Contrast-enhanced ultrasound time-intensity curve-assisted parametric imaging: a possible effective tool for differential diagnosis of thyroid nodules

Introduction

Thyroid cancer originating from follicular epithelial cells is the most common endocrine malignancy (1). The GLOBOCAN2020 database indicates that thyroid cancer ranks ninth in terms of global cancer incidence (2), with a yearly increasing trend (3). Ultrasound is the favored imaging technique for assessing thyroid nodules by analyzing features like their quantity, position, dimensions, and vascularization (4,5). The global rise in the use of thyroid ultrasound in recent years has resulted in increased tumor detection rates (6). Although most thyroid cancers have a good prognosis, some patients experience anxiety when suspicious nodules are detected, which may lead to overdiagnosis and unnecessary treatment (7). Accurate identification of whether thyroid nodules are benign or malignant is vital for directing patients towards suitable treatment options.

Due to overlapping ultrasound features, traditional ultrasound diagnosis of benign and malignant thyroid nodules may have an accuracy rate as low as 80.2% (8). Currently, the most precise and economical way to assess thyroid nodules is through ultrasound-guided fine needle aspiration (FNA) (9). However, the effectiveness of FNA for thyroid diagnosis is restricted by elements such as the operator’s expertise, the sampling path, and its invasiveness, leading to uncertainty in identifying thyroid malignancies that may be as high as 30% (10). Many countries do not widely perform FNA, resulting in reliance on ultrasound reports or other clinical examinations as references when formulating treatment plans (11).

In recent years, contrast-enhanced ultrasound (CEUS) technology has been developed rapidly and widely applied (12). CEUS involves injecting ultrasound contrast agents into peripheral veins to visualize microcirculatory perfusion of solid organs and observe nodules in real-time dynamically, including tiny feeding vessels. The varying vascular patterns of benign versus malignant thyroid nodules produce different CEUS images, providing a useful method for ultrasound doctors to tell them apart (13).

However, CEUS diagnosis requires doctors with a certain level of experience who specialize in CEUS. Junior doctors face certain difficulties in interpreting CEUS (14). This study investigates CEUS parametric imaging based on color coding and imaging according to the time of contrast agent arrival at target tissue post-injection and perfusion intensity, evaluating nodule perfusion patterns through qualitative parameters, accurately extracting dynamic contrast features to obtain more intuitive images, hoping to help identify lesion perfusion patterns, improve diagnostic accuracy, reduce subjectivity in contrast interpretation, and provide more objective evidence for disease differential diagnosis. Therefore, this study aims to evaluate the practicality of ultrasound contrast parametric imaging in determining if thyroid nodules are benign or malignant through qualitative parameter color-coded imaging, and compare whether it has auxiliary value for ultrasound physicians of different seniority levels in diagnosing suspicious thyroid nodules. We present this article in accordance with the STARD reporting checklist (available at https://gs.amegroups.com/article/view/10.21037/gs-2025-1-599/rc).

Methods

Data collection

This study retrospectively collected 195 thyroid nodules from 180 patients who underwent thyroid nodule CEUS and had further diagnostic puncture or surgery at Zhangzhou Affiliated Hospital of Fujian Medical University, Fujian Province, from October 2024 to March 2025. This study was conducted in accordance with the Declaration of Helsinki and its subsequent amendments. This retrospective study was approved by the Ethics Committee of Zhangzhou Affiliated Hospital of Fujian Medical University, Fujian Province (No. 2025LWB192), which waived the requirement for informed consent due to the retrospective nature of the study. The collected research content included patients’ general clinical information, conventional two-dimensional ultrasound of nodules, CEUS videos, and puncture or surgical pathological results. All participants provided written informed consent for CEUS and ultrasound-guided thyroid FNA.

Inclusion and exclusion criteria

Inclusion criteria: (I) nodules with complete clinical data, ultrasound examination results, and CEUS results; (II) nature of nodules confirmed by surgery or needle biopsy; (III) patients aged 18 years or older.

