Elastography is a non-invasive technique used to differentiate the elasticity of the diseased and normal tissue. Elastography is used in different modalities of radiology, including ultrasound and magnetic resonance imaging, while sonoelastography is most commonly used of all modalities. Since the mid-1990s, elastography has been in use for evaluation of stiffness and elasticity of soft tissues by giving external pressure(1). It is an alternative technique for biopsy as it is safe and non-invasive. It can detect stiffness and elasticity of muscles as well as other tissues of the body. When a disease develops in the body, the tissues of that particular area become stiff as compared with adjacent normal tissues. When compression is applied to abnormal tissue, it deforms less as compared with normal tissue. Malignant tumors, in which tissue becomes stiffer in comparison with normal tissues, may serve as an example. The standard method used for detection of lesions is palpation, but if a lesion is too small or if it is located too deep, palpation is not useful, and sonoelastography can help clinicians make an accurate diagnosis. Elastography is based on the principle of tissue deformity upon application of external pressure. During elastography, internal or external pressure is applied to tissues, which results in their displacement. If the examined tissue is malignant, it displaces to a lesser degree as malignant tissues become hard. On the other hand, if the tissue is benign, displacement is high because the tissue is soft(1,2). The variation in the soft tissue elasticity helps characterize focal and diffuse pathologies(1,3). During sonoelastography, images are obtained before and after compression and then deformation is evaluated. Tissue hardness or softness appears in the ultrasound monitor in a color box. On an elastogram, soft areas appear as red or yellow, the green color represents firm areas with intermittent stiff areas while hard areas appear as blue. Tissue hardness and elasticity increases due to increased fibrosis and desmoplastic reaction(2,4,5). There are generally 3 techniques of sonoelastography. The first of them is based on mechanical stress where tissue is stressed by internal or external forces. The technique in which the sonologist applies manual compression with the help of the transducer is known as quasi-static elastography (also known as strain imaging); it is a very common technique. The right angle and appropriate compression are necessary and, when not done properly, the image will contain many artifacts. Moreover, in order to obtain an appropriate elastogram, compression has to be applied at least twice(5–8). The third technique is supersonic elasticity imaging (SSI), or shear wave elasticity imaging (SWEI), in which acquisition time is <30 s. The speed of shear waves in soft tissues is a thousand to hundred times slower than longitudinal waves, but high in hard tissues. The propagation speed of shear waves is then directly related to tissue stiffness. Shear wave elastography is similar to acoustic radiation force impulse imaging (ARFI)(5,9). It can be applied clinically for the diagnosis of breast masses, lymph nodes, prostate carcinoma, liver diseases, salivary and parotid gland diseases, pancreatic masses, musculoskeletal diseases and renal diseases. The aim of our study is to evaluate the accuracy of sonoelastography in diagnosing different disorders with the help of previously published studies.
Two reviewers (I.M and R.B) searched the Google scholar, PubMed, NCBI, Medline and Medscape databases from 2007 up to 2015 with the following key terms: diagnostic accuracy, sonoelastography, sensitivity, specificity, superficial lymph nodes, neck nodules, malignancy in thyroid nodules, benign and malignant cervical lymph nodes, thyroid nodules, prostate carcinoma, benign and malignant breast abnormalities, liver diseases, parotid and salivary gland masses, pancreatic masses, musculoskeletal diseases and renal disorders.
Two reviewers (I.M and R.B) independently screened the titles and abstracts of the relevant articles and full articles for inclusion and extraction of data. Any disagreement between the reviewers was resolved by means of a consensus. Studies were eligible if they included information about superficial lymph nodes, neck nodules, malignancy in thyroid nodules, benign and malignant cervical lymph nodes, thyroid nodules, prostate carcinoma, benign and malignant breast abnormalities, liver diseases, parotid and salivary gland masses, pancreatic masses, musculoskeletal diseases, renal disorders and diagnostic accuracy of sonoelastography in these diseases.
