Note: This copy is for your personal, non-commercial use only. To order presentation-ready copies for distribution to your colleagues or clients, use

Transcription

1 Note: This copy is for your personal, non-commercial use only. To order presentation-ready copies for distribution to your colleagues or clients, use the Radiology Reprints form at the end of this article. ORIGINAL RESEARCH ULTRASONOGRAPHY Manjiri Dighe, MD, DMRE Unmin Bae, MS 2 Michael L. Richardson, MD Theodore J. Dubinsky, MD Satoshi Minoshima, MD, PhD Yongmin Kim, PhD Differential Diagnosis of Thyroid Nodules with US Elastography Using Carotid Artery Pulsation 1 Purpose: Materials and Methods: To explore the sensitivity and specificity of ultrasonographic (US) elastography using carotid arterial pulsation as the compression source for differential diagnosis of thyroid nodules. This HIPAA-compliant study was approved by the ethics committee of the institution, and all patients provided written informed consent. Fifty-eight patients (13 men and 45 women [mean age, 51 years; range, years]) were enrolled. A short US examination and elastography with pulsation of the carotid artery used as the thyroid compression source were performed before fine-needle aspiration. Baseband US data were downloaded for off-line analysis. Elastographic maps and the thyroid stiffness index were calculated. The Kruskal-Wallis nonparametric rank sum test was used to assess equality of population medians among the different types of thyroid nodules; the R software environment was used for statistical computing and graphics ( Results: Conclusion: Thyroid stiffness index calculated with elastography using carotid arterial pulsation as the compression source was effective in helping distinguish between papillary carcinomas (n 10) and other lesions (n 43) because papillary carcinomas were stiffer than other lesions (P.0039). It is possible to distinguish between papillary carcinomas and other lesions with the thyroid stiffness index calculated from US elastography using carotid arterial pulsation. RSNA, From the Departments of Radiology (M.D., M.L.R., T.J.D., S.M.), Electrical Engineering (U.B.), and Bioengineering (Y.K.), University of Washington Medical Center, Box , 1959 NE Pacific St, Seattle, WA From the 2006 RSNA Annual Meeting. Received October 4, 2007; revision requested November 27; revision received January 5, 2008; accepted February 18; final version accepted March 3. Address correspondence to M.D. ( dighe@u.washington.edu). 2 Current address: Department of Ultrasound, Philips Medical Systems, Bothell, Wash. RSNA, Radiology: Volume 248: Number 2 August 2008

2 The prevalence of thyroid nodules is about 3% 8% in the general population (1 4) and is greater than 50% after age 65 years (5). About 5% of adults in the United States have palpable thyroid nodules (6). The number of thyroid nodules being detected has increased because of improvements in medical imaging. Studies indicate a 5% 15% prevalence of malignancy for thyroid nodules (7,8). Ultrasonographic (US) examination is an accurate method for detecting thyroid nodules, but its use in differentiating between benign and malignant thyroid nodules is relatively low (9). Fine-needle aspiration (FNA) is the standard procedure to determine whether a thyroid nodule is cancerous. However, FNA is an invasive procedure, and about 10% 20% of FNAs yield inadequate results and lead to repeat biopsy (7). Palpation has been used in clinical examination to assess the degree of firmness of a thyroid nodule. However, palpation is a subjective method, and the assessment varies depending on the size and location of the nodule and on the examiner (10). US elastography was developed to obtain information on tissue stiffness noninvasively (11 14). Elastography has been successfully applied to breast lesions and more recently to the prostate (15). In previous ex vivo and in vivo studies, it has been suggested that there is considerable difference between the stiffness in the normal thyroid tissue and thyroid tumors (16,17). However, out-of-plane motion during external compression and the carotid arterial pulsation pushing the thyroid gland independently are limitations of this technique (18). In a feasibility study with 12 patients undergoing surgical treatment, Advances in Knowledge Thyroid stiffness index obtained with elastography can be used as a means of comparing the tissue stiffness in different