Exclusion criteria: (I) cases without surrounding thyroid tissue for comparison; (II) poor CEUS image quality unable to generate parametric imaging; (III) thyroid nodules with large calcification affecting contrast imaging. Based on a pilot study with an expected area under the curve (AUC) improvement of 0.10 and a power of 80% at α =0.05, a minimum of 140 nodules was required. To account for potential exclusions, we initially enrolled 195 nodules. According to inclusion and exclusion criteria, 134 cases (nodules: 146) were finally included, as detailed in Figure 1.

Figure 1 Flow diagram of included patients and number of thyroid nodules. CEUS, contrast-enhanced ultrasound; FNA, fine-needle aspiration; NEUS, nonenhanced ultrasound.

Examination method

Thyroid ultrasound examination was performed using a SIEMENS machine (model: ACUSON Sequoia Silver, Origin: Germany) equipped with a linear array probe (10L4) with working frequency of 4–10MHz. The contrast agent used was SonoVue^® (Bracco Milan, Italy). Patients were arranged in a supine position with their necks extended and entirely exposed. The mechanical index was set to 1.00, with lesion position slightly deeper than the target nodule (15). An ultrasound scan was employed to gather data on the position, dimensions, and blood circulation of thyroid nodules, as well as to examine cervical lymph nodes. A microbubble suspension was created by combining 5 mL of a 0.9% sodium chloride solution with SonoVue^® and agitating the mixture. With the probe held stationary, an entire image of the nodule and adjacent normal thyroid tissue was obtained, and patients were advised not to swallow. CEUS was used for dual-frame contrast imaging. 1.5 mL microbubble suspension was injected through the cubital vein, followed by 5 mL of 0.9% sodium chloride solution. A timer was started, and dynamic CEUS video of the entire process was stored in real-time for 120 seconds. All examinations were performed by senior physicians with attending physician or higher titles and more than 10 years of CEUS experience.

Image processing

This study utilized the independently developed TICTOOL CEUS analysis software: Time Intensity Curve Tool (v0.2.4 Beta, CET Beijing, China), to fit contrast time-intensity curves (TICs), calculate parameters related to the dynamic process of blood flow perfusion, encode parameter values into corresponding colors, and finally generate contrast perfusion time and intensity-related parametric images, achieving a perfusion imaging color coding method (Figure 2). This study uses different colors to represent the sequence of lesion and normal tissue perfusion arrival and the contrast of perfusion signal strength.

Figure 2 Parametric imaging generation using TICTOOL software. (A) Figure shows appropriate selection of target nodules and surrounding normal tissue to obtain regions of interest. (B) Signal arrival time sequence is represented using the color coding bar on the right side of the image. The color spectrum bar on the right side of Figure B represents the sequence of arrival time, with the color spectrum from red to orange, yellow, green, blue, and purple, sequentially representing signal arrival from early to late, and will display different colors in the ROI area. (C) Perfusion signal strength is also displayed in the ROI area using the same method. AUC, area under the curve; ROI, region of interest.

Image analysis and diagnostic criteria

Conventional ultrasound and CEUS examinations were performed on the thyroid, and conventional ultrasound and contrast videos were stored in DICOM format. Corresponding patient images and videos were imported into TICTOOL software for reading and analysis. Two-dimensional ultrasound was evaluated and diagnosed according to the “American College of Radiology Thyroid Imaging Reporting and Data System (ACR-TIRADS) Clinical Application Guidelines” published by the American College of Radiology for thyroid nodule examination. One junior physician (1–2 years of practice) and one senior physician (more than 10 years of practice) each performed ACR-TIRADS diagnostic classification of thyroid nodules in independent spaces without knowledge of pathological results. After randomizing the order 3 days later, they analyzed results by combining CEUS through the CEUS-Thyroid Imaging Reporting and Data System reporting system, recorded as junior CEUS-TIRADS and senior CEUS-TIRADS. Finally, cases were randomly selected and analyzed using ultrasound parametric imaging-assisted diagnostic system, recorded as junior CEUS-TIRADS + TIC and senior CEUS-TIRADS + TIC. After 2 weeks, senior and junior physicians randomly selected cases for evaluation, and the order of reading two-dimensional ultrasound, contrast video, and combined contrast parametric imaging assistance was randomly selected.

