Markedly hypoechoic: a new definition improves the diagnostic performance of thyroid ultrasound

OBJECTIVES

To determine the contribution of a modified definition of markedly hypoechoic in the differential diagnosis of thyroid nodules.

METHODS

A total of 1031 thyroid nodules were included in this retrospective multicenter study. All of the nodules were examined with US before surgery. The US features of the nodules were evaluated, in particular, the classical markedly hypoechoic and modified markedly hypoechoic (decreased or similar echogenicity relative to the adjacent strap muscles). The sensitivity, specificity, and AUC of classical/modified markedly hypoechoic and the corresponding ACR-TIRADS, EU-TIRADS, and C-TIRADS categories were calculated and compared. The inter- and intraobserver variability in the evaluation of the main US features of the nodules was assessed.

RESULTS

There were 264 malignant nodules and 767 benign nodules. Compared with classical markedly hypoechoic as a diagnostic criterion for malignancy, using modified markedly hypoechoic as the criterion resulted in a significant increase in sensitivity (28.03% vs. 63.26%) and AUC (0.598 vs. 0.741), despite a significant decrease in specificity (91.53% vs. 84.88%) (p < 0.001 for all). Compared to the AUC of the C-TIRADS with the classical markedly hypoechoic, the AUC of the C-TIRADS with the modified markedly hypoechoic increased from 0.878 to 0.888 (p = 0.01); however, the AUCs of the ACR-TIRADS and EU-TIRADS did not change significantly (p > 0.05 for both). There was substantial interobserver agreement (κ = 0.624) and perfect intraobserver agreement (κ = 0.828) for the modified markedly hypoechoic.

CONCLUSION

The modified definition of markedly hypoechoic resulted in a significantly improved diagnostic efficacy in determining malignant thyroid nodules and may improve the diagnostic performance of the C-TIRADS.

CLINICAL RELEVANCE STATEMENT

Our study found that, compared with the original definition, modified markedly hypoechoic significantly improved the diagnostic performance in differentiating malignant from benign thyroid nodules and the predictive efficacy of the risk stratification systems.

KEY POINTS

• Compared with the classical markedly hypoechoic as a diagnostic criterion for malignancy, the modified markedly hypoechoic resulted in a significant increase in sensitivity and AUC. • The C-TIRADS with the modified markedly hypoechoic achieved higher AUC and specificity than that with the classical markedly hypoechoic (p = 0.01 and  < 0.001, respectively).

Ultrasound screening for thyroid nodules and cancer in individuals with family history of thyroid cancer: a micro-costing approach

PURPOSE

Screening programs that target healthy populations are an important tool for identifying unrecognized, asymptomatic disease. However, ultrasound screening for thyroid cancer has no obvious advantage in terms of cost-effectiveness in asymptomatic adults. There is far less consensus (and data) on the indications for screening in high-risk individuals. The aim of the study was to estimate the costs of ultrasound screening for individuals with first-degree family history of thyroid cancer.

METHODS

We conducted a prospective cross-sectional study from January 1, 2009 through December 31, 2018 in the Thyroid Cancer Outpatient Clinic of a large teaching hospital in Rome, Italy. We estimated the costs of an ultrasound screening protocol using the micro-costing bottom-up method.

RESULTS

For individuals without thyroid nodules, the screening examination had an estimated cost of €66.21 per screenee. For those found to have unsuspicious nodules, the estimated cost rose to €119.52 per screenee, owing to the addition of thyroid function tests. The estimated cost of screening for a subject with newly diagnosed nodules that were submitted to cytology was €259.89. The total cost of screening for the entire population of 1176 individuals was € 118,133.85. The total expenditure to confirm a single thyroid cancer diagnosis was €10,598.71.

CONCLUSION

A sonographic screening implies a significant direct expenditure and is likely to detect a very large number of individuals with benign nodules (more than 45 asymptomatic individuals are diagnosed with a thyroid nodule for each newly detected cancer case), whose long-term follow-up will further increase healthcare costs.

The performance of deep learning on thyroid nodule imaging predicts thyroid cancer: A systematic review and meta-analysis of epidemiological studies with independent external test sets

BACKGROUND AND AIMS

It is still controversial whether deep learning (DL) systems add accuracy to thyroid nodule imaging classification based on the recent available evidence. We conducted this study to analyze the current evidence of DL in thyroid nodule imaging diagnosis in both internal and external test sets.

METHODS

Until the end of December 2022, PubMed, IEEE, Embase, Web of Science, and the Cochrane Library were searched. We included primary epidemiological studies using externally validated DL techniques in image-based thyroid nodule appraisal. This systematic review was registered on PROSPERO (CRD42022362892).

RESULTS

We evaluated evidence from 17 primary epidemiological studies using externally validated DL techniques in image-based thyroid nodule appraisal. Fourteen studies were deemed eligible for meta-analysis. The pooled sensitivity, specificity, and area under the curve (AUC) of these DL algorithms were 0.89 (95% confidence interval 0.87-0.90), 0.84 (0.82-0.86), and 0.93 (0.91-0.95), respectively. For the internal validation set, the pooled sensitivity, specificity, and AUC were 0.91 (0.89-0.93), 0.88 (0.85-0.91), and 0.96 (0.93-0.97), respectively. In the external validation set, the pooled sensitivity, specificity, and AUC were 0.87 (0.85-0.89), 0.81 (0.77-0.83), and 0.91 (0.88-0.93), respectively. Notably, in subgroup analyses, DL algorithms still demonstrated exceptional diagnostic validity.

