An improved diagnostic criterion based on Node-RADS MRI score for lymph node metastasis in papillary thyroid carcinoma

Introduction

Papillary thyroid carcinoma (PTC) accounts for approximately 90% of thyroid cancers and generally has a good prognosis (1,2). However, lymph node (LN) metastasis remains a frequent occurrence, affecting up to 60–70% of PTC patients and serving as a significant risk factor for loco-regional recurrence and cancer-specific mortality (3-5). Prophylactic neck dissection is not recommended, considering its complications and limited clinical benefit. Therapeutic neck dissection should be performed only for clinically diagnosed or biopsy-proven metastatic LNs (6). Therefore, accurate assessment of LN metastasis prior to surgery is of great importance to clinical management.

Ultrasonography (US) is commonly utilized as the primary imaging modality for noninvasive assessment of LN metastasis in PTC (7-9). However, its diagnostic accuracy is frequently compromised by operator-dependent variability and challenges in visualizing specific regions, particularly levels VI and VII (10,11). In contrast, cross-sectional imaging techniques including computed tomography (CT) and magnetic resonance imaging (MRI) are less operator-dependent than US and provide more comprehensive anatomical details regarding nodal locations relative to surgical landmarks (8,10,12,13). This allows clinicians to visualize structures and abnormalities clearly before performing surgical procedures. However, there is no consensus on CT or MRI criteria for diagnosing LN metastasis in PTC, although nodal size is generally used.

The structured oncological imaging reports have been increasingly used in recent years, leading to widespread implementation of Reporting and Data Systems (RADS) across multiple organs, including the breast (BI-RADS), prostate (PI-RADS), and liver (LI-RADS) (14-16). This development standardizes the communication of imaging findings and ensures consistent reporting among radiologists. Node-RADS 1.0 was initially introduced in 2021 for the systematic assessment of LNs in cancer, aiming to categorize the level of suspicion for LN metastasis through a comprehensive set of criteria (17). The scoring system can be assessed on CT or MRI based on nodal size and configuration criteria (texture, border, and shape). To be adapted to thyroid cancer, the system also identifies “entity-specific findings”. If cystic appearance or calcification is found in LNs of thyroid cancer patients, 3 points can be given in the “texture” category. The levels of suspicion for metastasis range from NR-1 (very low) to NR-5 (very high) (17). Some studies have been conducted on the role of Node-RADS in patients with malignant tumors in various organs (18-22), showing a high specificity (97.1–100%) but a very low sensitivity (16.7–48.6%). Its application in assessing LN metastasis in PTC remains unexplored yet. In addition to Node-RADS, some additional or supplementary MRI features, including T1 hyperintensity, nodal clustering, and restricted diffusion on diffusion-weighted imaging (DWI), have demonstrated potential in enhancing the diagnostic accuracy for LN metastasis in PTC (23-25).

Therefore, the objectives of this study were (I) to evaluate the performance of Node-RADS with MRI for the diagnosis of LN metastasis in PTC; and (II) to explore strategies to further improve the Node-RADS scoring, so that it can be used for clinical routine staging in PTC. We present this article in accordance with the STARD reporting checklist (available at https://qims.amegroups.com/article/view/10.21037/qims-2025-740/rc).

Methods

This prospective study was approved by the Institutional Review Board and Ethics Committee of Zhongshan Hospital (Xiamen), Fudan University (No. B2020-002). The study was conducted in accordance with the principles of the Declaration of Helsinki and its subsequent amendments. Written informed consent was obtained from all participants.

Study population

From August 2021 to September 2023, patients with suspected PTC scheduled for surgery were enrolled consecutively in Zhongshan Hospital (Xiamen), Fudan University according to the following inclusion criteria: (I) patients who were suspected of having PTC by routine thyroid ultrasound; (II) patients who had no prior treatment or biopsy; (III) who had no history of head and neck cancer. A total of 116 consecutive patients were included for thyroid contrast-enhanced MRI examination. Exclusion criteria were as follows: (I) patients who did not undergo subsequent surgery in our hospital; (II) who had prior treatment or biopsy; (III) patients whose MR images were inadequate to evaluate due to severe artifact; and (IV) those with no visible LNs on MR images. A total of 82 patients with 156 LNs were finally enrolled in the study (Figure 1).

Figure 1 Flowchart of the study population. MRI, magnetic resonance imaging; PTC, papillary thyroid carcinoma.

