Abstract
Objective
This article aims to evaluate the intrareader and interreader agreement of ultrasound (US) gallbladder reporting and data system (GB-RADS) and validate the risk of malignancy in each GB-RADS category.
Materials and methods
This retrospective study comprised consecutive patients with nonacute gallbladder wall thickening who underwent US evaluation between January 2019 and December 2022. Three radiologists independently read the static US images and cine-loops for GB-RADS findings and assigned GB-RADS categories. The intraobserver (static images) and interobserver (static images and cine-loops) agreement was calculated using kappa statistics and Krippendorff's alpha. Another radiologist assigned a consensus GB-RADS category. The percentage of malignancy in each GB-RADS category was calculated.
Results
Static US images of 414 patients (median age, 56 years; 288 women, benign = 45.6% and malignant = 54.4%) and cine-loops of 50 patients were read. There was weak to moderate intrareader agreement for most GB-RADS findings and moderate intrareader agreement for the GB-RADS category for all readers. On static images, the interreader agreement was acceptable for GB-RADS categories. On cine-loops, the interreader agreement for GB-RADS findings and categories was better than static images. The percentage of malignancy was 1.2%, 37%, 71.1%, and 89.1% in GB-RADS 2, 3, 4, and 5 categories.
Conclusion
GB-RADS has moderate intrareader for GB-RADS categories. As originally proposed, the risk of malignancy is negligible in GB-RADS 2 category and highest in GB-RADS 5 category. However, the discriminatory performance of GB-RADS 3 and 4 categories is low. Larger multicenter studies with more readers must assess the reader agreement and validate the GB-RADS systems for wider clinical utilization.
Keywords: gallbladder, gallbladder cancer, gallbladder diseases, data reporting, ultrasound
Gallbladder cancer (GBC) is prevalent in specific geographic regions.1 Advanced GBC has a poor prognosis with a five-year overall survival rate of less than 5%.2 Hence, early diagnosis and management are crucial.2 Ultrasound (US) is often the first imaging test to evaluate patients with suspected gallbladder diseases. Gallbladder wall thickening (GWT) is a common finding at US. It is nonspecific and represents both benign and malignant diseases. Around 20–30% of patients with GBC have GWT-type morphology.3 Unfortunately, there is a lack of literature on the objective categorization of GWT to assess the likelihood of malignancy. Gallbladder reporting and data system (GB-RADS) is a recently proposed US-based system for the GWT assessment and risk stratification in nonacute settings.4 GB-RADS was based on international expert consensus. The agreement of imaging parameters described in the classification system and the diagnostic performance of risk stratification are yet to be validated in the clinical setting.
This study aims to assess the intrareader and interreader agreement of the key findings proposed in GB-RADS and validate the risk of malignancy in GB-RADS categories with the final diagnosis.
Materials and Methods
The institutional ethics committee approved this single-center, retrospective study. The ethics committee waived the requirement for informed consent due to the study's retrospective nature. The study design and data reporting were in accordance with the Standards for Reporting of Diagnostic Accuracy Studies.5
Study Population
Between January 2019 and December 2022, consecutive patients who underwent US in the Radiology section of a Gastroenterology department of a tertiary care referral center, for evaluation of nonacute gallbladder diseases were screened for inclusion (Figure 1). Textual reports and US images were retrieved from the database and reviewed. Patients with GWT were included. As per existing definition, GWT was defined as wall thickness ≥3 mm.4 Patients with masses replacing gallbladder, intraluminal polypoidal lesions, inadequate US [due to technical factors, patient-related factors, gallbladder-related factors (GB-RADS 0, explained later)], and those without a final diagnosis were excluded.
Figure 1.
Patient flow in the study.
Reference Standard (Final Diagnosis)
The final diagnosis of all malignant GWT was based on image-guided fine needle aspiration cytology, histopathology of the core biopsy sample or surgical histopathology of cholecystectomy specimen. The diagnosis of benign disease was based on the histopathology of the cholecystectomy specimen or follow-up imaging (US, computed tomography, or magnetic resonance imaging) at least 6 months after the US scans read in this study.6 A benign diagnosis was considered if the GWT remained stable, reduced or resolved.