CONCLUSIONS

Current evidence suggests DL-based imaging shows diagnostic performances comparable to clinicians for differentiating thyroid nodules in both the internal and external test sets.

Use of molecular testing results to analyze the overuse of atypia of undetermined significance in thyroid cytology

INTRODUCTION

The suggested atypia of undetermined significance (AUS) rate for thyroid fine-needle aspiration biopsies is 10% or less. Prompted by a high institutional AUS rate, we examined using molecular testing results (MTR) as a potential quality metric tool to reduce the AUS rate. We correlated MTR with AUS cytologic findings, surgical pathology follow-up, and individual pathologist AUS rates.

MATERIALS AND METHODS

Demographic data, cytologic diagnoses, MTR, and surgical pathology diagnoses were retrospectively obtained. MTR were classified as either positive or negative. AUS rates and MTR proportions were compared among pathologists. The cytomorphologic features of 143 AUS cases were assessed and correlated with MTR.

RESULTS

Between 2017 and 2022, 710 of 3247 thyroid fine-needle aspirations were classified as AUS, with a yearly average rate of 22% (range = 19%-26%). AUS cases included: 331 (47%) with architectural atypia; 204 (29%) with oncocytic (Hürthle cell) atypia; 99 (14%) with combined architectural and cytologic atypia; and 76 (10%) with isolated cytologic atypia. Most AUS cases with molecular testing had negative MTR (360/492, 73%). AUS with cytologic atypia had higher positive MTR risk (logarithm of odds ratio = 1.27, 95% credible interval [0.5-2.04], P = 0.001). The average positive MTR rate by pathologist was 21.5% (range 0%-35%); higher positive MTR rates had better correlation with subsequent neoplastic/malignant histologic diagnoses. The MTR sensitivity for malignant disease was 89% and the negative predictive value was 91%.

CONCLUSIONS

MTR analysis reveals the importance of cytologic atypia as a determinant of malignancy risk in AUS cases. Periodic analysis of MTR data alongside individual pathologist AUS rates can help refine diagnostic criteria and potentially reduce AUS overuse.

Comparison of six machine learning methods for differentiating benign and malignant thyroid nodules using ultrasonographic characteristics

BACKGROUND

Several machine learning (ML) classifiers for thyroid nodule diagnosis have been compared in terms of their accuracy, sensitivity, specificity, negative predictive value (NPV), positive predictive value (PPV), and area under the receiver operating curve (AUC). A total of 525 patients with thyroid nodules (malignant, n = 228; benign, n = 297) underwent conventional ultrasonography, strain elastography, and contrast-enhanced ultrasound. Six algorithms were compared: support vector machine (SVM), linear discriminant analysis (LDA), random forest (RF), logistic regression (LG), GlmNet, and K-nearest neighbors (K-NN). The diagnostic performances of the 13 suspicious sonographic features for discriminating benign and malignant thyroid nodules were assessed using different ML algorithms. To compare these algorithms, a 10-fold cross-validation paired t-test was applied to the algorithm performance differences.

RESULTS

The logistic regression algorithm had better diagnostic performance than the other ML algorithms. However, it was only slightly higher than those of GlmNet, LDA, and RF. The accuracy, sensitivity, specificity, NPV, PPV, and AUC obtained by running logistic regression were 86.48%, 83.33%, 88.89%, 87.42%, 85.20%, and 92.84%, respectively.

CONCLUSIONS

The experimental results indicate that GlmNet, SVM, LDA, LG, K-NN, and RF exhibit slight differences in classification performance.

TS-DSANN: Texture and shape focused dual-stream attention neural network for benign-malignant diagnosis of thyroid nodules in ultrasound images

Recently, accurate diagnosis of thyroid nodules has played a critical role in precision medicine and healthcare system management. Due to complicated changes in ultrasound features of texture, and similar visual appearance of benign-malignant nodules, the identification of cancerous thyroid lesions from a given ultrasound image still faces challenges for even experienced radiologists. Learning-based computer-aided diagnosis (CAD) systems have accordingly attracted more and more attention recently. However, little research is committed to developing a deep learning-based CAD system that has greater conformity with radiologists' diagnostic decision-making. In this study, we devise a texture and shape focused dual-stream attention neural network, dubbed TS-DSANN. Specifically, in the texture focused stream, we utilize the ImageNet pre-trained ResNet34 to guide the network to recognize texture-related nodule attributes. Meanwhile, in the shape focused stream, in addition to using ResNet34 backbone, jointly learning from scratch with the contour obtained by contour detection module to enhance the extraction of shape features. Afterward, we employ a concatenation operation to aggregate the abovementioned two stream networks for capturing richer and more representative features. Finally, we further utilize an online class activation mapping mechanism to assist the dual-stream network in generating a localization heatmap to obtain more visualization attention to the nodule from the whole image, and supervise classifier's attention in decision-making. Experimental results conducted on the two-center thyroid nodule ultrasound datasets verify that our proposed method has improved the classification performance, superior to the state-of-the-art methods.

The auxiliary diagnosis of thyroid echogenic foci based on a deep learning segmentation model: A two-center study

OBJECTIVE

The aim of this study is to develop AI-assisted software incorporating a deep learning (DL) model based on static ultrasound images. The software aims to aid physicians in distinguishing between malignant and benign thyroid nodules with echogenic foci and to investigate how the AI-assisted DL model can enhance radiologists' diagnostic performance.