This study was designed to evaluate the diagnostic performance of the standard Node-RADS MRI score in detecting LN metastasis in PTC and to investigate the improved accuracy of a novel diagnostic criterion. Thus, the sample size was calculated based on the method for comparing areas under the receiver operating characteristic (ROC) curves. Given the previously reported (20) area under the curve (AUC) of the standard Node-RADS (0.583) and the expected sensitivity of the new criterion (approximately 0.900), a minimum sample size of 38 LNs was required to achieve a statistical power of 80% (1 − β =0.80) with a two-sided significance level of 0.05 (α=0.05). The calculation was performed using MedCalc 23.1.6 (Mariakerke, Belgium) following established guidelines for diagnostic test evaluation (26). Ultimately, a total of 82 patients with 156 LNs were enrolled in this study, fulfilling the required sample size to ensure reliable and generalizable results.

Imaging examination

All patients underwent neck contrast-enhanced MRI examinations using the 3.0T Discovery MR 750w (GE Healthcare, Milwaukee, WI, USA). The scanning protocol was in accordance with the American Thyroid Association (ATA) Statement on Preoperative Imaging for Thyroid Cancer Surgery (27) and technical considerations in Node-RADS (17). Although not strictly required by Node-RADS, contrast-enhanced MRI is still advantageous for evaluating “size” and “configuration” and is critical for tumor staging and treatment monitoring (28,29). Thus, Gd-DTPA (Magnevist; Bayer Schering Pharma AG, Berlin, Germany) was used as the contrast agent for enhanced MRI. A detailed overview of MR acquisition parameters is summarized in Table S1.

Preoperative localization and labeling

All patients were histopathologically diagnosed as PTC through surgical resection and underwent lymphadenectomy at our institution. Regional LNs were all resected and categorized into groups according to the N staging standard of the eighth edition of the American Joint Committee on Cancer (AJCC) Tumor-Node-Metastasis (TNM) staging system for PTC (30). The surgeon (X.L.; with 9 years of experience in head and neck surgery), who had previously identified the visible LNs in each group (levels II, III, IV, etc.) on MR images, carefully labeled the same largest LNs during surgery. For intraoperative identification, LN size, location, and distance from the tumor or anatomic landmarks were used for reference (31,32). All the identified LNs on MRI were also labelled by the same surgeon, in preparation for evaluation of the Node-RADS score by radiologists.

Node-RADS and other MRI features evaluation

All the images were assessed independently by 3 radiologists (J.Z., Q.T., and M.W.), with 30, 9, and 6 years of experience in head and neck imaging, respectively, who were blinded to the clinical history, imaging report, and pathological information of all observations. The three readers evaluated all the identified LNs according to Node-RADS scheme, including size, texture, border, and shape (17). Categories were scored from 1 to 5, reflecting the level of suspicion for malignancy: 1 (very low), 2 (low), 3 (equivocal), 4 (high), and 5 (very high). Each radiologist was trained by evaluating more than 30 patients prior to this formal review. This study assessed Node-RADS on a “per lymph node” basis to enable a direct node-for-node correlation with the histopathological reference standard. Although thyroid cancer staging (N1a vs. N1b) is determined by the location of metastatic nodes rather than their number, and clinical practice relies on the most suspicious node per station (33), our aim was specifically to validate the diagnostic accuracy of the Node-RADS criteria itself, which was distinct from the goal of patient-level staging.

According to previous studies, supplementary MRI features that possibly referred to metastasis included: (I) T1 hyperintensity; or (II) nodal clustering; or (III) restricted diffusion on DWI (23-25). T1 hyperintensity is defined as the presence of high signal on pre-contrast T1-weighted images with fat saturation. Nodal clustering is defined as the presence of three or more LNs in one group (23). DWI was processed on the GE AW4.6 Workstation (GE Healthcare) to generate apparent diffusion coefficient (ADC) map. A circular region of interest (ROI) was manually delineated on the darkest part of the LN on ADC map, avoiding necrotic, cystic, and hemorrhagic areas. The same reader measured the lesion three consecutive times and recorded the mean ADC value of the ROI. Restricted diffusion on DWI is characterized by an ADC value of the LN that falls below the optimal cutoff value in ROC curve analysis for identifying metastasis.

Histopathological analysis

Operative specimens of the LNs were collected by a pathologist with 8 years of experience. All specimens were fixed in formalin and embedded in paraffin. The section of each specimen was stained with hematoxylin and eosin (HE). Histopathological analysis was performed using a microscope (Leica, Wetzlar, Germany) by the same pathologist without knowing the MRI findings.