The ultrasound contrast scoring system was classified according to the thyroid CEUS reporting system (15), as shown in Table 1. On two-dimensional ultrasound, hypoechoic or very hypoechoic, aspect ratio greater than 1, irregular or lobulated margins, large calcification or marginal calcification, and presence of extrathyroidal invasion were malignant features, each scoring 1 point, with microcalcification scoring 2 points. In contrast mode, centripetal or centrifugal enhancement, hypoenhancement or hyperenhancement, and irregular ring enhancement were malignant features, each scoring 1 point. The total scores obtained from two-dimensional and contrast modes were added to obtain the corresponding total score and the corresponding malignancy probability of nodules.

Qualitative indicators for parametric analysis of thyroid target nodules include arrival time and perfusion intensity. The timing of arrival is classified based on when the contrast agent reaches the nodules compared to the surrounding normal thyroid tissue: before, at the same time as, or after the adjacent normal tissue. Perfusion intensity is based on contrast agent perfusion in nodules compared with surrounding normal thyroid tissue, classified as weaker than, equal to, and stronger than surrounding tissue, as detailed in Figure 3.

Figure 3 CEUS parametric imaging with histopathological correlation. (A-C) Benign nodule showing synchronous, isoenhancing perfusion (similar time-to-arrival and intensity to surrounding tissue), pathologically confirmed. (D) Histopathology of benign nodule. (E-G) Papillary thyroid carcinoma with delayed arrival, hypoenhancement, centripetal perfusion—malignant features. (H) Histopathology of papillary carcinoma. (I-K) Another case of papillary carcinoma demonstrating early central contrast arrival with centrifugal enhancement (center-to-periphery) and irregular rim enhancement. (L) Histopathology of papillary carcinoma. (M-O) Benign nodule with uniform peripheral ring enhancement and regular perfusion dynamics (P) Histopathology of benign nodule. AUC, area under the curve; CEUS, contrast-enhanced ultrasound; ROI, region of interest.

Statistical methods

SPSS 26.0 software was used for statistical processing. Quantitative data were expressed as x¯±s, and qualitative data were expressed as numbers or percentages. Using pathological results as the “gold standard”, receiver operating characteristic (ROC) curves were drawn, and Z-test was used to compare AUC of each group. According to different diagnostic malignancy cutoff values, sensitivity, specificity, Youden index, positive predictive value, negative predictive value, and diagnostic accuracy of each group were calculated. Kappa test was used to evaluate consistency of diagnostic results among physicians of the same seniority.

Results

General data

Among 146 thyroid nodules, pathology confirmed 72 malignant nodules (49.3%) and 74 benign nodules (50.7%). No indeterminate results were encountered in this study. All malignant nodules were confirmed by puncture biopsy or surgical pathology, all being papillary carcinomas. Benign nodules were confirmed by FNA biopsy cytological pathology, all being Bethesda II category.

Interpretation results of junior and senior physicians

Results of junior and senior physicians according to ACR-TIRADS, CEUS-TIRADS, and TIC-assisted diagnosis are shown in Table 2.

Diagnostic consistency analysis

Kappa tests for thyroid nodule classification judgment by physicians of different seniority showed good consistency (all P<0.001). Whether junior or senior physicians used contrast or combined contrast parametric imaging, classification diagnostic consistency was improved. See Table 3 for details.

Diagnostic efficacy analysis of different seniority physicians and combined TIC-assisted diagnostic system

Comparison results of diagnostic efficacy of thyroid nodule interpretation by physicians of different seniority are shown in Table 4 and Figure 4. The AUC of junior physicians’ diagnosis was lower than that of senior physicians (0.730 vs. 0.834), with statistically significant difference (P<0.001). After combining CEUS-assisted diagnosis, diagnostic efficacy of both junior and senior physicians improved significantly, but there were still differences in diagnostic efficacy between the two groups (P<0.001). Under contrast parametric imaging-assisted diagnosis, diagnostic efficacy of both junior and senior physicians improved significantly, with junior physicians showing more significant improvement, and no significant difference in diagnostic efficacy between junior and senior physicians for thyroid nodule diagnosis (0.898 vs. 0.937; P=0.07).