nodules. A statistically significant difference in the thyroid stiffness index was seen between papillary carcinoma and other thyroid lesions (P.0039). we showed that the pulsation of the carotid artery can serve as the thyroid compression source for elastography (19). The aim of the current prospective study was to explore the sensitivity and specificity of US elastography using carotid arterial pulsation as the compression source for differential diagnosis of thyroid nodules. Materials and Methods This HIPAA-compliant study was approved by the ethics committee of our institution, and all patients provided written informed consent. One of the authors (U.B.) joined industry during the review process of the manuscript. The nonindustry personnel were the study guarantors and had complete control of the data. Patients Fifty-eight consecutive patients, 13 men and 45 women, referred for thyroid FNA were recruited for this pilot study from February 2006 to January The mean age of the patients was 51 years (range, years). Sixty-two nodules were initially included (two patients had two nodules, and one patient had three nodules). Nine patients (each with one nodule) were later excluded because of inadequate sample at FNA (n 8) or incomplete data acquisition at elastography. Thus, our final study population consisted of 49 patients with 53 nodules. Mean size of the nodules was cm (range, cmto cm). The US characteristics, listed in Table 1, were retrospectively analyzed by T.J.D. (senior radiologist with more than 15 years experience in US) without knowledge of the final histopathologic diagnosis or the elastographic findings. Implications for Patient Care Papillary carcinoma can be noninvasively differentiated with elastography by using the thyroid stiffness index. Multiple nodules in a patient can be evaluated with elastography to select probable papillary carcinoma. None of the patients included in this study had significant carotid atherosclerosis, which was assessed by noting the presence of any carotid plaque, calcification, or mural thrombus in the neck. The patients blood pressures were recorded, along with age, sex, and history of previous thyroid or other diseases. Twenty patients had a history of thyroid disease. Ten patients had a history of malignancy elsewhere in the body (ie, other than the thyroid gland). None of the patients underwent iodine 131 scanning before the thyroid FNA. The Society of Radiologists in Ultrasound criteria (20) for selection of the thyroid nodules for aspiration were used to select patients and nodules for FNA. US Examination Before FNA, a short US examination was performed with the HiVision 5500 system (Hitachi Medical Systems America, Twinsburg, Ohio) by three lead sonographers with more than 5 years of experience in scanning and approximately 4 hours of dedicated training in acquiring elastograms. These sonographers were trained to acquire elastograms during our preliminary study assessing the feasibility of this technique. The probe was placed lightly on the patient s neck, with a large amount of US gel placed to create a stand-off pad. No external compression was applied for elastography because the carotid arterial pulsation was used as the compres- Published online before print /radiol Radiology 2008; 248: Abbreviations: FNA fine-needle aspiration ROC receiver operating characteristic Author contributions: Guarantors of integrity of entire study, M.D., U.B.; study concepts/study design or data acquisition or data analysis/interpretation, all authors; manuscript drafting or manuscript revision for important intellectual content, all authors; approval of final version of submitted manuscript, all authors; literature research, M.D., U.B., T.J.D.; clinical studies, M.D., U.B.; experimental studies, M.D., U.B., S.M., Y.K.; statistical analysis, M.D., U.B., M.L.R., Y.K.; and manuscript editing, all authors See Materials and Methods for pertinent disclosures. Radiology: Volume 248: Number 2 August