Statistical analysis

The percentage of LN metastasis in each category (i.e., NR-1 to NR-5) and each imaging feature (i.e., size, texture, border, and shape) was calculated. The diagnostic performance of the Node-RADS categories for LN metastasis was described by the sensitivity, specificity, accuracy, and positive and negative predictive values with 95% confidence intervals (CIs). Diagnostic performances according to different criteria were compared using the McNemar test. Student’s t-test was used for continuous variables, and the chi-squared test or Fisher’s exact test was used for categorical variables. Factors with a P value of less than 0.05 were enrolled into the multivariate logistic regression analysis to explore independent risk factors of metastasis. Odds ratios and 95% CIs were calculated. ROC curve analysis was performed to evaluate the diagnostic effectiveness. The AUC and optimal cutoff value were calculated. The difference between ROC curves was compared using the DeLong method. The interobserver agreement for qualitative and quantitative measurements between three readers was assessed by Kendall W coefficient with the following criteria: <0.20, slight; 0.21–0.40, fair; 0.41–0.60, moderate; 0.61–0.80, substantial; and 0.81–1.00, perfect. A 2-sided P value of less than 0.05 was considered statistically significant. All statistical analyses were performed using SPSS 22.0 (IBM Corp., Armonk, NY, USA) and MedCalc 23.1.6.

Results

Patient characteristics

A total of 82 patients (24 males and 58 females; age: 41±12 years) with 156 LNs were finally enrolled in the study. According to the eighth edition of AJCC TNM staging system (30), 74 of the 82 PTCs were stage I (90.3%), 7 were stage II (8.5%), and 1 was stage IV (1.2%). The distribution of pathological T classification was 48 (58.5%) with T1, 21 (25.6%) with T2, 10 (12.2%) with T3, and 3 (3.7%) with T4. The distribution of pathological N classification was 12 (14.6%) with N0, 23 (28.1%) with N1a, and 47 (57.3%) with N1b. Only one (1.2%) PTC was classified as M1, the rest (98.8%) were classified as M0 in clinical M classification (Table 1).

Interobserver agreement

Overall agreement between three readers was excellent for Node-RADS categories (W=0.978, P<0.001), size (W=0.970, P<0.001), texture (W=0.972, P<0.001), border (W=0.911, P<0.001), shape (W=0.962, P<0.001), T1 hyperintensity (W=0.925, P<0.001), nodal clustering (W=0.992, P<0.001), and ADC values (W=0.859, P<0.001). Thus, the following results are based on Reader 1.

Node-RADS categories

The final pathologic diagnoses of 156 lesions were 109 metastatic LNs and 47 non-metastatic LNs (Table 2). Metastatic LNs were diagnosed in 8 (32.0%) of the 25 NR-1, 13 (41.9%) of 31 NR-2, 32 (82.1%) of 39 NR-3, 33 (91.7%) of 36 NR-4, and 23 (92.0%) of 25 NR-5 (Figures 2,3). Texture and border were statistically different between metastatic and non-metastatic LNs (both P<0.001). LN size and shape had no significant difference between the metastasis and non-metastasis groups (P=0.145 and P=0.687, respectively).

Figure 2 A 49-year-old female with an enlarged lymph node in the right VI compartment (arrow) with a short-axis diameter of 7 mm, heterogeneous texture, irregular border and oval shape, which was categorized as Node-RADS 3, with T1 hyperintensity reinforcing the probability of metastasis. Final diagnosis was metastasis. (A) Fat-saturated axial T2-weighted image. (B) Fat-saturated axial T1-weighted image. (C) Fat-saturated axial enhanced T1-weighted image. Node-RADS, Node Reporting and Data System.

Figure 3 A 33-year-old female with an enlarged lymph node in the left IV compartment (white arrows) with a short-axis diameter of 8 mm, cystic appearance, smooth border and oval shape, which was categorized as Node-RADS 4, with T1 hyperintensity reinforcing the probability of metastasis. Final diagnosis was metastasis. (A) Fat-saturated axial T2-weighted image. (B) Fat-saturated axial T1-weighted image. (C) Fat-saturated axial enhanced T1-weighted image. Note the left internal jugular vein was compressed by the lymph node on T1-weighted image and T2-weighted image, but was full and round after injection of contrast agent (yellow arrows). Node-RADS, Node Reporting and Data System.