METHODS

For this retrospective study, a total of 2724 ultrasound (US) scans were collected from two independent institutions, encompassing 1038 echogenic foci nodules. All echogenic foci were confirmed by pathology. Three DL segmentation models (DeepLabV3+, U-Net, and PSPNet) were developed, with each model using two different backbones to extract features from the nodular regions with echogenic foci. Evaluation indexes such as Mean Intersection over Union (MIoU), Mean Pixel Accuracy (MPA), and Dice coefficients were employed to assess the performance of the segmentation model. The model demonstrating the best performance was selected to develop the AI-assisted diagnostic software, enabling radiologists to benefit from AI-assisted diagnosis. The diagnostic performance of radiologists with varying levels of seniority and beginner radiologists in assessing high-echo nodules was then compared, both with and without the use of auxiliary strategies. The area under the receiver operating characteristic curve (AUROC) was used as the primary evaluation index, both with and without the use of auxiliary strategies.

RESULTS

In the analysis of Institution 2, the DeepLabV3+ (backbone is MobileNetV2 exhibited optimal segmentation performance, with MIoU = 0.891, MPA = 0.945, and Dice = 0.919. The combined AUROC (0.693 [95% CI 0.595-0.791]) of radiology beginners using AI-assisted strategies was significantly higher than those without such strategies (0.551 [0.445-0.657]). Additionally, the combined AUROC of junior physicians employing adjuvant strategies improved from 0.674 [0.574-0.774] to 0.757 [0.666-0.848]. Similarly, the combined AUROC of senior physicians increased slightly, rising from 0.745 [0.652-0.838] to 0.813 [0.730-0.896]. With the implementation of AI-assisted strategies, the accuracy, sensitivity, specificity, positive predictive value (PPV), and negative predictive value (NPV) of both senior physicians and beginners in the radiology department underwent varying degrees of improvement.

CONCLUSIONS

This study demonstrates that the DL-based auxiliary diagnosis model using US static images can improve the performance of radiologists and radiology students in identifying thyroid echogenic foci.

Comparative Study of C-TIRADS, ACR-TIRADS, and EU-TIRADS for Diagnosis and Management of Thyroid Nodules

RATIONALE AND OBJECTIVES

Chinese Thyroid Imaging Reporting and Data Systems (C-TIRADS) was developed to provide a more simplified tool for stratifying thyroid nodules. Here we aimed to validate the efficacy of C-TIRADS in distinguishing benign from malignant and in guiding fine-needle aspiration biopsies in comparison with the American College of Radiology TIRADS (ACR-TIRADS) and European TIRADS (EU-TIRADS).

MATERIALS AND METHODS

This study retrospectively included 3438 thyroid nodules (≥10 mm) in 3013 patients (mean age, 47.1 years ± 12.9) diagnosed between January 2013 and November 2019. Ultrasound features of the nodules were evaluated and categorized according to the lexicons of the three TIRADS. We compared these TIRADS by using the area under the receiver operating characteristic curve (AUROC), the area under the precision-recall curve (AUPRC), sensitivity, specificity, net reclassification improvement (NRI), and unnecessary fine-needle aspiration biopsy (FNAB) rate.

RESULTS

Of the 3438 thyroid nodules, 707 (20.6%) were malignant. C-TIRADS showed higher discrimination performance (AUROC, 0.857; AUPRC, 0.605) than ACR-TIRADS (AUROC, 0.844; AUPRC, 0.567) and EU-TIRADS (AUROC, 0.802; AUPRC, 0.455). The sensitivity of C-TIRADS (85.3%) was lower than that of ACR-TIRADS (89.1%) but higher than that of EU-TIRADS (78.4%). The specificity of C-TIRADS (76.9%) was similar to that of EU-TIRADS (78.9%) and higher than that of ACR-TIRADS (69.5%). The unnecessary FNAB rate was lowest with C-TIRADS (21.2%), followed by ACR-TIRADS (41.7%) and EU-TIRADS (58.3%). C-TIRADS obtained significant NRI for recommending FNAB over ACR-TIRADS (19.0%, P < 0.001) and EU-TIRADS (25.5%, P < 0.001).

CONCLUSION

C-TIRADS may be a clinically applicable tool to manage thyroid nodules, which warrants thorough tests in other geographic settings.

Quantitative analysis of dual-phase enhanced CT in cervical lymph node metastasis of papillary thyroid carcinoma: a comparative study along with pathological manifestations

PURPOSE

This study aimed to investigate the diagnostic value of dual-phase enhanced computed tomography (CT) in the cervical lymph node metastasis (LNM) of papillary thyroid carcinoma (PTC) by analyzing the dual-phase enhanced Hounsfield units (HUs) of lymph node and sternocleidomastoid muscle, and the ratio and difference.

METHODS

The CT arterial-phase and venous-phase imaging data of 143 metastasis-positive lymph nodes (MPLNs) in 88 cases and 172 metastasis-negative lymph nodes (MNLNs) in 128 cases with PTC were retrospectively analyzed. All lymph nodes were confirmed by surgical pathology. The arterial-phase HU of lymph nodes (ANHU), venous-phase HU of lymph nodes (VNHU), arterial-phase HU of the sternocleidomastoid muscle (AMHU) and venous-phase HU of the sternocleidomastoid muscle (VMHU) were measured, and their difference and ratio (ANHU-AMHU, ANHU/AMHU, VNHU-VMHU, VNHU/VMHU) were calculated. The cutoff values and corresponding diagnostic efficacy for diagnosing LNM in PTC were sought by performing the receiver operating characteristic curves. The maximum pathological diameter (MPD) measured on pathological sections of lymph nodes was compared with the maximum transverse diameter (MTD) and maximum sagittal diameter (MSD) and their average values on CT images.