US Examination and Analysis
All US exams had been performed in supine position using 2–5 MHz curvilinear probes after at least 6 h of fasting by radiologists with 1–3 years of posttraining experience in abdominal sonography. Patients had been changed to different positions (including lateral decubitus and prone) to scan the entire gallbladder lumen and wall.4 Sagittal and axial static images had been obtained. Following acquisition of static images, a cine loop covering the entire gallbladder (and the gallbladder lesion) had been stored.
The US data were retrieved and anonymized by research fellows (SS, RSi) who were not involved in reading static images or cine-loops. Three radiologists (PS, RS, and PR with one year, 3.5 years, and five years of posttraining experience in the abdominal US) read the static US images independently. The readers with variable experience were employed to mirror the clinical practice. All the radiologists involved in reading were blinded to the patient's clinical data and the final diagnoses. The following findings were assessed on US. Distension (adequate/contracted) and intraluminal contents (calculus/sludge/tumor) were recorded. Mural changes, including the extent of thickening (focal/diffuse), site of thickening (neck/body/fundus), and GB-RADS findings [mural layering, intramural cysts, intramural echogenic foci, interface with liver, extramural extension (biliary involvement, vascular involvement, and liver infiltration)] were recorded.
Based on the above US findings, GWT was categorized using GB-RADS into GB-RADS 2: benign, when the wall thickening showed layered appearance and/or intramural cysts and/or intramural echogenic foci; GB-RADS 3: equivocal, when there was no layered appearance, or intramural cysts/echogenic foci but the interface of GWT with the liver was distinct; GB-RADS 4: malignancy is likely, when there was no layered appearance or intramural features (cysts or echogenic foci) and the interface with the liver was indistinct; and GB-RADS 5: malignancy is highly likely when there was a definite extramural extension in the form of liver infiltration (extramural mass), biliary involvement, or vascular involvement by the GWT. As per the study design, we did not include GB-RADS 0 and 1 cases.
Intrareader and Interreader Agreement
All the readers were initially asked to read the GB-RADS consensus document.4 The senior radiologist (PG, with ten years of experience in abdominal US) trained them to detect the key GB-RADS findings. The static images of ten patients not included in the subsequent readings were used for this face-to-face training session. Within one week of training, all the readers were asked to read the static images of all the patients in consensus. Four weeks after this consensus reading, all the readers performed their 1st independent reading. The second independent reading was performed after four weeks of completing the first independent reading. Each independent reading was completed in 5 days by all the readers. For independent readings, the different radiologists were given images (by research fellows) in different order (generated by computer). After four weeks washout period following second independent reading, all the radiologists independently read the anonymized cine-loops of 50 patients selected randomly based on computed based random numbers (generated by research fellows). The research fellows provided these cine-loops to all the radiologists in different order (which was different from that of the static images). The cine-loops were read in two days by all the readers. The intrareader agreement and the interreader agreement were assessed for the GB-RADS findings and the GB-RADS category on the static images. The interreader agreement for the GB-RADS findings and the GB-RADS category was also evaluated on cine-loops.
GB-RADS Validation and Diagnostic Performance
In cases with discordance among the readers, the senior radiologist (PG, with ten years of experience in abdominal US) who was blinded to the patient's clinical data and final diagnosis read the US images and reassigned a GB-RADS (consensus GB-RADS).
The consensus GB-RADS category was compared with the final diagnosis, and the percentage of malignancy in each GB-RADS category was calculated. The specific diagnosis in benign cases (where histopathology was available) in GB-RADS 4 and 5 categories were reported. Additionally, the diagnostic accuracy of the GB-RADS category for benign and malignant diseases was calculated. These analyses were performed by research fellows (SS, RSi) not involved in reading US data.