3 sion source. A 7.5-MHz linear transducer (L53; Hitachi Medical Systems America) was used, and the average section thickness of the elastogram acquired was about 5 mm. Multiple scans (two transverse scans and two longitudinal scans) were obtained. The baseband US data were acquired from these scans, and the corresponding strain images were generated offline by using direct strain estimation by software developed by our bioengineering department (21). The processing time for these data sets was approximately 24 hours. In strain estimation, we used a large correlation window of 2.9 mm with a small window separation (ie, distance between correlation windows) of 0.1 mm to increase the strain signal-to-noise ratio without substantially compromising the strain resolution. We defined the thyroid stiffness index as follows: strain near the carotid artery (high strain area)/lowest thyroid strain (19). Elastographic maps and the thyroid stiffness index (19) (Figs 1, 2) were calculated by U.B. and Y.K. For each nodule, two thyroid strain index values were computed from two transverse scans of the nodule and were then averaged to provide the thyroid stiffness index of the nodule. The mean absolute difference between the two thyroid stiffness index values from two transverse scans was The size of the regions used for computation of the thyroid stiffness index was 2 2 mm. The diagnosis determined at FNA and the surgical results were the reference standards. Statistical Analysis The data were analyzed by using the R software environment for statistical computing and graphics ( The Kruskal-Wallis nonparametric rank sum test was used to test for equality of population medians among the different types of thyroid nodules. Maximum likelihood estimation of the curve parameters for a binomial receiver operating characteristic (ROC) curve was performed by using the ROCKIT software package (University of Chicago, Chicago, Ill; Table 1 US Features of 53 Nodules in 49 Patients Variable No. of Nodules Location Right 26 (49) Left 27 (51) US features Absent normal thyroid tissue around nodule 10 (19) Composition Solid 31 (58) 50% solid 16 (30) 50% solid 6 (11) Cystic 0 (0) Complex lesions (incompletely solid lesions) Spongelike 3 (6) Cyst with mural nodule 1 (2) Cyst with thick wall 4 (8) Cyst with multiple thick irregular septa 0 (0) Solid lesion with scattered regular cystic regions 5 (9) Solid lesion with irregular cystic region 9 (17) Echogenicity if 25% solid Hyperechoic 6 (11) Isoechoic 22 (42) Slightly hypoechoic 9 (17) Moderately hypoechoic 12 (23) Very hypoechoic 2 (4) Echotexture if 25% solid Homogeneous 10 (19) Heterogeneous diffuse 30 (57) Heterogeneous focal 4 (8) Halo* Regular, thin 25 (47) Irregular, thick 16 (30) Indeterminate 12 (23) Margins Well-defined 35 (66) Ill-defined 17 (32) Calcifications None 35 (66) Microcalifications 4 (8) Colloid 3 (6) Microcalcifications vs colloid 1 (2) Macrocalcifications 12 (23) Peripheral/eggshell 2 (4) Doppler Central 6 (11) Peripheral 22 (42) Central and peripheral 18 (34) None 10 (19) Note. Numbers in parentheses are percentages. * The halo in 12 lesions could not be classified because they did not fit into the thin and regular or thick and irregular options. Two nodules had both macrocalcifications and colloid-appearing echotexture, and one lesion had both macrocalcifications and peripheral calcifications. 664 Radiology: Volume 248: Number 2 August 2008

4 /software_index6.htm). ROCKIT was also used to estimate the false-positive fraction and the true-positive fraction for different values of the critical value of thyroid stiffness index. Results The median thyroid stiffness index of papillary carcinoma (26.04) was significantly different from those of the other lesions (15.53) (P.0039) (Table 2). However, there was no significant difference in thyroid stiffness index among nodular goiter/colloid cyst, follicular lesion, Hürthle cell lesion, and lymphocytic thyroiditis (P.99) (Fig 3). There was no correlation between the pathologic diagnosis and patients blood pressures. In addition, 10 of 53 nodules did not have a considerable Figure 1 amount (eg, 2 2 mm) of thyroid tissue outside the tumor. Nine of the 10 papillary carcinomas were correctly identified at FNA (Table 3). One lesion was not identified at FNA because it represented a small (2-mm) focus at histopathologic examination. The thyroid stiffness index in this lesion was One lesion with a thyroid stiffness index of 29.4 was suspected of being a thyroid neoplasm at FNA; however, this lesion was a fibrotic nodule at histopathologic examination. The ROC curve for distinguishing papillary carcinoma from other lesions by using thyroid stiffness index is shown in Figure 4. The area under the ROC curve for diagnosing papillary carcinomas was (95% confidence interval: 0.78, 0.97). A critical value of thyroid stiffness index of 18.0 corresponds to a sensitivity of 87.8% and