RESULTS

The ANHU, and VNHU of MPLNs and MNLNs were 111.89 ± 33.26 and 66.12 (56.81-76.86) (P < 0.001), and 99.07 ± 23.27 and 75.47 ± 13.95 (P < 0.001), respectively. The area under the curve, sensitivity, and specificity of the arterial-phase three parameters (ANHU, ANHU-AMHU, ANHU/AMHU) for diagnosing LNM were (0.877-0.880), (0.755-0.769), and (0.901-0.913), respectively, and the venous-phase three parameters (VNHU, VNHU-VMHU, VNHU/VMHU) were (0.801-0.817), (0.650-0.678), and (0.826-0.901), respectively. Compared with MPD, MTD (Z = -2.686, P = 0.007) and MSD (Z = -3.539, P < 0.001) were significantly different, while (MTD + MSD)/2 was not statistically different (Z = -0.038b, P = 0.969).

CONCLUSION

In the differential diagnosis of cervical LNM of PTC by dual-phase enhanced CT angiography, the arterial phase had higher diagnostic efficacy.

Radiofrequency ablation for autonomously functioning nodules as treatment for hyperthyroidism: subgroup analysis of toxic adenoma and multinodular goitre and predictors for treatment success

PURPOSE

Treatment of hyperthyroidism caused by autonomously functioning thyroid nodules (AFTN) with 131I often leads to undesirable hypothyroidism. Radiofrequency ablation (RFA) has emerged as a promising alternative. This retrospective analysis aimed to examine the efficacy of, and postprocedural hypothyroidism after, RFA for AFTN.

METHODS

Patients with hyperthyroidism caused by AFTN and treated with RFA were included if follow-up of at least 1 year was available. Cure was defined as thyroid medication-free biochemical euthyroidism. To predict cure, patient and treatment factors were analysed. A distinction was made between solitary toxic adenoma (STA) and toxic multinodular goitre (TMG).

RESULTS

Forty-eight patients (36 STA, 12 TMG) were included. One year post-RFA cure rate was 72% in STA versus 25% in TMG (p = 0.004). One patient developed hypothyroidism. In 11 patients that remained hyperthyroid, a second RFA was successful in 83% of STA and 40% of TMG patients. At last available follow-up, this amounted to a total cure rate of 81% in STA and 33% in TMG (p = 0.002). In STA, cured patients had a higher baseline TSH and a lower FT3 than non-cured patients (p = 0.026 and 0.031). Cure was observed in 91% of patients when > 2.1 kJ/mL was delivered during RFA, compared to 44% when less energy was administered.

CONCLUSION

The efficacy of RFA was nearly 3 times higher in STA patients compared to TMG. Severity of hyperthyroidism and kJ/mL delivered during RFA predicts cure. Direct comparison to the current standard of care is needed to implement RFA in treatment of hyperthyroidism caused by AFTN.

Diagnostic performance of shear wave elastography in thyroid nodules with indeterminate cytology: A systematic review and meta-analysis

Purpose

Thyroid nodules classified as indeterminate in previous fine-needle aspiration cytology often necessitate additional evaluation to determine their histology, while shear wave elastography (SWE) offers an alternative option in this regard. The objective of this study was to assess the diagnostic effectiveness of SWE in evaluating indeterminate nodules.

Methods

The PubMed, EMBASE, and Web of Science databases were searched from 1st January 1970 to 1st March 2023. The studies were reviewed and the data was extracted by two separate reviewers. A Bayesian bivariate model was utilized to quantitatively synthesize the diagnostic accuracy and yield of the studies in R.

Results

A total of seven studies, involving indeterminate thyroid nodules undergoing SWE were included, and the overall malignancy rate was 34.1% (307/900). The summarized estimates of sensitivity and specificity were 0.792 (95% credible interval [CI], 0.727-0.850) and 0.845 (95% CI, 0.797-0.887), respectively. The summarized estimate for the diagnostic odds ratio (DOR) was 17.8 (95% CI, 14.0-22.6). Summarized receiver operating characteristic (SROC) plots indicated a trade-off between sensitivity and specificity, and the estimate of AUC was 0.866 (95% CI, 0.834-0.895). The summary estimates for positive and negative likelihood ratios were 4.67 (95% CI, 3.98-5.85) and 0.26 (95% CI, 0.23-0.28), respectively.

Conclusions

The overall accuracy of SWE remains satisfactory in indeterminate thyroid nodules. However, it should be noted that the available data are still extremely limited, and more studies or guidelines are required to provide further insights.

Comparison of the Efficiency of Radiofrequency and Microwave Ablation Methods in the Treatment of Benign Thyroid Nodules

RATIONALE AND OBJECTIVES

To compare the efficacy of two thermal ablation techniques (radiofrequency vs microwave ablation) in the treatment of benign thyroid nodules.

MATERIALS AND METHODS

A total of 80 patients with 80 nodules underwent thermal ablation of thyroid nodules with either radiofrequency ablation (RFA) (23 females and 14 males; mean age 41 ± 9years) or microwave ablation (MWA) (28 females and 15 males; mean age 45 ± 11years). Ultrasound assessments were made at the 1st, 3rd, 6th, and 12th months after the ablation procedure.