Statistical Analysis
Quantitative variables were assessed for normality using the Kolmogorov–Smirnov test. Continuous variables were expressed as median with range. Categorical variables are reported as frequencies and percentages. Intrareader agreement for GB-RADS findings was assessed using kappa (k), while that for GB-RADS categories was assessed using linear-weighted kappa. Interpretation of k was as follows7: k < 0.20, no agreement; k = 0.21–0.39, minimal agreement; k = 0.40–0.59, weak agreement; k = 0.60–0.79, moderate agreement, k = 0.80–0.90, strong agreement; and k > 0.90, almost perfect agreement. Interreader agreement for GB-RADS findings and GB-RADS category was assessed using Krippendorff's alpha (α). An α value above 0.67 was considered acceptable for tentative conclusions, and a value above 0.8 indicated reliable agreement.8 Cross-tabulation analysis was done between the consensus GB-RADS category and the final diagnosis to evaluate the percentage of malignancy in each category. The sensitivity, specificity, positive predictive value (PPV), and negative predictive value (NPV) of the GB-RADS 2 category to identify benign GWT and GB-RADS 4 and 5 categories for malignant GWT were calculated. Similarly, the diagnostic performance of GB-RADS 3 to detect both benign and malignant GWT was assessed. A P-value <0.05 was considered statistically significant. Statistical analysis was done using IBM Statistical Package for the Social Sciences Statistics, release 23; SPSS, Chicago, Ill.
Results
Patient Characteristics
Initially, 774 patients with non-acute gallbladder lesions were screened for inclusion. Among them, 307 patients had mass-replacing gallbladder/intraluminal polypoidal lesions and were excluded from the study. There was inadequate US evaluation of gallbladder in 30 patients, and a final diagnosis was unavailable in 23 patients. Thus, 414 patients were included in the study (median age, 56 years; range, 25–85 years; 288 women [69.5%]). The final diagnosis was benign GWT in 45.6 % (n = 189) and malignant GWT in 54.4% (n = 225) patients (Table 1). In 43 patients, the benign diagnosis was based on six months of follow-up imaging.
Table 1.
Demographic Characteristics and US Findings in 414 Patients.
| Parameter | Value |
|---|---|
| Gender (n) | |
| Male | 126 (30.4) |
| Female | 288 (69.6) |
| Age (years) | |
| Median | 56 |
| Range | 25–85 |
| Intraluminal contents (n) | |
| Calculus | 189 (45.6) |
| Sludge | 144 (34.8) |
| Extent of GWT | |
| Diffuse | 209 (50.5) |
| Focal | 205 (49.5) |
| Site of GWT in cases with focal thickening | |
| Neck | 56 (27.3) |
| Body | 8 (39) |
| Fundus | 33 (16.1) |
| Multifocal | 108 (52.7) |
| Consensus GB-RADS category | |
| 2 | 81 (19.6) |
| 3 | 102 (24.6) |
| 4 | 101 (24.4) |
| 5 | 130 (31.4) |
| Final diagnosis | |
| Benign | 189 (45.6) |
| Malignant | 225 (54.4) |
GB-RADS, gallbladder reporting and data system; GWT, gallbladder wall thickening; Figures in parentheses indicate percentages.
Intrareader Agreement
The intrareader agreement for GB-RADS findings and category are shown in Table 2. There was weak to moderate intrareader agreement for most GB-RADS findings for all readers. There was a moderate intrareader agreement for the GB-RADS category for all readers (Tables S1–S8).
Table 2.
Intrareader Agreement for GB-RADS Findings and Category on Static Images in 414 Patients.
| GB-RADS findings (kappa) | |||||||||
|---|---|---|---|---|---|---|---|---|---|
| Readera | Mural layering | Intramural cysts | Intramural echogenic foci | Interface with liver | Extramural extension | Biliary involvement | Vascular involvement | Liver infiltration | GB-RADS categoryb |
| 1 | 0.518 | 0.515 | 0.535 | 0.643 | 0.618 | 0.509 | 0.517 | 0.509 | 0.648 |
| 2 | 0.618 | 0.422 | 0.506 | 0.708 | 0.716 | 0.602 | 0.564 | 0.602 | 0.714 |
| 3 | 0.572 | 0.422 | 0.447 | 0.760 | 0.680 | 0.562 | 0.559 | 0.562 | 0.718 |
GB-RADS, gallbladder reporting and data system.
Increasing experience.
Linear-weighted kappa.