a specificity Figure 1: Example of selecting region of interest (ROI) to compute thyroid stiffness index within a nodule. (a) Transverse US image showing a nodule (arrows) within thyroid gland (Tg), the trachea (Tr), and the carotid artery (C). (b) Combined US image and elastogram of this nodule; average strain of stiffest region inside a thyroid nodule (ROI #1) is used as denominator of thyroid stiffness index. (c) Combined US image and elastogram of this nodule with a different scale to better visualize the areas of highest strain near the carotid artery. Average strain of the highest-strain regions near the carotid artery (ROI #2, ROI #3) is used as numerator of thyroid stiffness index. Range of the color map is adjusted to better visualize ROIs. of 77.5% for prediction of papillary carcinoma (Table 4). Discussion US elastography was developed to determine tissue stiffness and strain information noninvasively (11 14). Strain represents the amount of deformation; thus, stiff tissue shows less strain than softer tissue. The strain images or elastograms are displayed with a color map. Elastography has been shown to be useful in the differential diagnosis of breast and prostate tumors (14,15). Traditionally, strain estimators aim to accurately derive tissue displacements between before and after compression and to compute strain from the displacements. However, the displacement can be as large as 1000 times the strain for typical compression levels used in US elasticity imaging. Error in displacement estimation leads to a large variance in strain, thereby resulting in poor signal-to-noise ratio for the estimated strain. Bae and Kim (21) have developed a novel strain estimator that can be used to directly estimate strain from the phase of temporal and spatial correlation instead of estimating small strain from large displacements. Signalto-noise ratio of the elastogram and contrast-to-noise ratio of the elastogram measured by using the direct strain estimator are at least three times and six times larger, respectively, than values obtained by using conventional displacement-based strain estimators. This indicates that the direct strain estimator can substantially improve accuracy and lesion detectability in US elasticity imaging. In addition, the direct strain estimator is computationally efficient compared with conventional estimators, thus enabling the real-time implementation and clinical use of this new US imaging mode. A thyroid lesion may have different levels of stiffness within it, depending on the cellularity and composition of the nodule. Information from these elastograms helps assess the relative stiffness of the lesion compared with its surrounding tissues and within itself. However, this information cannot be used to Radiology: Volume 248: Number 2 August

5 compare the stiffness of thyroid lesions from different patients because strain changes with applied compression. When pulsation of the carotid artery is used as the compression source, strain near the carotid artery can indicate the amount of compression applied by carotid arterial pulsation. Because the thyroid gland is located between the trachea and the carotid artery, lateral expansion of the carotid artery during systole compresses the thyroid gland against the trachea. As a result, the thyroid gland expands in the anteroposterior direction, that is, in the direction of the US probe (beam axis). Because US is sensitive to detecting motion along the beam axis, deformation of the thyroid gland in this direction can be readily detected. Because the thyroid stiffness index is a ratio between the carotid strain and the strain in the thyroid nodule, the varying strain from change in the blood pressure did not affect the ratio. Lyshchik et al (18) concluded that elastography is a promising imaging technique that can assist in differential diagnosis of thyroid nodules. Among the criteria evaluated at off-line elastography, only a tumor to normal tissue strain ratio greater than 4 was strongly associated with thyroid cancer (P.001), with a specificity of 96% and a sensitivity of 82%. The authors reported two limitations of their study: the out-of-plane motion during freehand external compression and the artifact caused by the interference of carotid arterial pulsation with the external compression (18). In our study, we evaluated the utility of carotid arterial pulsation induced elastography rather than using freehand