RESULTS

The mean initial volume of the nodules (RFA: 15.6 mL [min 2.5-max 74]; MWA: 40 mL [min 2-max 205]) was statistically significantly different (P < .001). The following were the volume reduction rates of nodules at 1, 3, 6, and 12months, respectively: after RFA: 46.8 ± 13.5%, 62.9 ± 13.6%, 71.6 ± 11.9%, and 77.9 ± 10.3%. After MWA: 38.7 ± 12.5%, 54 ± 15.3%, 59.6 ± 12.5%, and 65 ± 11.3%. For all months, volume reduction rates in the RFA group were significantly higher than those in the MWA group (P < .05). One patient treated by RFA reported an abscess formation and another patient treated by RFA had a self-limiting hematoma, who recovered without any further treatment. Also, in the MWA group, one patient had abscess formation and another patient had transient recurrent nerve paralysis, who recovered with appropriate treatment.

CONCLUSION

Both methods are effective in treating benign thyroid nodules; however, RFA provides a better volume reduction.

Monocyte-to-High-Density Lipoprotein Cholesterol Ratio Together With the Lymphocyte-to-Monocyte Ratio in Predicting the Malignancy of the Thyroid Nodule in Patients Complicated With Type 2 Diabetes

Background

Thyroid nodules, although mostly benign and symptomless, have a small chance of being cancerous, necessitating accurate diagnosis. This study aims to develop and validate a nomogram for differentiating malignant and non-malignant thyroid nodules in individuals with type 2 diabetes.

Methods

The study included 484 patients with both thyroid nodules and type 2 diabetes who underwent thyroid gland lobectomy at Wenzhou Medical University Hospital. Optimal cutoff values for continuous variables were determined using ROC curve analysis. Significant factors identified in univariable analysis were used to construct the nomogram. The monocyte-to-high-density lipoprotein cholesterol ratio (MHR) was visualized through a histogram and scatter diagram. Discriminatory power was assessed using ROC analysis, and calibration curves ensured consistency. Decision curve analysis (DCA) evaluated clinical benefits.

Results

The cohort was divided into a training group (70%) and an internal validation group (30%). The scatter diagram revealed a correlation between MHR levels and the proportion of goiter cases, with higher MHR levels associated with increased goiter incidence. The histogram showed higher average MHR levels in goiter patients compared to those with papillary thyroid carcinoma (PTC) in both groups. Multivariate logistic regression identified age, total cholesterol (TC), triglyceride (TG), fasting blood sugar (FSG), fibrinogen, lymphocyte-to-monocyte ratio (LMR), and MHR as independent predictive factors for malignancy in thyroid nodules with type 2 diabetes. The nomogram achieved high discrimination, with C-index values of 0.901 (training data set) and 0.760 (internal validation data set). Calibration curves displayed good agreement, and DCA demonstrated significant net clinical benefits.

Conclusions

MHR is associated with sex, serum cholesterol levels, and peripheral blood cell counts, making it a potential novel biomarker for differentiating between PTC and goiter in type 2 diabetes patients.

Performance of multigene testing in cytologically indeterminate thyroid nodules and molecular risk stratification

Objective

Thyroid cancer is the third most prevalent cancer among females. Genetic testing based on next-generation sequencing may provide an auxiliary diagnosis to reduce cytologically diagnostic uncertainty. However, commercial multigene tests are not widely available and are not well-tested in the Chinese population.

Methods

In this study, we designed a multigene testing panel and evaluated its performance in 529 cytologically indeterminate thyroid nodules (Bethesda III, IV and V). The molecular data of the DNA mutations and RNA fusions of fine needle aspiration samples were reviewed in conjunction with a clinical diagnosis, pathological reports, and definitive surgery for retrospective analysis. Then, the molecular risk stratification was investigated for its accuracy in malignant risk prediction.

Results

The overall combined consistency revealed substantial agreement (Kappa = 0.726) with the sensitivity, specificity, positive predictive value, and negative predictive values of 97.80%, 82.14%, 98.99%, and 67.65%, respectively. The most common aberration was BRAFV600E (82.59%), followed by NRAS mutants (4.07%), RET fusions (3.70%), and KRAS mutants (3.15%). Two cases (0.44%) were categorized into a high-risk group, 426 cases (94.67%) were categorized into a BRAF-like group with totally histopathologic papillary patterned tumors, and 22 cases (4.89%) were categorized into a RAS-like group with 14 papillary and eight follicular patterned tumors when the cohort concurrent aberrations were excluded. Potentially aggressive features may be related to concurrent molecular alterations of BRAFV600E with TERTQ302R, and AKT1L52R, NRASG12C, NRASQ61R, and CCDC6-RET fusions.

Conclusions

This study provided a multigene panel for identifying benign nodules from cytologically indeterminate thyroid nodules to avoid unnecessary surgery. We provide further evidence for using molecular risk stratification as a promising predictor of disease outcomes. The results of this study may be limited by the extremely high prevalence of cancer in the cohort for clinical reference.

Clinical value of grayscale ultrasound combined with real-time shear wave elastography nomogram in risk prediction of thyroid cancer

OBJECTIVES

This study constructed a nomogram based on grayscale ultrasound features and real-time shear wave elastography (SWE) parameters to predict thyroid cancer.

METHODS

Clinical data of 217 thyroid nodules of 201 patients who underwent grayscale ultrasound, real-time SWE, and thyroid function laboratory examination in Ma'anshan People's Hospital from January 2019 to December 2020 were retrospectively analyzed. The subjects were divided into a benign nodule group (106 nodules) and a malignant nodule group (111 nodules). The differences in grayscale ultrasound features, quantitative parameters of real-time SWE, and laboratory results of thyroid function between benign and malignant thyroid nodules were analyzed. We used a chi-square test for categorical variables and a t-test for continuous variables. Then, the independent risk factors for thyroid cancer were analyzed using multivariate logistic regression. Based on the independent risk factors, a nomogram for predicting thyroid cancer risk was constructed using the RMS package of the R software.