Diagnostic performances of Node-RADS and the new diagnostic criterion

Among the evaluated thresholds for Node-RADS, a score of ≥3 demonstrated the highest diagnostic performance, achieving an AUC of 0.776 (95% CI: 0.702 to 0.839), a sensitivity of 80.7%, and a specificity of 74.5%. In comparison, a threshold of ≥4 yielded an AUC of 0.704 (95% CI: 0.625 to 0.774), a sensitivity of 51.4%, and a specificity of 89.4%. A threshold of ≥5 resulted in an AUC of 0.584 (95% CI: 0.503 to 0.662), a sensitivity of 21.2%, and a specificity of 95.7%. The superior performance of Node-RADS score ≥3 threshold compared to Node-RADS score ≥4 and Node-RADS score ≥5 was statistically significant (P=0.034; P<0.001, respectively).

For supplementary MRI features, 86.6% (84 of 97) of LNs with T1 hyperintensity, and 86.8% (33 of 38) with nodal clustering were metastatic, which were all statistically significant (P<0.001; P=0.009, respectively). The ADC values of the metastasis group were significantly lower than those of the non-metastasis group (1.54±0.47 vs. 1.70±0.42; P=0.038) (Table 3). The optimal cutoff values of ADC in identifying metastasis were 1.26×10⁻³ mm²/s (sensitivity 30.3%, specificity 87.2%). Therefore, Node-RADS score ≥3, T1 hyperintensity, nodal clustering, and ADC value ≤1.26×10⁻³ mm²/s were finally included in the multivariate analyses.

In the multivariate analysis, only the Node-RADS score ≥3 (odds ratio: 9.013; 95% CI: 3.651 to 22.248; P<0.001) and T1 hyperintensity (odds ratio: 6.736; 95% CI: 2.748 to 16.510; P<0.001) were independent risk predictors for metastasis. The combined predictor of Node-RADS score and T1 hyperintensity was generated from the logistic regression equation. The new criterion showed a better diagnostic performance (sensitivity 75.2%; specificity 87.2%; AUC =0.856) than Node-RADS alone (sensitivity 80.7%; specificity 74.5%; AUC =0.776) (P=0.003). Table 4 and Figure 4 summarize the diagnostic performance of Node-RADS and the new criterion incorporating Node-RADS and T1 hyperintensity.

Figure 4 ROC curves of the diagnostic performance of Node-RADS and the new criterion incorporating Node-RADS and T1 hyperintensity, with AUCs of 0.776 and 0.856, respectively. AUC, area under the curve; Node-RADS, Node Reporting and Data System; ROC, receiver operating characteristic.

Discussion

The study demonstrated that the Node-RADS MRI score was valuable in clinical evaluation of LN metastasis in PTC with excellent reproducibility and good diagnostic performance. By combining T1 hyperintensity with Node-RADS, the new criterion showed a better diagnostic performance than Node-RADS alone.

Node-RADS was initially proposed to meet the necessity to standardize reporting of cancer involvement of LNs at any anatomical site based on CT and MRI (17). According to the Node-RADS guidelines, categories 1 and 2 should be interpreted as N(−), while categories 4 and 5 should be interpreted as N(+). The reporting of category 3 requires tailored consideration based on the primary tumor’s stage and histological grade. Multiple studies have demonstrated that Node-RADS shows promising diagnostic performance for both cutoff values of ≥3 and ≥4 (34). Although a Node-RADS score of 4 can identify more true positive cases, a score of 3 demonstrates higher accuracy in recognizing true negative cases. Consistently, Node-RADS score ≥3 showed the highest diagnostic performance in our study, in comparison to Node-RADS ≥4 and Node-RADS ≥5. Therefore, a Node-RADS score of 3 may represent the most appropriate cutoff for PTC LN metastasis.

In contrast to the study by Yu et al. (35), which utilized CT-based Node-RADS for assessing lateral cervical LNs in thyroid cancer (AUC =0.602), our investigation demonstrated that Node-RADS when applied to MRI achieved a superior predictive performance (AUC =0.776). This comparison suggests a possible advantage of MRI over CT for this specific application. We believe that the higher soft tissue resolution of MRI was pivotal, as it allows for a more detailed characterization of LN architecture. This includes superior delineation of internal features such as nodal texture, necrosis, and margins, all of which are critical criteria in the Node-RADS. Consequently, MRI likely enables a more accurate and confident assignment of Node-RADS categories.