external compression. Because we applied limited external compression in our study and we used the movement of the carotid artery as the compression source, we were able to avoid both artifacts seen in the study by Lyshchik et al. In addition, we developed a quantitative metric called thyroid stiffness index so that the stiffness of lesions from different patients could be compared. On the basis of the fact that the carotid arterial pulsation is the compression Figure 2 Figure 2: (a) Left: Transverse US image in a 62-year-old man shows a predominantly solid nodule (arrows) occupying entire thyroid gland. Right: Elastogram shows the predominantly homogeneous appearance within nodule, with small areas of decreased stiffness (blue indicates stiffer part; red indicates softer part). This was diagnosed as a follicular lesion at FNA. (b) Left: Transverse US image in a 60-year-old woman shows a predominantly solid nodule with small cystic areas (arrows). Right: Elastogram shows the heterogeneous appearance within nodule. This was diagnosed as a nodular goiter at FNA. (c) Left: Transverse US image in a 36-year-old woman shows a hypoechoic nodule (arrows) with irregular borders and tiny punctate calcifications in it. Right: Elastogram at same level shows stiff areas within lesion (blue region). This was diagnosed as papillary carcinoma at FNA. C carotid artery; Tr trachea. source, we estimated the amount of compression to compensate for the strain variation due to a change in compression. The thyroid stiffness index can be computed for any region of the thyroid, including a tumor and normal thyroid tissue, whereas the tumor to normal tissue strain ratio (18) is a measure of relative stiffness between a tumor and its surrounding normal tissue. Computing the tumor to normal tissue strain ratio assumes that there is some normal thyroid tissue outside the tumor. However, this assumption may not always be valid because thyroid tumors can occupy the whole thyroid lobe. Furthermore, the tissue outside the tumor may not be normal; histopathologic re- 666 Radiology: Volume 248: Number 2 August 2008

6 Figure 3 sults available for a few of these cases revealed that the tissue outside the papillary carcinoma was lymphocytic thyroiditis or nodular goiter. On the contrary, the thyroid stiffness index can be computed for all thyroid lesions regardless of the existence and normality of thyroid tissue outside a nodule. The median thyroid stiffness index of papillary carcinoma (26.04) was sig- Table 2 Diagnosis at FNA and Correlation with Thyroid Stiffness Index in 53 Nodules Thyroid Stiffness Index Diagnosis at FNA Median Range 2 SDs Figure 3: Box plot shows that thyroid stiffness index values for lesions classified as papillary carcinoma (CA) (n 10), follicular lesions (n 25), multinodular goiter or colloid cysts (n 12), thyroiditis (n 4), and Hürthle cell lesions (n 2) at FNA were 26.04, 12.75, 14.04, 11.97, and 13.28, respectively. This suggests that there was no significant difference in thyroid stiffness index among multinodular goiter/colloid cyst, follicular lesion, Hürthle cell lesion, and lymphocytic thyroiditis (P.99). Papillary carcinoma (n 10) Follicular lesion (n 25) Multinodular goiter or colloid cyst (n 12) Thyroiditis (n 4) Hürthle cell lesion (n 2) Note. SD standard deviation. Table 3 Diagnosis at Histopathologic Evaluation (after Surgery) and Correlation with Thyroid Stiffness Index in 20 Nodules Thyroid Stiffness Index Histopathologic Diagnosis Median Range 2 SDs Papillary carcinoma (n 10) Fibrotic nodule (n 1) Multinodular goiter (n 8) Follicular adenoma (n 1) Figure 4 Note. SD standard deviation. Table 4 Estimated Relationship between Critical Test Result Value and Corresponding Operating Point on Fitted Binomial ROC Curve Critical Test Result Value of Thyroid Stiffness Index False-Positive Fraction True-Positive Fraction Figure 4: ROC curve for distinguishing between papillary carcinoma (n 10) and other thyroid lesions (n 43). Blue line represents the observed (empirical) ROC curve. Black curved line represents the best fit for a binomial ROC curve using maximum likelihood estimation of the curve parameters. Dashed red lines represent the 95% confidence interval for the best-fit curve. The area under the best-fit ROC curve for diagnosing papillary carcinomas was (95% confidence interval: 0.78, 0.97) Note. Critical test result value separates positive results from negative results. Radiology: Volume 248: Number 2 August