Figure 4 ROC curves for the diagnosis of thyroid cancer in different groups. ACR, American College of Radiology; CEUS, contrast-enhanced ultrasound; ROC, receiver operating characteristic; TIC, time-intensity curve.

The figure shows ROC curves comparing diagnostic performance of different approaches, demonstrating progressive improvement in AUC values with the addition of contrast enhancement and TIC assistance, particularly for junior physicians.

Discussion

Thyroid nodule ultrasound examination is a rapid procedure with monochromatic imaging. The recognition and interpretation of ultrasound features are highly subjective, influenced by the physician’s experience, as well as the objective environment and the physician’s subjective state at the time. This leads to variability in nodule classification and characterization both between different physicians and for the same physician at different times (16,17). Therefore, our study is based on previous research on peak time parametric imaging and TIC-based CEUS, with a focus on clinical validation and application rather than methodological innovation (18,19). The results of this study demonstrate that both junior and senior physicians achieved more consistent thyroid nodule classification when using CEUS and TIC parametric imaging tools. Moreover, the malignant risk stratification of nodules better aligned with diagnostic requirements.

As an advanced imaging technique, CEUS allows for real-time observation of microcirculation perfusion in lesions, improving microvascular detection and adding to the morphological data from standard ultrasound (20). Currently, researchers believe that papillary thyroid carcinoma (PTC) is primarily a hypovascular tumor, as the blood vessel formation inside the tumor is often not fully developed. As the tumor grows invasively, it compresses blood vessels, resulting in varying levels of vascular collapse or damage (21). Therefore, our study builds upon previous research on time-to-peak parametric imaging and TIC-based CEUS, focusing on clinical validation and application rather than methodological innovation. CEUS enhancement in PTC often exhibits delayed enhancement compared to surrounding thyroid tissue. Studies on CEUS for thyroid nodules, both domestically and internationally, have emphasized key malignant perfusion features such as “slow perfusion”, “heterogeneous hypo-enhancement”, “centripetal enhancement”, “extrathyroidal extension”, and “irregular peripheral rim-like enhancement” (22). However, in small nodules, the human eye may struggle to capture arterial-phase perfusion characteristics, potentially leading to oversight of critical diagnostic information. Additionally, the technique heavily relies on the operator’s subjective judgment, limiting its diagnostic efficacy.

To address these limitations, this study utilized the self-developed software TICTOOL to compare qualitative features in 146 thyroid nodules, accurately extracting parameters such as regional time-to-peak and perfusion intensity. This approach allows for intuitive assessment of enhancement patterns, offering a simpler and more straightforward evaluation compared to dynamic CEUS. The results indicate that diagnostic features such as hypo-enhancement with delayed peak time relative to surrounding thyroid tissue, centripetal or centrifugal enhancement, and irregular rim-like enhancement provide significant diagnostic value for malignant thyroid nodules (23).

In this study, senior radiologists demonstrated significantly higher intra-observer consistency in ACR-TIRADS classification compared to junior radiologists when re-evaluating cases after two weeks (Kappa: 0.690 vs. 0.555), indicating more stable morphological assessment by experienced practitioners. After implementing parametric imaging with contrast enhancement, both junior and senior radiologists showed significantly improved diagnostic consistency across different time points (junior Kappa increased to 0.709; senior to 0.745). This suggests that parametric imaging provides objective diagnostic parameters (e.g., perfusion intensity, time-to-peak) that compensate for the experience gap among junior radiologists. Notably, parametric imaging converts dynamic contrast-enhanced videos into standardized 2D parametric maps, offering a more efficient and accurate interpretation method than real-time video analysis. Our findings support incorporating dynamic CEUS with TIC analysis into thyroid nodule diagnostic workflows, particularly in primary hospitals or during training, as it may standardize interpretation and reduce experience-related variability. However, senior radiologists still maintained slightly better performance when using all modalities combined (Kappa 0.745 vs. 0.709; AUC: 0.937 vs. 0.898), underscoring that experience remains an irreplaceable factor requiring continuous training for optimal diagnostic performance.