The eligible studies were first categorized, and the analysis of the data was performed according to the target conditions. We retrieved the sensitivity and specificity relating to the selected diseases for each individual study and made forest plots. A table was also made for predefined subgroups of types of articles, country, sample sizes as well as sensitivity and specificity values (Tab. 1). Data analysis was performed with the help of Microsoft excel 2017 and Statistical Package for the Social Sciences version 24 (SPSS 24, IBM, Armonk, NY, United States of America).
In total, 69 studies were found after the search. Four were excluded due to duplication, 10 did not include sufficient data for our research, and 9 were rejected on the basis of the title and abstract. The flow chart summarizes the flow records through review in Figure 1. Ultimately, 46 studies were included in the analysis, 16 of which were devoted to lymph nodes, 9 to prostate carcinoma, 6 to breast masses, 5 to liver diseases, 3 to pancreatic masses, 3 to musculoskeletal diseases, 2 to renal diseases and 2 to salivary and parotid gland diseases. Thirteen authors were contacted to supplement the data, but sufficient information was not obtained. All the analyses were performed in the clinical and radiology departments of hospitals.
The data analysis is presented in Fig. 1, Fig. 2, Fig. 3, Fig. 4, Fig. 5, Fig. 6, Fig. 7, Fig. 8 and Fig. 9. Characteristics (study year, country, disease, sensitivity, specificity, sample size and journal name) of the included studies are presented in Table 1. Pooled results of overall sensitivity and specificity of sonoelastography in diagnosing different diseases are shown in Table 2. The overall pooled sensitivity and specificity values of sonoelastography in 16 studies concerning the lymph nodes were 79.31% and 86.52%, respectively. As for the 9 studies on prostate carcinoma, the overall pooled sensitivity and specificity of sonoelastography were 78.43% and 79.71%, respectively. The pooled sensitivity and specificity of sonoelastography in diagnosing breast masses in 6 studies addressing this problem were 86.40% and 75.73%, respectively. In 5 studies on liver diseases, the pooled sensitivity of sonoelastography was 94.94% and specificity was 86.22%. In 2 studies addressing the salivary and parotid gland diseases, the pooled sensitivity and specificity of sonoelastography were 64.95% and 52.75%, respectively. As for the 3 studies on pancreatic masses, the overall pooled sensitivity and specificity of sonoelastography were 94.80% is 76.95%, respectively. The overall pooled sensitivity and specificity in diagnosing musculoskeletal diseases in the 3 selected studies were 92.17% and 83.73%, respectively. In the 2 studies addressing renal diseases, the overall pooled sensitivity of sonoelastography was 82.85% and specificity was 85.00% (Tab. 2). None of the analyses found significant heterogeneity between the studies.
Characteristics of the included studies
Study year | Country | Type of article | Technique | Disease | Sensitivity % | Specificity% | Sample size | Journal |
---|---|---|---|---|---|---|---|---|
2012(10) | China | Meta-analysis | Sonoelastography | Superficial malignant lymph nodes | 74 | 90 | 9 articles |
|
Sonoelastography | Superficial Benign lymph nodes | 90 | 88 | |||||
2009(11) | Italy | Original research | Sonoelastography | Thyroid nodules | 82 | 88 | 25 |
|
Sonoelastography | Deep lymph nodes in mediastinum or abdomen | 85 | 92 | |||||
Sonoelastography | Cervical lymph nodes | 75 | 80 | |||||
2015(12) | USA | Original research | Sonoelastography | Malignancy in thyroid nodules | 79 | 77 | not reported |
|
2009(13) | Different centers of Europe | Original research | Sonoelastography | Superficial lymph nodes | 92 | 83 | 101 |
|