RESULTS

Multivariate logistic regression showed that the grayscale ultrasound features of thyroid nodules were the shape, margin, echogenicity, and echogenic foci of the nodules,the maximum Young's modulus (SWE-max) of thyroid nodules, and the ratio of thyroid nodule and peripheral gland (SWE-ratio) measured by real-time SWE were independent risk factors for thyroid cancer (all p < 0.05), and the other variables had no statistical difference (p > 0.05). Based on the shape (OR = 5.160, 95% CI: 2.252-11.825), the margin (OR = 9.647, 95% CI: 2.048-45.443), the echogenicity (OR = 6.512, 95% CI: 1.729-24.524), the echogenic foci (OR = 2.049, 95% CI: 1.118-3.756), and the maximum Young's modulus (SWE-max) (OR = 1.296, 95% CI: 1.140-1.473), the SWE-ratio (OR = 2.001, 95% CI: 1.403-2.854) of the thyroid nodule to peripheral gland was used to establish the related nomogram prediction model. The bootstrap self-sampling method was used to verify the model. The consistency index (C-index) was 0.979, ROC curve was used to analyze the nomogram scores of all patients, and the AUC of nomogram prediction of thyroid cancer was 0.976, indicating that the nomogram model had high accuracy in the risk prediction of thyroid cancer.

CONCLUSIONS

The nomogram model of grayscale ultrasound features combined with SWE parameters can accurately predict thyroid cancer.

Can Thyroid Nodules be Managed with Radiofrequency Ablation?

Radiofrequency ablation (RFA) is an established and growing minimally-invasive technique with an impressive safety profile used to manage thyroid nodules. Beyond shorter operative and recovery times, the main advantages of RFA include the lack of an incisional scar as well as maximizing the potential for maintenance of normal thyroid function. RFA can significantly reduce nodular volume, achieving rates of 60% to 95% in a breadth of thyroid disease, including benign nodules, primary carcinomas, and recurrent malignancies. Thorough patient counselling is imperative for RFA candidates, including a discussant of complications, nodule regrowth, and the potential for a subsequent ablation session.

Correlation of the ultrasound thyroid imaging reporting and data system with cytology findings among patients in Uganda

BACKGROUND

Ultrasonography is a noninvasive modality for the initial assessment of thyroid nodules. The American College of Radiology Thyroid Imaging Reporting and Data System (ACR TI-RADS) has demonstrated good performance in differentiating malignant thyroid nodules. However, the combination of ACR TI-RADS categories and cytology has not been studied extensively, in Uganda. The study aims to correlate ACR TI-RADS with cytology among patients referred for US-guided fine-needle aspiration at Mulago National Referral Hospital.

METHODS

This was a hospital-based cross-sectional study that recruited 132 patients with thyroid nodules. Spearman's correlation was used to establish a relationship between TI-RADS and cytology findings. The diagnostic accuracy of TI-RADS was assessed using sensitivity, specificity, positive and negative predictive values, and positive and negative likelihood ratios.

RESULTS

Of 132 study participants, 90% (n = 117) were females, and the mean age was 41 ± 13 years. One hundred sixty-one thyroid nodules were analyzed. More than half of the thyroid nodules (54.7%, n = 87) were solid or almost solid, 96.9% (n = 154) were shaped wider than tall, 57.2% (n = 91) had smooth margins, 83.7% (n = 133) were hyperechoic or isoechoic, and 88.7% (n = 141) had no echogenic foci. TI-RADS 3 was the most common at 42.9% (n = 69). The proportions of malignancy for TI-RADS 4 and TI-RADS 5 were 73.3% and 85.7%, respectively. The correlation between ACR TI-RADS and the Bethesda system of thyroid classification scores was r = 0.577. The sensitivity, specificity, positive and negative predictive values, and positive and negative likelihood ratios of ACR TI-RADS were 90.9%, 98.5%, 90%, 99.3%, 62.3, and 0.1, respectively.

CONCLUSION

We found that ACR TI-RADS classification is an appropriate and noninvasive method for assessing thyroid nodules in routine practice. It can safely reduce the number of unnecessary fine-needle aspiration in a significant proportion of benign thyroid lesions. Thyroid nodules classified as TI-RADS 3 should be followed routinely. ACR TI-RADS should be standardized as the screening tool in resource-limited areas.

[An Improved Object Detection Algorithm for Thyroid Nodule Ultrasound Image Based on Faster R-CNN]

Objective

To propose an improved algorithm for thyroid nodule object detection based on Faster R-CNN so as to improve the detection precision of thyroid nodules in ultrasound images.

Methods

The algorithm used ResNeSt50 combined with deformable convolution (DC) as the backbone network to improve the detection effect of irregularly shaped nodules. Feature pyramid networks (FPN) and Region of Interest (RoI) Align were introduced in the back of the trunk network. The former was used to reduce missed or mistaken detection of thyroid nodules, and the latter was used to improve the detection precision of small nodules. To improve the generalization ability of the model, parameters were updated during backpropagation with an optimizer improved by Sharpness-Aware Minimization (SAM).

Results

In this experiment, 6 261 thyroid ultrasound images from the Affiliated Hospital of Xuzhou Medical University and the First Hospital of Nanjing were used to compare and evaluate the effectiveness of the improved algorithm. According to the findings, the algorithm showed optimization effect to a certain degree, with the AP50 of the final test set being as high as 97.4% and AP@50:5:95 also showing a 10.0% improvement compared with the original model. Compared with both the original model and the existing models, the improved algorithm had higher detection precision and improved capacity to detect thyroid nodules with better accuracy and precision. In particular, the improved algorithm had a higher recall rate under the requirement of lower detection frame precision.