Interreader Agreement
The interreader agreement is shown in Table 3. The GB-RADS findings showed acceptable interreader agreement for intramural cyst, intramural echogenic foci, liver interface, and extramural extension in the second reading. The interreader agreement was acceptable to reliable, for the GB-RADS category in both readings (α = 0.749 and 0.825).
Table 3.
Interreader Agreement for GB-RADS Findings and Category on Static Images in 414 Patients.
| GB-RADS findings (Krippendorff's alpha) | |||||||||
|---|---|---|---|---|---|---|---|---|---|
| Reading | Mural layering | Intramural cysts | Intramural echogenic foci | Interface with liver | Extramural extension | Biliary involvement | Vascular involvement | Liver infiltration | GB-RADS category |
| 1 | 0.534 | 0.458 | 0.617 | 0.625 | 0.535 | 0.473 | 0.421 | 0.404 | 0.749 |
| 2 | 0.569 | 0.646 | 0.654 | 0.713 | 0.692 | 0.647 | 0.519 | 0.587 | 0.825 |
GB-RADS, gallbladder reporting and data system.
In 50 patients with static images and cine-loops, the interreader agreement was higher for cine-loops for both the GB-RADS findings and the GB-RADS category (Table 4).
Table 4.
Interreader Agreement for GB-RADS Findings and Category on Static Images and Cine-loops in 50 Patients.
| GB-RADS findings (Krippendorff's alpha) | |||||||||
|---|---|---|---|---|---|---|---|---|---|
| Reading | Mural layering | Intramural cysts | Intramural echogenic foci | Interface with liver | Extramural extension | Biliary involvement | Vascular involvement | Liver infiltration | GB-RADS category |
| Static | 0.500 | 0.561 | 0.561 | 0.736 | 0.548 | 0.540 | 0.413 | 0.549 | 0.716 |
| Cine-loops | 0.739 | 0.843 | 0.744 | 0.728 | 0.732 | 0.688 | 0.661 | 0.673 | 0.790 |
GB-RADS, gallbladder reporting and data system.
GB-RADS Validation and Diagnostic Performance
The diagnostic performance of GB-RADS for individual readers is shown in Table S9. A total of 88 cases had discordance among the readers. The consensus GB-RADS categories were GB-RADS 2: 19.6%, GB-RADS 3: 24.6%, GB-RADS 4: 24.4%, and GB-RADS 5: 31.4%. The percentage of malignancy was 1.2% in GB-RADS 2, 37% in GB-RADS 3, 71.1 % in GB-RADS 4, and 89.1% in GB-RADS 5 categories. Of the 31 benign cases in GB-RADS 4 category, histopathological diagnosis was available in 16 patients. The benign lesions in GB-RADS 4 category included chronic cholecystitis (n = 10), xanthogranulomatous cholecystitis (XGC) (n = 4), adenomyomatosis (ADM) (n = 1), and IgG4 cholecystitis (n = 1). In GB-RADS 5 category, there were 15 benign lesions, and the histopathological diagnosis was available in 9 cases. The benign lesions in GB-RADS 5 category included chronic cholecystitis (n = 5) and XGC (n = 4).
There was high specificity (99.5%) and positive predictive value (PPV) (98.7%) of GB-RADS 2 for benign GWT. Similarly, GB-RADS 5 showed a high specificity (92.1%) and PPV (88.5%) for malignant GWT. The diagnostic performance of GB-RADS to identify benign/malignant GWT is shown in Table 5 and Figure 2.
Table 5.
Diagnostic Performance of GB-RADS Categories.
| GB-RADS category | Sensitivity (%) | Specificity (%) | PPV (%) | NPV (%) | |
|---|---|---|---|---|---|
| 2 | 42.3 (35.1–49.7) |
99.5 (97.5–99.9) |
98.7 (93.3–99.9) |
67.2 (61.9–72.2) |
|
| 3 | For benign thickening | 33.3 (26.6–40.5) |
82.6 (77.1–87.3) |
61.7 (51.6–71.2) |
59.6 (53.9–65.1) |
| For malignant thickening | 17.3 (12.6–22.9) |
66.6 (59.4–73.3) |
38.2 (28.7–48.3) |
40.3 (34.8–46.1) |
|
| 4 | 31.1 (25.1–37.6) |
83.6 (77.5–88.5) |
69.3 (59.3–78.1) |
50.4 (44.8–56.1) |
|
| 5 | 51.1 (44.3–57.8) |
92.1 (87.2–95.4) |
88.4 (81.6–93.4) |
61.2 (55.3–66.9) |
|
GB-RADS, gallbladder reporting and data system; PPV, Positive predictive value; NPV, Negative predictive value.