2012(14) | Romania | Original research | Sonoelastography | Benign cervical lymph nodes | 67 | 97 | 69 |
|
Sonoelastography | Malignant cervical lymph nodes | 71 | 97 | |||||
2013(15) | Romania | Review article | Sonoelastography | Superficial lymphadenopathy | 42 | 100 | not reported |
|
2008(16) | Japan | Original research | Sonoelastography | Cervical lymph nodes | 83 | 100 | 85 |
|
2009(17) | Republic of Korea | Original research | Sonoelastography | Thyroid nodules | 70 | 100 | 45 |
|
201218) | Turkey | Original research | Sonoelastography | Thyroid nodules | 86 | 82 | 74 |
|
Sonoelastography | Thyroid nodules | 89 | 82 | |||||
2013(19) | USA | Review article | Sonoelastography | Lymph nodes | 86 | 66 | 24 articles |
|
2012(20) | China | Original research | Sonoelastography | Enlarged cervical lymph nodes | 98 | 64 | 93 |
|
2010(21) | Original research | Sonoelastography | Prostate carcinoma | 90 | 79 |
|
||
2009(22) | Japan | Original research | Sonoelastography | Prostate carcinoma | 73 | 90 | 311 |
|
2010(23) | Japan | Original research | Sonoelastography | Prostate carcinoma | 72 | 86 | 87 |
|
2008(24) | Japan | Original research | Sonoelastography | Prostate carcinoma | 68 | 81 | 107 |
|
2008(25) | Germany | Original research | Sonoelastography | Prostate carcinoma | 75 | 77 | 109 |
|
2008(26) | Austria | Original research | Sonoelastography | Prostate carcinoma | 88 | 72 | not reported |
|
2007(27) | Austria | Original research | Sonoelastography | Prostate carcinoma | 80 | 79 | 15 |
|
2010(28) | NR | Original research | Sonoelastography | Prostate carcinoma | 88 | 79 | not reported |
|
2015(29) | USA | Pictorial Essay | Sonoelastography | Prostate carcinoma | 72 | 76 | not reported |
|
2011(30) | Republic of Korea | Original research | Sonoelastography | Axillary lymph nodes in breast cancer | 81 | 67 | 62 |
|
2009(31) | China | Original research | Sonoelastography | Breast lesions | 98 | 44 | 104 |
|
2008(32) | Italy | Original research | Sonoelastography | Non-palpable breast lesions | 80 | 81 | 278 |
|
2012(33) | USA | Review article | Sonoelastography | Malignant breast abnormalities | 88 | 98 | 9 articles |
|
Sonoelastography | Benign breast abnormalities | 83 | 72 | |||||
2010(34) | Italy | Original research | Sonoelastography | Breast nodules | 89 | 93 | 110 |
|
2009(35) | Italy | Original research | Sonoelastography | Fibrosis staging of chronic liver disease F2-F4 | 91 | 80 | 74 |
|
Sonoelastography | F3-F4 | 96 | 79 | |||||
Sonoelastography | F4 | 94 | 87 | |||||
2003(36) | France | Original research | Sonoelastography | Hepatic fibrosis | 93 | 94 | 106 |
|
2010(37) | Japan | Original research | Sonoelastography | Non-alcoholic fatty liver disease | 100 | 91 | 54 |
|
2013(38) | Turkey | Original research | Sonoelastography | Parotid gland masses | 61 | 59 | 75 |
|
2010(39) | Romania | Original research | Sonoelastography | Pleomorphic adenoma of salivary glands | 69 | 46 | 70 |
|
2012(40) | India | Original research | Sonoelastography | Inflammatory pancreatic disease | 97 | 93 | 166 |
|
2015(29) | USA | Pictorial Essay | Sonoelastography | Pancreatic masses | 95 | 69 |
|
|
2009(13) | Different centers of Europe | Original research | Sonoelastography | Pancreatic masses | 92 | 69 | 101 |
|
2013(41) | Tokyo | Original research | Sonoelastography | Achilles tendon | 100 | 86 | 10 |
|
2009(42) | Austria | Original research | Sonoelastography | Lateral epicondylitis | 100 | 89 | 38 |
|
2011(43) | Republic of Korea | Original research | Sonoelastography | Lateral epicondylitis | 77 | 76 | 48 |
|
2015(44) | USA | Pictorial Essay | Sonoelastography | Fibrosis in kidney disease | 86 | 95 | not reported |
|
201245) | USA | Original research | Sonoelastography | Chronic kidney disease | 80 | 75 | 25 |
|
Pooled sensitivity and specificity
Disease | No of studies | Mean sensitivity | Mean specificity | Std. deviation |
---|---|---|---|---|