Building upon the inherent strengths of MRI, we explored the value of supplementary MRI features, including T1 hyperintensity, nodal clustering, and DWI findings, in improving diagnostic accuracy. T1 hyperintensity emerged as a particularly promising predictor of metastasis, which is likely attributed to the high thyroglobulin or colloid content in thyroid cancer metastases (24). Thyroglobulin is a specific glycoprotein produced only by PTC cells or normal thyroid follicular cells. Thyroglobulin in LN fine needle aspiration washout has also been performed to confirm cervical LNs suspected to be metastases from PTC (36). Colloid production is also a common phenomenon both in primary and metastatic PTC (37). Node-RADS combined with T1 hyperintensity significantly surpassed the performance of the Node-RADS alone, achieving an AUC of 0.856, with 75.2% sensitivity and 87.2% specificity. Therefore, we suggest including T1 hyperintensity as a Node-RADS “entity specific finding” of texture in thyroid tumors in the future.

This study’s findings support the broader application of Node-RADS, as recently summarized in the comprehensive review by Parillo et al. (38). As noted therein, the diagnostic performance of Node-RADS exhibits considerable heterogeneity across different primary tumors and anatomic regions, with reported AUC values ranging from 0.59 (pelvic LNs in prostate cancer) to 0.97 (locoregional LNs in breast cancer). The diagnostic performance of our proposed MRI-based new criterion (AUC: 0.856) ranks high among the reported tumors, underscoring its potential utility specifically for PTC. Regarding inter-reader reliability, the same review indicates that agreement levels for Node-RADS scoring vary from fair to almost perfect across studies, with LN border and shape evaluation consistently being identified as major sources of discordance. Our results directly corroborate this finding. While the overall interobserver agreement in our study was excellent for final Node-RADS categories, agreement on the “border” criterion was relatively lower. This may be attributed to the inherent subjectivity in differentiating between “irregular” and “smooth” margins on imaging.

Future efforts to optimize and standardize the definitions of these specific morphological features could further enhance the reproducibility of the Node-RADS system. Therefore, multicenter studies with larger sample sizes are warranted to validate our promising results and to solidify the role of Node-RADS in PTC clinical practice.

There are some limitations in this study. Firstly, the relatively small sample size and the high level of expertise required for radiological-pathological correlation represent limitations of the study. These findings suggest that larger, multicenter studies are needed to validate this diagnostic criterion. Secondly, although we did our best to perform radiological-pathological matching of LNs, mismatches may still exist. Finally, microcalcification in metastatic LNs in PTC was sometimes equivocal on MRI, which might have an effect on the evaluation of Node-RADS.

Conclusions

MRI-based Node-RADS presents a valuable tool for LN metastasis assessment in PTC, with commendable reproducibility and good diagnostic performance. A threshold of Node-RADS ≥3 demonstrated the highest diagnostic performance, in comparison to Node-RADS ≥4 and Node-RADS ≥5. Through the incorporation of Node-RADS and T1 hyperintensity, we developed a new criterion demonstrating significantly improved diagnostic performance in LN metastasis of PTC.

Acknowledgments

We are grateful to the Radiological Society of North America (RSNA) for selecting a portion of this manuscript as an oral presentation at the 2024 RSNA Annual Meeting.

Footnote

Reporting Checklist: The authors have completed the STARD reporting checklist. Available at https://qims.amegroups.com/article/view/10.21037/qims-2025-740/rc

Data Sharing Statement: Available at https://qims.amegroups.com/article/view/10.21037/qims-2025-740/dss

Funding: This work was supported by the Health and Technology Project of Fujian Province (grant No. 2022QNB020); Natural Science Foundation Project of Fujian Province (grant No. 2022J01120676); and Medical and Health Guided Project of Xiamen (grant No. 3502220214201076).

Conflicts of Interest: All authors have completed the ICMJE uniform disclosure form (available at https://qims.amegroups.com/article/view/10.21037/qims-2025-740/coif). Q.T. reports this work was supported by the Health and Technology Project of Fujian Province (grant No. 2022QNB020). M.W. reports this work was supported by Medical and Health Guided Project of Xiamen (grant No. 3502220214201076). Y.J. reports this work was supported by Natural Science Foundation Project of Fujian Province (grant No. 2022J01120676). The other 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 trial was conducted in accordance with the Declaration of Helsinki and its subsequent amendments. The study was approved by the Institutional Review Board and Ethics Committee of Zhongshan Hospital (Xiamen), Fudan University (No. B2020-002) and informed consent was taken from all individual participants.