The accuracy of GB-RADS 2 corresponds to benign thickening and that of GB-RADS 4 and 5 pertain to malignant thickening. The accuracy of GB-RADS 3 was calculated for both benign and malignant diagnoses. Data in parentheses indicate 95% confidence intervals.
Figure 2.
2 × 2 tables depicting the benign vs. malignant diagnoses in GB-RADS categories 2, 3, 4, and 5. GB-RADS, gallbladder reporting and data system.
Discussion
In this retrospective multireader study evaluating the agreement for GB-RADS, we found weak to moderate intrareader agreement for GB-RADS findings and GB-RADS categories among all the readers. There was acceptable interreader agreement for GB-RADS categories. However, the interreader agreement was low for individual GB-RADS findings. The interreader agreement was better for cine-loops than static images. We found that, as originally proposed, there was an extremely low risk of malignancy in the GB-RADS 2 category and the highest risk of malignancy in the GB-RADS 5 category.
GBC is prevalent in specific geographic regions.1 GBC presents in three distinct morphological forms. The most common form is the mass-replacing gallbladder.9 This form is the easiest to diagnose—polypoidal and mural thickening types of GBC present diagnostic challenges.10,11 The Society of Radiologists in Ultrasound recently published their consensus conference recommendations for gallbladder polyps.12 The proposal allows objective classification of polyps into extremely low-risk, low-risk, and indeterminate risk on US.
Recently, an international consensus (GB-RADS) was also reported for risk stratification of non-acute GWT on US.4 This system stratifies GWT into six categories with a gradually increasing risk of malignancy based on key US findings.
However, the agreement for the key GB-RADS US findings and GB-RADS category is not previously reported. Moreover, the percentage risk of malignancy in each GB-RADS category has not been validated. Finally, the diagnostic performance of GB-RADS is not known. We performed this study to address these limitations of GB-RADS.
The diagnosis of benign GWT (GB-RADS 2) is based on the identification of layered appearance or intramural cysts/echogenic foci. The layered appearance of the GWT is defined as the visualization of both the inner and outer layers of the gallbladder.4 The layered appearance is usually seen in diffuse GWT. The layered appearance is lost in malignant GWT as the mucosa is disrupted, and tumor cells infiltrate deeper layers.13,14 There was weak to moderate intrareader agreement for this finding on static images. The intramural findings, including intramural cysts and echogenic foci, also favor benign GWT.4 These can be focal or diffuse. Intramural cysts appear as anechoic areas within the thickened gallbladder wall, represent enlarged Rokitansky Aschoff sinuses (RAS), and are considered pathognomonic of ADM.14 Intramural echogenic foci are secondary to the deposition of cholesterol crystals in the RAS.15 There was a weak intrareader agreement for intramural cysts and echogenic foci. Joo et al. reported a kappa value between 0.581 and 0.765 for the agreement among the three readers at HRUS for diagnosing ADM. Yoshimitsu et al. reported a kappa value 0.67 between the two readers for diagnosing ADM.16 Bang et al. reported an interreader agreement of 0.78 for diagnosing ADM at HRUS.17 We found a high specificity and PPV of GB-RADS 2 findings for diagnosing benign GWT. This is similar to previous studies that also reported high diagnostic accuracy of intramural changes for differentiating ADM from GBC.14, 15, 16
An indistinct gallbladder–liver interface may suggest infiltration by GBC.18 We found a high specificity and PPV of an indistinct liver interface (GB-RADS 4) for diagnosing malignant GWT. We found moderate intrareader for the gallbladder–liver interface.