7 nificantly different from those of the other lesions (15.53) (P.0039). However, there was no significant difference in thyroid stiffness index among nodular goiter/colloid cyst, follicular lesion, Hürthle cell lesion, and lymphocytic thyroiditis. This finding suggests that papillary carcinoma is stiffer than other lesions, such as nodular goiter and follicular lesion, and is consistent with the findings of two previous studies (17,18). The area under the ROC curve in our study was Sensitivity and specificity of the thyroid stiffness index obviously vary depending on which critical value of thyroid stiffness index is chosen. For our ROC curve, a critical value of 18.0 corresponds to a sensitivity of 87.8% and a specificity of 77.5%. Our study recruited patients who underwent FNA; thus, thyroid lesions in a broad range of categories (from low to high probability of malignancy) were included. It is important to evaluate the efficacy of elastography for the population scheduled for FNA because noninvasive differential diagnosis of thyroid lesions at this stage can benefit many patients by reducing the number of FNA procedures and improving the diagnostic performance of FNA. According to our data, patients with a thyroid stiffness index greater than 18 would require FNA to allow distinction between benign lesions and papillary carcinoma. In our study, we compared the strain in the tumor with the strain from the carotid artery (the source of the strain). Our method would avoid any bias in the estimation of strain in the tumor from a background of diffuse thyroid abnormality (eg, thyroiditis). Because our study used the FNA diagnosis as the reference standard, limitations of FNA also became limitations of our study. Inadequate or nondiagnostic smears often occur in the setting of cystic or vascular lesions (7,22), as seen in eight cases in our study. It is difficult to distinguish between follicular carcinoma and follicular adenoma on the basis of the thyroid stiffness index because FNA was unable to help differentiate follicular carcinoma from follicular adenoma. Another limitation is that many of our cases were classified as both nodular goiter and follicular lesion at FNA; this limited our ability to distinguish between these lesions at elastography. Because of these reasons, we evaluated the elastographic results mainly for papillary carcinoma and other lesions. In our study, there were only 10 cases of papillary carcinoma out of 62 thyroid lesions. Because of the small number of papillary carcinomas, the 90% sensitivity we found is preliminary and needs to be further evaluated with more cases of papillary carcinoma. Another limitation of the study is the longer postprocessing time for the thyroid stiffness index. This increase in time limits this assessment to the research arena until the process can be performed more rapidly. We acquired US data for two transverse views of a thyroid nodule rather than covering the whole thyroid nodule. Thus, we may have missed small focal areas of cancer, which may explain the low thyroid stiffness index values for two cases of papillary carcinoma. Another issue pertains to a relatively large section thickness ( 5 mm) of the ultrasound probe used in the study. A thicker section and sample volume would mean that information from smaller lesions would be averaged out with that from other tissues. In our future study, we plan to obtain multiple scans spanning the entire thyroid nodule. This may help us detect a small focal region of cancer. In our study, strain images showed the varying level of stiffness within a lesion. This, we postulate, may be due to the varying histologic characteristics within a nodule. At histopathologic examination, a papillary carcinoma has complex papillae, with a central fibrovascular stalk of variable thickness interspersed with neoplastic follicles that have similar nuclear features, although in various proportions. A fibrous stroma with broad hyaline bands dividing the tumor into irregular lobules is a common feature (23). Psammoma bodies or other calcific concretions may be associated with papillae. Including preoperative patients will help us determine the histologic features of a thyroid nodule and correlate the histologic map with the stiffness distribution obtained from elastography. In conclusion, our pilot study has shown that thyroid elastography using carotid arterial pulsation