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. Lim H, Devesa SS, Sosa JA, Check D, Kitahara CM. Trends in Thyroid Cancer Incidence and Mortality in the United States, 1974-2013. JAMA 2017;317:1338-48. [Crossref] [PubMed]
2. Papaleontiou M, Evron JM, Esfandiari NH, Reyes-Gastelum D, Ward KC, Hamilton AS, Worden F, Haymart MR. Patient Report of Recurrent and Persistent Thyroid Cancer. Thyroid 2020;30:1297-305. [Crossref] [PubMed]
3. Choi SY, Cho JK, Moon JH, Son YI. Metastatic Lymph Node Ratio of Central Neck Compartment Has Predictive Values for Locoregional Recurrence in Papillary Thyroid Microcarcinoma. Clin Exp Otorhinolaryngol 2016;9:75-9. [Crossref] [PubMed]
4. Kim M, Jeon MJ, Oh HS, Park S, Song DE, Sung TY, Kim TY, Chung KW, Kim WB, Shong YK, Lee YM, Kim WG. Prognostic Implication of N1b Classification in the Eighth Edition of the Tumor-Node-Metastasis Staging System of Differentiated Thyroid Cancer. Thyroid 2018;28:496-503.
5. Kim M, Kim HK, Kim HI, Kim EH, Jeon MJ, Yi HS, Kim ES, Kim H, Kim TH, Kim BH, Kim TY, Kang HC, Kim WB, Chung JH, Shong YK, Kim SW, Kim WG. Modification of the eight-edition tumor-node-metastasis staging system with N1b for papillary thyroid carcinoma: A multi-institutional cohort study. Oral Oncol 2018;86:48-52. [Crossref] [PubMed]
6. Haddad RI, Bischoff L, Ball D, Bernet V, Blomain E, Busaidy NL, et al. Thyroid Carcinoma, Version 2.2022, NCCN Clinical Practice Guidelines in Oncology. J Natl Compr Canc Netw 2022;20:925-51. [Crossref] [PubMed]
7. Zhao H, Li H. Meta-analysis of ultrasound for cervical lymph nodes in papillary thyroid cancer: Diagnosis of central and lateral compartment nodal metastases. Eur J Radiol 2019;112:14-21. [Crossref] [PubMed]
8. Lee DW, Ji YB, Sung ES, Park JS, Lee YJ, Park DW, Tae K. Roles of ultrasonography and computed tomography in the surgical management of cervical lymph node metastases in papillary thyroid carcinoma. Eur J Surg Oncol 2013;39:191-6. [Crossref] [PubMed]
9. Hwang HS, Orloff LA. Efficacy of preoperative neck ultrasound in the detection of cervical lymph node metastasis from thyroid cancer. Laryngoscope 2011;121:487-91. [Crossref] [PubMed]
10. Moon HJ, Kim EK, Yoon JH, Kwak JY. Differences in the diagnostic performances of staging US for thyroid malignancy according to experience. Ultrasound Med Biol 2012;38:568-73. [Crossref] [PubMed]
11. Haugen BR, Alexander EK, Bible KC, Doherty GM, Mandel SJ, Nikiforov YE, Pacini F, Randolph GW, Sawka AM, Schlumberger M, Schuff KG, Sherman SI, Sosa JA, Steward DL, Tuttle RM, Wartofsky L. 2015 American Thyroid Association Management Guidelines for Adult Patients with Thyroid Nodules and Differentiated Thyroid Cancer: The American Thyroid Association Guidelines Task Force on Thyroid Nodules and Differentiated Thyroid Cancer. Thyroid 2016;26:1-133. [Crossref] [PubMed]
12. Park JE, Lee JH, Ryu KH, Park HS, Chung MS, Kim HW, Choi YJ, Baek JH. Improved Diagnostic Accuracy Using Arterial Phase CT for Lateral Cervical Lymph Node Metastasis from Papillary Thyroid Cancer. AJNR Am J Neuroradiol 2017;38:782-8. [Crossref] [PubMed]
13. Lesnik D, Cunnane ME, Zurakowski D, Acar GO, Ecevit C, Mace A, Kamani D, Randolph GW. Papillary thyroid carcinoma nodal surgery directed by a preoperative radiographic map utilizing CT scan and ultrasound in all primary and reoperative patients. Head Neck 2014;36:191-202. [Crossref] [PubMed]