A definite extramural extension is diagnosed based on the involvement of the bile ducts or vessels by the gallbladder lesion or by the presence of an extramural mass infiltrating the liver.4 These US findings make a diagnosis of GBC highly likely (GB-RADS 5).4 We found high PPV and NPV of these findings for diagnosing malignant GWT. There was moderate intrareader agreement for extramural extension.
The intrareader agreement was better for cine-loops than static images. This is because the cine-loops allow a more confident assessment of key US findings. Progressive training may also have influenced the interreader agreement as cine-loops were read after multiple readings of static images.18
Around 50% of patients in our study were categorized as GB-RADS 3 or 4 after the consensus, with a higher proportion of malignancy in these categories (37% and 71.1 % respectively). Although this could represent an inherent tendency for poor discrimination in GB-RADS 3 and 4, the percentage of malignancy is in line with the proposed risk stratification.4 Modifications in the criteria, especially for GB-RADS 3, may improve the discriminatory ability of this category.
There were a few limitations to our study. First, due to the retrospective nature, we did not evaluate the influence of scan-related variables or ancillary techniques (including harmonic imaging, HRUS, Doppler, and speckle reduction) on the intrareader and interreader agreement. Second, the initial consensus reading may have led to a higher interreader agreement. However, we introduced a washout period of 4 weeks and the images were given in different order to the radiologists to reduce this bias. Third, the cine-loops for only some randomly selected patients were read. This may have led to bias. However, we could demonstrate that in clinical practice where the sonologist performs and interprets a dynamic US examination, the interobserver agreement of GB-RADS will be higher than with static images. Finally, the diagnosis of benign disease in some patients was based on follow-up imaging rather than histopathology of the cholecystectomy specimen. The availability of the histological diagnosis in all cases would have provided better insight into the benign gallbladder diseases manifesting as GB-RADS 4 and 5 lesions.
In conclusion, GB-RADS has moderate intrareader and acceptable interreader agreement for GB-RADS categories. However, intrareader agreement is weak to moderate and interreader agreement is low for individual GB-RADS findings. Although, the risk of malignancy is negligible in GB-RADS 2 category and highest in GB-RADS 5 category, the discriminatory performance of GB-RADS 3 and 4 categories is lower. Larger multicenter studies with more readers must assess the reader agreement and validate the GB-RADS systems for wider clinical utilization.
CRediT authorship contribution statement
Raghuraman Soundararajan, Pavithra Subramanian, Pratyaksha Rana, Manika Chhabra, Shravya Singh, Ruby Siddiqui: Methodology, Writing – Original Draft, Formal analysis.
Pankaj Gupta – Conceptualization, Methodology, Writing – Original Draft, Writing – Review and Editing, Formal analysis.
Thakur Deen Yadav, Vikas Gupta, Lileswar Kaman, Harjeet Singh, Santosh Irrinki, Parikshaa Gupta, Uma Nahar Saikia, Ritambhra Nada, Usha Dutta, Manavjit Singh Sandhu – Methodology, Writing – Review and Editing.
Conflicts of interest
The authors have none to declare.
FUNDING
None.
Footnotes
Supplementary data to this article can be found online at https://doi.org/10.1016/j.jceh.2024.101393.
Appendix A. Supplementary data
The following are the supplementary data to this article.