as a compression source is feasible. Differentiation between types of nodules prior to FNA is also possible with use of thyroid elastography, especially when a quantitative metric such as the thyroid stiffness index is used. However, larger studies are needed to confirm our findings and to distinguish between benign and malignant follicular lesions. References 1. Wiest PW, Hartshorne MF, Inskip PD, et al. Thyroid palpation versus high-resolution thyroid ultrasonography in the detection of nodules. J Ultrasound Med 1998;17: Tomimori E, Pedrinola F, Cavaliere H, Knobel M, Medeiros-Neto G. Prevalence of incidental thyroid disease in a relatively low iodine intake area. Thyroid 1995;5: Carroll BA. Asymptomatic thyroid nodules: incidental sonographic detection. AJR Am J Roentgenol 1982;138: Brander A, Viikinkoski P, Nickels J, Kivisaari L. Thyroid gland: US screening in a random adult population. Radiology 1991; 181: Mortensen JD, Woolner LB, Bennett WA. Gross and microscopic findings in clinically normal thyroid glands. J Clin Endocrinol Metab 1955;15: Utiger RD. The multiplicity of thyroid nodules and carcinomas. N Engl J Med 2005; 352: Gharib H, Goellner JR. Fine-needle aspiration biopsy of the thyroid: an appraisal. Ann Intern Med 1993;118: Hegedus L. Clinical practice: the thyroid nodule. N Engl J Med 2004;351: Takashima S, Fukuda H, Nomura N, Kishimoto H, Kim T, Kobayashi T. Thyroid nodes: re-evaluation with ultrasound. J Clin Ultrasound 1995;23: Tan GH, Gharib H, Reading CC. Solitary thyroid nodule: comparison between palpation and ultrasonography. Arch Intern Med 1995;155: Ophir J, Alam SK, Garra B, et al. Elastography: ultrasonic estimation and imaging of the elastic properties of tissues. Proc Inst Mech Eng H 1999;213: Radiology: Volume 248: Number 2 August 2008

8 12. Gao L, Parker KJ, Lerner RM, Levinson SF. Imaging of the elastic properties of tissue: a review. Ultrasound Med Biol 1996;22: Greenleaf JF, Fatemi M, Insana M. Selected methods for imaging elastic properties of biological tissues. Annu Rev Biomed Eng 2003; 5: Garra BS, Cespedes EI, Ophir J, et al. Elastography of breast lesions: initial clinical results. Radiology 1997;202: Cochlin DL, Ganatra RH, Griffiths DF. Elastography in the detection of prostatic cancer. Clin Radiol 2002;57: Lyshchik A, Higashi T, Asato R, et al. Elastic moduli of thyroid tissues under compression. Ultrason Imaging 2005;27: Rago T, Santini F, Scutari M, et al. Elastography: new developments in ultrasound for predicting malignancy in thyroid nodules. J Clin Endocrinol Metab 2007;92: Lyshchik A, Higashi T, Asato R, et al. Thyroid gland tumor diagnosis at US elastography. Radiology 2005;237: Bae U, Dighe M, Dubinsky T, et al. Ultrasound thyroid elastography using carotid artery pulsation: preliminary study. J Ultrasound Med 2007;26: Frates MC, Benson CB, Charboneau JW, et al. Management of thyroid nodules detected at US: Society of Radiologists in Ultrasound consensus conference statement. Radiology 2005;237: Bae U, Kim Y. Direct phase-based strain estimator for ultrasound tissue elasticity imaging. Conf Proc IEEE Eng Med Biol Soc 2004;2: de los Santos ET, Keyhani-Rofagha S, Cunningham JJ, Mazzaferri EL. Cystic thyroid nodules: the dilemma of malignant lesions. Arch Intern Med 1990;150: Rosai J, Carcangiu ML, Delellis RA. Atlas of tumor pathology: tumors of the thyroid gland, fasc 5, ser 3. Washington, DC: Armed Forces Institute of Pathology, Radiology: Volume 248: Number 2 August

9 Radiology 2008 This is your reprint order form or pro forma invoice (Please keep a copy of this document for your records.) Reprint order forms and purchase orders or prepayments must be received 72 hours after receipt of form either by mail or by fax at It is the policy of Cadmus Reprints to issue one invoice per order. Please print clearly. Author Name Title of Article Issue of Journal Reprint # Publication Date Number of Pages KB # Symbol Radiology Color in Article? Yes / No (Please Circle) Please include the journal name and reprint number or manuscript number on your purchase order or other correspondence. Order and Shipping Information Reprint Costs (Please see page 2 of 2 for reprint costs/fees.) Number of reprints ordered $ Number of color reprints ordered $ Number of covers ordered $ Subtotal $ Taxes $ (Add appropriate sales tax for Virginia, Maryland, Pennsylvania, and the District of Columbia or Canadian GST to the reprints if your order is to be shipped