Furthermore, junior radiologists showed significantly lower diagnostic efficacy using conventional ultrasound compared to seniors (AUC 0.730 vs. 0.834, P<0.05), consistent with Ye FY et al.’s report of generally inferior performance among less-experienced practitioners (24). When applying the CEUS-TIRADS system, both groups showed improved AUC values (junior: 0.730 vs. 0.784; senior: 0.832 vs. 0.890), attributable to CEUS detecting additional malignant features invisible on conventional ultrasound, aligning with prior studies (15). However, due to limited experience in interpreting contrast dynamics, juniors benefited less from conventional CEUS than seniors did. Parametric imaging assistance significantly enhanced diagnostic performance for both groups (all P<0.05), with juniors showing greater improvement - particularly in specificity (37.8% vs. 62.2%). Based on the improved specificity achieved with TIC-assisted parametric imaging, we estimate that unnecessary FNA procedures could be reduced by approximately 24.4% in clinical practice. This suggests that the integration of parametric imaging may meaningfully impact clinical decision-making by reducing invasive procedures in patients with benign nodules.

The perfusion intensity provided in this study represents the cumulative overall perfusion intensity of nodules during the enhancement phase. For malignant thyroid nodules exhibiting hypo-enhancement, centripetal enhancement, or slow enhancement, the imaging characteristics demonstrate progressively weaker perfusion signal intensity toward the lesion center, allowing for more accurate assessment. By combining perfusion intensity with time-to-peak parameters, the dual-parametric imaging analysis provides a more comprehensive evaluation of perfusion patterns. The parametric imaging in this study offers qualitative parameters that assist sonographers in identifying perfusion features difficult to discern visually, thereby compensating for the diagnostic inefficiency of junior radiologists due to limited experience. This approach reduces inter-observer variability in CEUS pattern analysis, ultimately improving diagnostic objectivity and reliability.

In conclusion, the application of our self-developed TICTOOL tool significantly enhanced both the diagnostic accuracy and interobserver consistency among physicians with varying experience levels in evaluating thyroid nodules. This advancement holds substantial clinical significance as it enables more precise diagnosis of suspicious nodules, thereby potentially reducing unnecessary fine-needle aspiration procedures. Particularly for junior physicians with limited diagnostic experience in thyroid nodules, this tool provides crucial decision-support value, especially in the interpretation of CEUS findings.

Study limitations

This study has the following limitations:

• Retrospective design with small sample size, and all malignant cases were papillary carcinomas, which may significantly limits the generalizability of our findings to other thyroid cancer subtypes, such as follicular, medullary, or anaplastic carcinomas.
• Perfusion phase analysis was limited to the arterial phase and did not encompass later perfusion stages, which might provide additional diagnostic information.
• The study was conducted at a single center, and multi-center validation is needed to confirm the broader applicability of the TIC-assisted diagnostic system.
• Long-term follow-up data were not available to assess the clinical impact of improved diagnostic accuracy on patient outcomes.

Conclusions

The parametric imaging-assisted diagnostic system based on CEUS demonstrates high application value in the differential diagnosis of benign and malignant thyroid nodules. Combined use can significantly improve the diagnostic efficacy of physicians with different levels of experience, particularly benefiting junior physicians by reducing inter-observer variability and improving diagnostic consistency. This technology shows promise for standardizing thyroid nodule evaluation and may be particularly valuable in clinical settings with limited access to experienced radiologists.

Acknowledgments

We extend our heartfelt thanks to our colleagues in the Department of Ultrasound for their support and collaboration during this research. Their expertise, feedback, and encouragement have been instrumental in the completion of this paper. It has been a privilege to work with such a dedicated and inspiring team.