Lymph nodes | 16 | 79.31 | 86.52 | 13.196 |
Prostate carcinoma | 9 | 78.43 | 79.71 | 8.327 |
Breast masses | 6 | 86.40 | 75.73 | 6.800 |
Liver diseases | 5 | 94.94 | 86.22 | 3.324 |
Salivary and Parotid Gland | 2 | 64.95 | 52.75 | 5.303 |
Pancreatic masses | 3 | 94.80 | 76.95 | 2.406 |
Musculoskeletal diseases | 3 | 92.17 | 83.73 | 13.568 |
Renal diseases | 2 | 82.85 | 85.00 | 4.031 |
|
|
|
|
|
Real-time elastography is an innovation in the field of radiology. It is non-invasive and complimentary to conventional B-mode ultrasound. Elastography reduces the number of unwanted biopsies by differentiating between benign and malignant masses. Sonoelastography has been in use for many years for diagnosing many diseases. In a study published in 2008, conducted to investigate the differentiation between malignant and benign breast masses before biopsy, 278 women were included with 293 lesions classified in the BIRADS system (Breast Imaging Reporting and Data System). Sonoelastography was performed for all the lesions and 110 of them were found to be malignant and 183 were benign, which also was histologically proven (32). In another study conducted in 2008, the authors wished to learn about the accuracy of B-mode ultrasonography and sonoelastography in the diagnosis of enlarged cervical lymph nodes. For that purpose, 37 patients were enrolled and scanned with B-mode ultrasonography and sonoelastography. The results showed that the accuracy of B-mode ultrasonography was 84%, while the accuracy of sonoelastography was 93%(16). Moreover, sonoelastography is an effective and useful technique for detection of the intratendinous and peritendinous alterations of lateral epicondylitis. It also plays the fundamental role in differentiating between control and diseased extensor tendon origins with high sensitivity and strong correlation with ultrasound findings(42). A prospective study conducted in 2009 aimed to evaluate the accuracy of acoustic radiation force impulse (ARFI) elastography in assessing liver fibrosis in patients with chronic HCV (hepatitis C). For that purpose, 74 patients were enrolled in the study and underwent tests for aspartate aminotransferase (AST)- to-platelet ratio index (APRI) as well as fibro-max and ARFI elastography. The results show that ARFI elastography has a strong correlation with the results of liver biopsy and that it is accurate and reliable for predicting liver fibrosis(35). Sonoelastography is one of the useful qualitative scoring methods in the diagnosis of salivary gland masses, including parotid and sub-mandibular lesions, in terms of detecting benign and malignant masses(45). During sonoelastography of parotid gland tumors, different signs can be frequently seen, such as: garland sign more for malignant tumors than for benign ones, dense core sign for Pleomorphic adenomas, half-half sign for Warthin’s tumor and bull’s eye sign for parotid cysts(46). The results of the previous studies match with our review article that sonoelastography is highly accurate in diagnosing different clinical disorders.
It is concluded that sonoelastography is an easy, rapid and non-invasive technique for detection of many diseases and has high sensitivity and specificity. Tissue elasticity not only varies across different tissues, but also seems to reflect disease-induced alternations in tissue properties. Real-time sonoelastography has been recently applied to the normal and pathologic tissues in muscle and tendon disorders, and it showed promising results and new potential. Therefore, it is expected to be a useful modality for providing novel diagnostic information in musculoskeletal diseases because tissue elasticity is closely related to musculoskeletal pathology. It can also be used as a research tool to provide insight into the biomechanics and pathophysiology of tissue abnormality.