Conclusion

The improved method proposed in the study is an effective object detection algorithm for thyroid nodules and can be used to detect thyroid nodules with accuracy and precision.

[Sociodemographic, clinical and ultrasound characterization of thyroid nodule pathology and its association with malignancy in a Colombian high-complexity center]

BACKGROUND AND OBJECTIVE

Thyroid nodules are among the most frequent conditions, with a 10% risk of malignancy. The objective is to describe the frequency of demographic, clinical, and ultrasonographic characteristics of thyroid nodule pathology in adults and to explore the relationship with tumor malignancy.

METHODS

An analytical, retrospective cross-sectional study in adults with thyroid nodules and nodular fine-needle aspiration performed in adult patients from a Colombian reference center between 2009-2019. Data were obtained from the clinical history, descriptive measures of the patient's demographic, clinical, and ultrasound variables were estimated, and their relationship with the malignancy of the tumor was explored.

RESULTS

A total of 445 patients and 515 nodules were included. The median age was 55 years (IQR 44-64), 86.8% of women, and 54.8% had a single lesion. Percentages of 80.2 and 19.8 were benign and malignant nodules, with a median of 15.7mm (IQR 11-25) and 12.7mm (IQR 8.5-18.3), respectively (p<0.001). Hypothyroidism and levothyroxine consumption were higher in those with malignant nodules (p<0.001). The echographic characteristics were statistically different between the nodules. In the malignant ones, there was a higher frequency of solid composition, hypoechogenicity, and irregular margins. In contrast, in the benign ones, the absence of echogenic focus stood out (p<0.001).

CONCLUSION

The ultrasound characteristics are essential to define the risk of malignancy of a thyroid nodule. Therefore, considering the most frequent ones can help in the most appropriate approach to primary care.

Identification of NIFTP-Specific mRNA Markers for Reliable Molecular Diagnosis of Thyroid Tumors

Non-invasive follicular thyroid neoplasm with papillary-like nuclear features (NIFTP) is a low-risk thyroid tumor with a favorable prognosis. Nonetheless, differentiating NIFTP from other thyroid tumors remains challenging, necessitating reliable diagnostic markers. This study is aimed at discovering NIFTP-specific mRNA markers through RNA sequencing analysis of thyroid tumor tissues. We performed mRNA expression profiling for 74 fresh frozen thyroid tissue samples, including NIFTP and benign and malignant follicular-cell-derived tumors. NIFTP/malignant tumors showed 255 downregulated genes and 737 upregulated genes compared to benign tumors. Venn diagram analysis revealed 19 significantly upregulated and 7 downregulated mRNAs in NIFTP. Akaike information criterion analysis allowed us to select OCLN, ZNF423, LYG1, and AQP5 mRNA markers. We subsequently developed a predictive model based on logistic regression analysis using these four mRNAs, which we validated in independent samples (n = 90) using a qRT-PCR assay. This model demonstrated high accuracy in predicting NIFTP in discovery dataset (AUC (area under the receiver operating characteristic) = 0.960) and the validation dataset (AUC = 0.757). Our results suggest that OCLN, ZNF423, LYG1, and AQP5 mRNA markers might serve as reliable molecular markers for identifying NIFTP among other thyroid tumors, ultimately aiding in accurate diagnosis and management of NIFTP patients.

Delayed Cancer Diagnosis in Thyroid Nodules Initially Treated as Benign With Radiofrequency Ablation: Ultrasound Characteristics and Predictors for Cancer

OBJECTIVE

Regrowth after radiofrequency ablation (RFA) of symptomatic large thyroid nodules, initially treated as benign, sometimes turns out to be malignancies. This study aimed to assess the ultrasound (US) characteristics of thyroid nodules initially treated as benign with RFA and later diagnosed as cancers, predictive factors for cancers masquerading as benign, and methods to avoid RFA in these cancers.

MATERIALS AND METHODS

We reviewed the medical records of 134 consecutive patients with 148 nodules who underwent RFA between February 2008 and November 2016 for the debulking of symptomatic thyroid nodules diagnosed as benign using US-guided biopsy. We investigated the pre-RFA characteristics of the thyroid nodules, changes at follow-up after RFA, and the final surgical pathology.

RESULTS

Nodule regrowth after RFA was observed in 36 (24.3%) of the 148 benign nodules. Twenty-two of the 36 nodules were surgically removed, and malignancies were confirmed in seven (19.4% of 36). Of the 22 nodules removed surgically, pre-RFA median volume (range) was significantly larger for malignant nodules than for benign nodules: 22.4 (13.9-84.5) vs. 13.4 (7.3-16.8) mL (P = 0.04). There was no significant difference in the regrowth interval between benign and malignant nodules (P = 0.49). The median volume reduction rate (range) at 12 months was significantly lower for malignant nodules than for benign nodules (51.4% [0-57.8] vs. 83.8% [47.9-89.6]) (P = 0.01). The pre-RFA benignity of all seven malignant nodules was confirmed using two US-guided fine-needle aspirations (FNAs), except for one nodule, which was confirmed using US-guided core-needle biopsy (CNB). Regrown malignant nodules were diagnosed as suspicious follicular neoplasms by CNB. Histological examination of the malignant nodules revealed follicular thyroid carcinomas, except for one follicular variant, a papillary thyroid carcinoma.