Footnote

Reporting Checklist: The authors have completed the STARD reporting checklist. Available at https://gs.amegroups.com/article/view/10.21037/gs-2025-1-599/rc

Data Sharing Statement: Available at https://gs.amegroups.com/article/view/10.21037/gs-2025-1-599/dss

Peer Review File: Available at https://gs.amegroups.com/article/view/10.21037/gs-2025-1-599/prf

Funding: This work was supported by the Natural Science Foundation of Fujian Province, China (No. 2024J011562).

Conflicts of Interest: All authors have completed the ICMJE uniform disclosure form (available at https://gs.amegroups.com/article/view/10.21037/gs-2025-1-599/coif). The authors have no conflicts of interest to declare.

Ethical Statement: The authors are accountable for all aspects of the work in ensuring that questions related to the accuracy or integrity of any part of the work are appropriately investigated and resolved. The study was conducted in accordance with the Declaration of Helsinki and its subsequent amendments, and was approved by the Ethics Committee of Zhangzhou Affiliated Hospital of Fujian Medical University (No. 2025LWB192). All personal identifiers were removed from the data to ensure participant confidentiality. The need for written informed consent was waived due to the retrospective nature of the study.

Open Access Statement: This is an Open Access article distributed in accordance with the Creative Commons Attribution-NonCommercial-NoDerivs 4.0 International License (CC BY-NC-ND 4.0), which permits the non-commercial replication and distribution of the article with the strict proviso that no changes or edits are made and the original work is properly cited (including links to both the formal publication through the relevant DOI and the license). See: https://creativecommons.org/licenses/by-nc-nd/4.0/.