References
- 1.Dutta U., Bush N., Kalsi D., Popli P., Kapoor V.K. Epidemiology of gallbladder cancer in India. Chin Clin Oncol. 2019;8:1–33. doi: 10.21037/cco.2019.08.03. [DOI] [PubMed] [Google Scholar]
- 2.Goetze T.O. Gallbladder carcinoma: Prognostic factors and therapeutic options. World J Gastroenterol. 2015;21:12211–12217. doi: 10.3748/wjg.v21.i43.12211. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3.Soundararajan R., Marodia Y., Gupta P., et al. Imaging patterns of wall thickening type of gallbladder cancer. Clin Exp Hepatol. 2022;8:255–266. doi: 10.5114/ceh.2022.122285. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4.Gupta P., Dutta U., Rana P., et al. Gallbladder reporting and data system (GB-RADS) for risk stratification of gallbladder wall thickening on ultrasonography: an international expert consensus. Abdom Radiol (NY) 2022;47:554–565. doi: 10.1007/s00261-021-03360-w. [DOI] [PubMed] [Google Scholar]
- 5.Cohen J.F., Korevaar D.A., Altman D.G., et al. STARD 2015 guidelines for reporting diagnostic accuracy studies: explanation and elaboration. BMJ Open. 2016;6 doi: 10.1136/bmjopen-2016-012799. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6.Gupta P., Basu S., Rana P., et al. Deep-learning enabled ultrasound based detection of gallbladder cancer in northern India: a prospective diagnostic study. Lancet Reg Health South East Asia. 2023 doi: 10.1016/j.lansea.2023.100279. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7.McHugh M.L. Interrater reliability: the kappa statistic. Biochem Med (Zagreb) 2012;22:276–282. [PMC free article] [PubMed] [Google Scholar]
- 8.Hayes A.F., Krippendorff K. Answering the call for a standard reliability measure for coding data. Commun Methods Meas. 2007;1:77–89. [Google Scholar]
- 9.Lopes Vendrami C., Magnetta M.J., Mittal P.K., Moreno C.C., Miller F.H. Gallbladder carcinoma and its differential diagnosis at MRI: what radiologists should know. Radiographics. 2021;41:78–95. doi: 10.1148/rg.2021200087. [DOI] [PubMed] [Google Scholar]
- 10.Kim J.H., Lee J.Y., Baek J.H., et al. High-resolution sonography for distinguishing neoplastic gallbladder polyps and staging gallbladder cancer. AJR Am J Roentgenol. 2015;204:W150–W159. doi: 10.2214/AJR.13.11992. [DOI] [PubMed] [Google Scholar]
- 11.Gupta P., Marodia Y., Bansal A., et al. Imaging-based algorithmic approach to gallbladder wall thickening. World J Gastroenterol. 2020;26:6163–6181. doi: 10.3748/wjg.v26.i40.6163. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12.Kamaya A., Fung C., Szpakowski J.L., et al. Management of incidentally detected gallbladder polyps: society of radiologists in ultrasound consensus conference recommendations. Radiology. 2022;305:277–289. doi: 10.1148/radiol.213079. [DOI] [PubMed] [Google Scholar]
- 13.Wibbenmeyer L.A., Sharafuddin M.J., Wolverson M.K., Heiberg E.V., Wade T.P., Shields J.B. Sonographic diagnosis of unsuspected gallbladder cancer: imaging findings in comparison with benign gallbladder conditions. AJR Am J Roentgenol. 1995;165:1169–1174. doi: 10.2214/ajr.165.5.7572497. [DOI] [PubMed] [Google Scholar]
- 14.Joo I., Lee J.Y., Kim J.H., et al. Differentiation of adenomyomatosis of the gallbladder from early-stage, wall-thickening-type gallbladder cancer using high-resolution ultrasound. Eur Radiol. 2013;23:730–738. doi: 10.1007/s00330-012-2641-9. [DOI] [PubMed] [Google Scholar]
- 15.Oh S.H., Han H.Y., Kim H.J. Comet tail artifact on ultrasonography: is it a reliable finding of benign gallbladder diseases? Ultrasonography. 2019;38:221–230. doi: 10.14366/usg.18029. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16.Yoshimitsu K., Honda H., Aibe H., et al. Radiologic diagnosis of adenomyomatosis of the gallbladder: comparative study among MRI, helical CT, and transabdominal US. J Comput Assist Tomogr. 2001;25:843–850. doi: 10.1097/00004728-200111000-00003. [DOI] [PubMed] [Google Scholar]
- 17.Bang S.H., Lee J.Y., Woo H., et al. Differentiating between adenomyomatosis and gallbladder cancer: revisiting a comparative study of high-resolution ultrasound, multidetector CT, and MR imaging. Korean J Radiol. 2014;15:226–234. doi: 10.3348/kjr.2014.15.2.226. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18.Seifert P., Maikowski I., Winkens T., et al. Ultrasound cine loop standard operating procedure for benign thyroid diseases-evaluation of non-physician application. Diagnostics (Basel) 2021;11:67. doi: 10.3390/diagnostics11010067. [DOI] [PMC free article] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.