to these locations.) First address included, add $32 for each additional shipping address TOTAL $ $ Shipping Address (cannot ship to a P.O. Box) Please Print Clearly Name Institution Street City State Zip Country Quantity Fax Phone: Day Evening Address Additional Shipping Address* (cannot ship to a P.O. Box) Name Institution Street City State Zip Country Quantity Fax Phone: Day Evening Address * Add $32 for each additional shipping address Payment and Credit Card Details Enclosed: Personal Check Credit Card Payment Details Checks must be paid in U.S. dollars and drawn on a U.S. Bank. Credit Card: VISA Am. Exp. MasterCard Card Number Expiration Date Signature: Please send your order form and prepayment made payable to: Cadmus Reprints P.O. Box Charlotte, NC Note: Do not send express packages to this location, PO Box. FEIN #: Invoice or Credit Card Information Invoice Address Please Print Clearly Please complete Invoice address as it appears on credit card statement Name Institution Department Street City State Zip Country Phone Fax Address Cadmus will process credit cards and Cadmus Journal Services will appear on the credit card statement. If you don t mail your order form, you may fax it to with your credit card information. Signature Date Signature is required. By signing this form, the author agrees to accept the responsibility for the payment of reprints and/or all charges described in this document. RB-9/26/07 Page 1 of 2

10 Radiology 2008 Black and White Reprint Prices Domestic (USA only) # of Pages $221 $233 $268 $285 $303 $ $355 $382 $432 $466 $510 $ $466 $513 $595 $652 $714 $ $576 $640 $749 $830 $912 $ $694 $775 $906 $1,017 $1,117 $1, $809 $906 $1,071 $1,200 $1,321 $1, $928 $1,041 $1,242 $1,390 $1,544 $1, $1,042 $1,178 $1,403 $1,568 $1,751 $1,924 Covers $97 $118 $215 $323 $442 $555 International (includes Canada and Mexico) # of Pages $272 $283 $340 $397 $446 $ $428 $455 $576 $675 $784 $ $580 $626 $805 $964 $1,115 $1, $724 $786 $1,023 $1,232 $1,445 $1, $878 $958 $1,246 $1,520 $1,774 $2, $1,022 $1,119 $1,474 $1,795 $2,108 $2, $1,176 $1,291 $1,700 $2,070 $2,450 $2, $1,316 $1,452 $1,936 $2,355 $2,784 $3,209 Covers $156 $176 $335 $525 $716 $905 Minimum order is 50 copies. For orders larger than 500 copies, please consult Cadmus Reprints at Reprint Cover Cover prices are listed above. The cover will include the publication title, article title, and author name in black. Shipping Shipping costs are included in the reprint prices. Domestic orders are shipped via UPS Ground service. Foreign orders are shipped via a proof of delivery air service. Multiple Shipments Orders can be shipped to more than one location. Please be aware that it will cost $32 for each additional location. Delivery Your order will be shipped within 2 weeks of the journal print date. Allow extra time for delivery. Color Reprint Prices Domestic (USA only) # of Pages $223 $239 $352 $473 $597 $ $349 $401 $601 $849 $1,099 $1, $486 $517 $852 $1,232 $1,609 $1, $615 $651 $1,105 $1,609 $2,117 $2, $759 $787 $1,357 $1,997 $2,626 $3, $897 $924 $1,611 $2,376 $3,135 $3, $1,033 $1,071 $1,873 $2,757 $3,650 $4, $1,175 $1,208 $2,122 $3,138 $4,162 $5,180 Covers $97 $118 $215 $323 $442 $555 International (includes Canada and Mexico)) # of Pages $278 $290 $424 $586 $741 $ $429 $472 $746 $1,058 $1,374 $1, $604 $629 $1,061 $1,545 $2,011 $2, $766 $797 $1,378 $2,013 $2,647 $3, $945 $972 $1,698 $2,499 $3,282 $4, $1,110 $1,139 $2,015 $2,970 $3,921 $4, $1,290 $1,321 $2,333 $3,437 $4,556 $5, $1,455 $1,482 $2,652 $3,924 $5,193 $6,462 Covers $156 $176 $335 $525 $716 $905 Tax Due Residents of Virginia, Maryland, Pennsylvania, and the District of Columbia are required to add the appropriate sales tax to each reprint order. For orders shipped to Canada, please add 7% Canadian GST unless exemption is claimed. Ordering Reprint order forms and purchase order or prepayment is required to process your order. Please reference journal name and reprint number or manuscript number on any correspondence. You may use the reverse side of this form as a proforma invoice. Please return your order form and prepayment to: Cadmus Reprints P.O. Box Charlotte, NC Note: Do not send express packages to this location, PO Box. FEIN #: Please direct all inquiries to: Rose A. Baynard (toll free number) (direct number) (FAX number) baynardr@cadmus.com ( ) Reprint Order Forms and purchase order or prepayments must be received 72 hours after receipt of form. Page 2 of 2