14. Lunkiewicz M, Forte S, Freiwald B, Singer G, Leo C, Kubik-Huch RA. Interobserver variability and likelihood of malignancy for fifth edition BI-RADS MRI descriptors in non-mass breast lesions. Eur Radiol 2020;30:77-86.
15. Turkbey B, Rosenkrantz AB, Haider MA, Padhani AR, Villeirs G, Macura KJ, Tempany CM, Choyke PL, Cornud F, Margolis DJ, Thoeny HC, Verma S, Barentsz J, Weinreb JC. Prostate Imaging Reporting and Data System Version 2.1: 2019 Update of Prostate Imaging Reporting and Data System Version 2. Eur Urol 2019;76:340-51. [Crossref] [PubMed]
16. Mitchell DG, Bruix J, Sherman M, Sirlin CB. LI-RADS (Liver Imaging Reporting and Data System): summary, discussion, and consensus of the LI-RADS Management Working Group and future directions. Hepatology 2015;61:1056-65. [Crossref] [PubMed]
17. Elsholtz FHJ, Asbach P, Haas M, Becker M, Beets-Tan RGH, Thoeny HC, Padhani AR, Hamm B. Introducing the Node Reporting and Data System 1.0 (Node-RADS): a concept for standardized assessment of lymph nodes in cancer. Eur Radiol 2021;31:6116-24. [Crossref] [PubMed]
18. Leonardo C, Flammia RS, Lucciola S, Proietti F, Pecoraro M, Bucca B, Licari LC, Borrelli A, Bologna E, Landini N, Del Monte M, Chung BI, Catalano C, Magliocca FM, De Berardinis E, Del Giudice F, Panebianco V. Performance of Node-RADS Scoring System for a Standardized Assessment of Regional Lymph Nodes in Bladder Cancer Patients. Cancers (Basel) 2023;15:580. [Crossref] [PubMed]
19. Loch FN, Beyer K, Kreis ME, Kamphues C, Rayya W, Schineis C, Jahn J, Tronser M, Elsholtz FHJ, Hamm B, Reiter R. Diagnostic performance of Node Reporting and Data System (Node-RADS) for regional lymph node staging of gastric cancer by CT. Eur Radiol 2024;34:3183-93. [Crossref] [PubMed]
20. Lucciola S, Pisciotti ML, Frisenda M, Magliocca F, Gentilucci A, Del Giudice F, et al. Predictive role of node-rads score in patients with prostate cancer candidates for radical prostatectomy with extended lymph node dissection: comparative analysis with validated nomograms. Prostate Cancer Prostatic Dis 2023;26:379-87. [Crossref] [PubMed]
21. Maggialetti N, Greco CN, Lucarelli NM, Morelli C, Cianci V, Sasso S, Rubini D, Scardapane A, Stabile Ianora AA. Applications of new radiological scores: the Node-rads in colon cancer staging. Radiol Med 2023;128:1287-95. [Crossref] [PubMed]
22. Meyer HJ, Schnarkowski B, Pappisch J, Kerkhoff T, Wirtz H, Höhn AK, Krämer S, Denecke T, Leonhardi J, Frille A. CT texture analysis and node-RADS CT score of mediastinal lymph nodes - diagnostic performance in lung cancer patients. Cancer Imaging 2022;22:75. [Crossref] [PubMed]
23. Forghani R, Yu E, Levental M, Som PM, Curtin HD. Imaging evaluation of lymphadenopathy and patterns of lymph node spread in head and neck cancer. Expert Rev Anticancer Ther 2015;15:207-24. [Crossref] [PubMed]
24. Renkonen S, Lindén R, Bäck L, Silén R, Mäenpää H, Tapiovaara L, Aro K. Accuracy of preoperative MRI to assess lateral neck metastases in papillary thyroid carcinoma. Eur Arch Otorhinolaryngol 2017;274:3977-83. [Crossref] [PubMed]
25. Wang P, Hu S, Wang X, Ge Y, Zhao J, Qiao H, Chang J, Dou W, Zhang H. Synthetic MRI in differentiating benign from metastatic retropharyngeal lymph node: combination with diffusion-weighted imaging. Eur Radiol 2023;33:152-61. [Crossref] [PubMed]
26. Obuchowski NA. Sample size calculations in studies of test accuracy. Stat Methods Med Res 1998;7:371-92. [Crossref] [PubMed]