CONCLUSION

Symptomatic large benign thyroid nodules showing regrowth or suboptimal reduction after RFA may have malignant potential. The confirmation of these nodules is better with CNB than with FNA.

An adaptive multi-modal hybrid model for classifying thyroid nodules by combining ultrasound and infrared thermal images

BACKGROUND

Two types of non-invasive, radiation-free, and inexpensive imaging technologies that are widely employed in medical applications are ultrasound (US) and infrared thermography (IRT). The ultrasound image obtained by ultrasound imaging primarily expresses the size, shape, contour boundary, echo, and other morphological information of the lesion, while the infrared thermal image obtained by infrared thermography imaging primarily describes its thermodynamic function information. Although distinguishing between benign and malignant thyroid nodules requires both morphological and functional information, present deep learning models are only based on US images, making it possible that some malignant nodules with insignificant morphological changes but significant functional changes will go undetected.

RESULTS

Given the US and IRT images present thyroid nodules through distinct modalities, we proposed an Adaptive multi-modal Hybrid (AmmH) classification model that can leverage the amalgamation of these two image types to achieve superior classification performance. The AmmH approach involves the construction of a hybrid single-modal encoder module for each modal data, which facilitates the extraction of both local and global features by integrating a CNN module and a Transformer module. The extracted features from the two modalities are then weighted adaptively using an adaptive modality-weight generation network and fused using an adaptive cross-modal encoder module. The fused features are subsequently utilized for the classification of thyroid nodules through the use of MLP. On the collected dataset, our AmmH model respectively achieved 97.17% and 97.38% of F1 and F2 scores, which significantly outperformed the single-modal models. The results of four ablation experiments further show the superiority of our proposed method.

CONCLUSIONS

The proposed multi-modal model extracts features from various modal images, thereby enhancing the comprehensiveness of thyroid nodules descriptions. The adaptive modality-weight generation network enables adaptive attention to different modalities, facilitating the fusion of features using adaptive weights through the adaptive cross-modal encoder. Consequently, the model has demonstrated promising classification performance, indicating its potential as a non-invasive, radiation-free, and cost-effective screening tool for distinguishing between benign and malignant thyroid nodules. The source code is available at https://github.com/wuliZN2020/AmmH .

Analysis and prediction of regrowth in benign thyroid nodules undergoing radiofrequency ablation: a retrospective study with a 5-year follow-up

OBJECTIVES

To evaluate the 5-year follow-up results of radiofrequency ablation (RFA) for benign thyroid nodules (BTNs), and construct pre- and postablation nomogram models to predict regrowth in BTNs undergoing RFA.

METHODS

According to the occurrence of regrowth, BTNs were divided into two subgroups, the regrowth group and the nonregrowth group, and the variables were compared between these two subgroups. Then, univariate and multivariate Cox regression analyzes were utilized to filter the independent prognostic factors of regrowth, which then were introduced into the pre- and postablation prognostic nomograms, respectively. The discrimination powers and prediction performances of the nomograms were appraised by receiver operating characteristic (ROC) curves and calibration charts.

RESULTS

The 5-year mean volume reduction rate was 88.80%, with a complication rate of 0.35% (7/200). Within 5 years of follow-up, the regrowth rate was 19% (38/200). Pre- and postablation prognostic nomograms were established to predict the probability of nonregrowth at 1, 3, and 5 years after RFA. The preablation nomogram included initial volume, perinodular vascularity, and high enhancement ring. The postablation nomogram incorporated total volume, perinodular vascularity, and energy applied per volume. The area under the ROC curves and concordance index values of these models were all above 0.7, indicating that the prognostic nomograms achieved satisfactory discrimination powers and prediction performances.

CONCLUSION

RFA for BTNs has long-term efficacy and safety under the influence of key techniques. The pre- and postablation nomograms constructed in the present study might facilitate clinical decision-making before RFA and for the follow-up management after RFA.

KEY POINTS

• Perinodular vascularity, rather than intranodular vascularity was an independent predictor of regrowth, both before and after the RFA procedure. • A high enhancement ring on preablation contrast-enhanced ultrasound was an independent predictor of regrowth. • Pre- and postablation nomograms might facilitate clinical decision-making before RFA and follow-up management after RFA.

Leveraging deep learning to identify calcification and colloid in thyroid nodules

Background

Both calcification and colloid in thyroid nodules are reflected as echogenic foci in ultrasound images. However, calcification and colloid have significantly different probabilities of malignancy. We explored the performance of a deep learning (DL) model in distinguishing the echogenic foci of thyroid nodules as calcification or colloid.

Methods

We conducted a retrospective study using ultrasound image sets. The DL model was trained and tested on 30,388 images of 1127 nodules. All nodules were pathologically confirmed. The area under the receiver-operator characteristic curve (AUC) was employed as the primary evaluation index.

Results

The YoloV5 (You Only Look Once Version 5) transfer learning model for thyroid nodules based on DL detection showed that the average sensitivity, specificity, and accuracy of distinguishing echogenic foci in the test 1 group (n = 192) was 78.41%, 91.36%, and 77.81%, respectively. The average sensitivity, specificity, and accuracy of the three radiologists were 51.14%, 82.58%, and 61.29%, respectively. The average sensitivity, specificity, and accuracy of distinguishing small echogenic foci in the test 2 group (n = 58) was 70.17%, 77.14%, and 73.33%, respectively. Correspondingly, the average sensitivity, specificity, and accuracy of the radiologists were 57.69%, 63.29%, and 59.38%.

Conclusions

The study demonstrated that DL performed far better than radiologists in distinguishing echogenic foci of thyroid nodules as calcifications or colloid.