References

1. Wang C, Wu Z, Lei L, et al. Geographic disparities in trends of thyroid cancer incidence and mortality from 1990 to 2019 and a projection to 2030 across income-classified countries and territories. J Glob Health 2023;13:04108. [Crossref] [PubMed]
2. Pizzato M, Li M, Vignat J, et al. The epidemiological landscape of thyroid cancer worldwide: GLOBOCAN estimates for incidence and mortality rates in 2020. Lancet Diabetes Endocrinol 2022;10:264-72. [Crossref] [PubMed]
3. Song SS, Huang S, Park S. Association of Polygenetic Risk Scores Related to Cell Differentiation and Inflammation with Thyroid Cancer Risk and Genetic Interaction with Dietary Intake. Cancers (Basel) 2021;13:1510. [Crossref] [PubMed]
4. Solymosi T, Hegedűs L, Bonnema SJ, et al. Considerable interobserver variation calls for unambiguous definitions of thyroid nodule ultrasound characteristics. Eur Thyroid J 2023;12:e220134. [Crossref] [PubMed]
5. Zhang D, Wang XN, Jiang L, et al. Conventional ultrasonography and elastosonography in diagnosis of malignant thyroid nodules: A systematic review and meta-analysis. Front Endocrinol (Lausanne) 2022;13:1082881. [Crossref] [PubMed]
6. Xu L, Cao ZX, Weng X, et al. Global thyroid cancer incidence trend and age-period-cohort model analysis based on Global Burden of Disease Study from 1990 to 2019. Front Endocrinol (Lausanne) 2023;14:1133098. [Crossref] [PubMed]
7. Giovanella L, Garo ML, Campenní A, et al. Thyroid Hormone Withdrawal versus Recombinant Human TSH as Preparation for I-131 Therapy in Patients with Metastatic Thyroid Cancer: A Systematic Review and Meta-Analysis. Cancers (Basel) 2023;15:2510. [Crossref] [PubMed]
8. Huang P, Zheng B, Li M, et al. The Diagnostic Value of Artificial Intelligence Ultrasound S-Detect Technology for Thyroid Nodules. Comput Intell Neurosci 2022;2022:3656572. [Crossref] [PubMed]
9. Nicholson KJ, Yip L. An update on the status of molecular testing for the indeterminate thyroid nodule and risk stratification of differentiated thyroid cancer. Curr Opin Oncol 2018;30:8-15. [Crossref] [PubMed]
10. Bao G, Chang C, Yin A. PSAT267 Performance of ThyroSeq V3 molecular testing in assessing indeterminate thyroid nodules for thyroid cancer at an urban endocrinology clinic. J Endocr Soc 2022;6:A809-10.
11. Trimboli P, Castellana M, Virili C, et al. Performance of contrast-enhanced ultrasound (CEUS) in assessing thyroid nodules: a systematic review and meta-analysis using histological standard of reference. Radiol Med 2020;125:406-15. [Crossref] [PubMed]
12. Ou Q, Xie W, Yu Y, et al. Contrast-enhanced ultrasound enables precision diagnosis of preoperative muscle invasion in bladder cancer: a prospective study. MedComm (2020) 2025;6:e70106.
13. Yang YP, Zhang GL, Zhou HL, et al. Diagnostic efficacy of the contrast-enhanced ultrasound thyroid imaging reporting and data system classification for benign and malignant thyroid nodules. Quant Imaging Med Surg 2024;14:5721-36. [Crossref] [PubMed]
14. Chi J, Chen JH, Wu B, et al. A Dual-Branch Cross-Modality-Attention Network for Thyroid Nodule Diagnosis Based on Ultrasound Images and Contrast-Enhanced Ultrasound Videos. IEEE J Biomed Health Inform 2025;29:1269-82. [Crossref] [PubMed]
15. Ruan J, Xu X, Cai Y, et al. A Practical CEUS Thyroid Reporting System for Thyroid Nodules. Radiology 2022;305:149-59. [Crossref] [PubMed]
16. Jiang N, Feng QQ, Li Y, et al. Arrival-Time Parametric Imaging in Contrast-Enhanced Ultrasound for Thyroid Nodule Differentiation. Med Sci Monit 2024;30:e945793. [Crossref] [PubMed]
17. Kim GR, Lee E, Kim HR, et al. Convolutional Neural Network to Stratify the Malignancy Risk of Thyroid Nodules: Diagnostic Performance Compared with the American College of Radiology Thyroid Imaging Reporting and Data System Implemented by Experienced Radiologists. AJNR Am J Neuroradiol 2021;42:1513-9. [Crossref] [PubMed]
18. He L, Li Y, Zhong D, et al. Enhancing CEUS Value in Papillary Thyroid Carcinoma: A Dual-Model Framework for Patient Selection Based on Conventional Ultrasound Characteristics. Ultrasound Med Biol 2026;52:714-20. [Crossref] [PubMed]
19. Zhan J, Li F, Ma J, et al. Contrast Enhanced Ultrasound: A Quantitative Ultrasound Imaging Mode for the Differential Diagnosis of Benign and Malignant Thyroid Nodules. J Clin Ultrasound 2025;53:1504-12. [Crossref] [PubMed]
20. Lu J, Zhou P, Jin C, et al. Diagnostic Value of Contrast-Enhanced Ultrasonography With SonoVue in the Differentiation of Benign and Malignant Breast Lesions: A Meta-Analysis. Technol Cancer Res Treat 2020;19:1533033820971583. [Crossref] [PubMed]
21. Zhang J, Chen C, Yang Y, et al. Surgical resection following therapy with anlotinib in locally advanced papillary thyroid carcinoma: a case description. Quant Imaging Med Surg 2023;13:5456-62. [Crossref] [PubMed]
22. Zhang Y, Zhang X, Li J, et al. Contrast-enhanced ultrasound: a valuable modality for extracapsular extension assessment in papillary thyroid cancer. Eur Radiol 2021;31:4568-75. [Crossref] [PubMed]
23. Liu Y, Liu H, Zhan J, et al. Contrast-Enhanced Ultrasound for Diagnosing Thyroid Nodules With Indeterminate Cytology: A Retrospective Study. Clin Endocrinol (Oxf) 2025;102:223-31. [Crossref] [PubMed]
24. Ye FY, Lyu GR, Li SQ, et al. Diagnostic Performance of Ultrasound Computer-Aided Diagnosis Software Compared with That of Radiologists with Different Levels of Expertise for Thyroid Malignancy: A Multicenter Prospective Study. Ultrasound Med Biol 2021;47:114-24. [Crossref] [PubMed]