27. Yeh MW, Bauer AJ, Bernet VA, Ferris RL, Loevner LA, Mandel SJ, Orloff LA, Randolph GW, Steward DLAmerican Thyroid Association Surgical Affairs Committee Writing Task Force. American Thyroid Association statement on preoperative imaging for thyroid cancer surgery. Thyroid 2015;25:3-14. [Crossref] [PubMed]
28. Parillo M, Mallio CA, Van der Molen AJ, Rovira À, Dekkers IA, Karst U, Stroomberg G, Clement O, Gianolio E, Nederveen AJ, Radbruch A, Quattrocchi CCESMRMB-GREC Working Group. The role of gadolinium-based contrast agents in magnetic resonance imaging structured reporting and data systems (RADS). MAGMA 2024;37:15-25. [Crossref] [PubMed]
29. Parillo M, van der Molen AJ, Asbach P, Elsholtz FHJ, Laghi A, Ronot M, Wu JS, Mallio CA, Quattrocchi CC. The role of iodinated contrast media in computed tomography structured Reporting and Data Systems (RADS): a narrative review. Quant Imaging Med Surg 2023;13:7621-31. [Crossref] [PubMed]
30. Amin MB, Greene FL, Edge SB, Compton CC, Gershenwald JE, Brookland RK, Meyer L, Gress DM, Byrd DR, Winchester DP. The Eighth Edition AJCC Cancer Staging Manual: Continuing to build a bridge from a population-based to a more “personalized” approach to cancer staging. CA Cancer J Clin 2017;67:93-9.
31. Liu X, Ouyang D, Li H, Zhang R, Lv Y, Yang A, Xie C. Papillary thyroid cancer: dual-energy spectral CT quantitative parameters for preoperative diagnosis of metastasis to the cervical lymph nodes. Radiology 2015;275:167-76. [Crossref] [PubMed]
32. Yang Z, Zhang X, Fang M, Li G, Duan X, Mao J, Shen J. Preoperative Diagnosis of Regional Lymph Node Metastasis of Colorectal Cancer With Quantitative Parameters From Dual-Energy CT. AJR Am J Roentgenol 2019;213:W17-25. [Crossref] [PubMed]
33. Cavalheiro BG, de Matos LL, Leite AKN, Kulcsar MAV, Cernea CR, Kowalski LP. Survival in differentiated thyroid carcinoma: Comparison between the 7th and 8th editions of the AJCC/UICC TNM staging system and the ATA initial risk stratification system. Head Neck 2021;43:2913-22.
34. Zhong J, Mao S, Chen H, Wang Y, Yin Q, Cen Q, Lu J, Yang J, Hu Y, Xing Y, Liu X, Ge X, Jiang R, Song Y, Lu M, Chu J, Zhang H, Zhang G, Ding D, Yao W. Node-RADS: a systematic review and meta-analysis of diagnostic performance, category-wise malignancy rates, and inter-observer reliability. Eur Radiol 2025;35:2723-35. [Crossref] [PubMed]
35. Yu P, Wang C, Zhang H, Zheng G, Jia C, Liu Z, Wang Q, Mu Y, Yang X, Mao N, Song X. Deep learning-based automatic pipeline system for predicting lateral cervical lymph node metastasis in patients with papillary thyroid carcinoma using computed tomography: A multi-center study. Chin J Cancer Res 2024;36:545-61. [Crossref] [PubMed]
36. Pak K, Suh S, Hong H, Cheon GJ, Hahn SK, Kang KW, Kim EE, Lee DS, Chung JK. Diagnostic values of thyroglobulin measurement in fine-needle aspiration of lymph nodes in patients with thyroid cancer. Endocrine 2015;49:70-7. [Crossref] [PubMed]
37. Ng JKM, Chan ABW, Li JJX. Colloid and pigmented histiocytes in lymph node aspirates as a clue to metastasis in patients with a history of papillary thyroid carcinoma. Diagn Cytopathol 2024;52:22-9. [Crossref] [PubMed]
38. Parillo M, Quattrocchi CC. Node Reporting and Data System 1.0 (Node-RADS) for the Assessment of Oncological Patients’ Lymph Nodes in Clinical Imaging. J Clin Med 2025;14:263. [Crossref] [PubMed]