cancer risks, Medical Ultrasonography 17(3) (2015), 327–332.

33.

Szczepanek-Parulska

, Wolinski

, Stangierski

, Gurgul

and Ruchala

, Biochemical and ultrasonographic parameters influencing thyroid nodules elasticity, Endocrine 47(2) (2014), 519–527.

34.

Veer

and Puttagunta

, The role of elastography in evaluating thyroid nodules: A literature review and meta-analysis, European Archives of Oto-Rhino-Laryngology: Official Journal of the European Federation of Oto-Rhino-Laryngological Societies 272(8) (2015), 1845–1855.

35.

Wendl

C.M.

, Janker

, Jung

, Stroszczysnski

and Jung

E.M.

, Contrast-enhanced ultrasound with perfusion analysis for the identification of malignant and benign tumours of the thyroid gland, Clinical Hemorheology and Microcirculation (2015) Oct 27. [Epub ahead of print].

36.

Wong

K.T.

and Ahuja

A.T.

, Ultrasound of thyroid cancer, Cancer Imaging: The Official Publication of the International Cancer Imaging Society 5 (2005), 157–166.

37.

, Li

and Wang

, Diagnostic value of “absent” pattern in contrast-enhanced ultrasound for the differentiation of thyroid nodules, Clinical Hemorheology and Microcirculation (2015) Dec 15. [Epub ahead of print].

38.

J.M.

, Xu

X.H.

, Xu

H.X.

, Zhang

Y.F.

, Zhang

, Guo

L.H.

, et al., Conventional US, US elasticity imaging, and acoustic radiation force impulse imaging for prediction of malignancy in thyroid nodules, Radiology 272(2) (2014), 577–586.

39.

Zhang

Y.F.

, He

, Xu

H.X.

, Xu

X.H.

, Liu

, Guo

L.H.

, et al., Virtual touch tissue imaging on acoustic radiation force impulse elastography: A new technique for differential diagnosis between benign and malignant thyroid nodules, Journal of Ultrasound in Medicine: Official Journal of the American Institute of Ultrasound in Medicine 33(4) (2014), 585–595.

40.

Zhang

Y.F.

, Xu

H.X.

, Xu

J.M.

, Liu

, Guo

L.H.

, Liu

L.N.

, et al., Acoustic radiation force impulse elastography in the diagnosis of thyroid nodules: Useful or not useful? Ultrasound in Medicine & Biology 41(10) (2015), 2581–2593.