A Clinical Decision Support System Using Ultrasound Textures and Radiologic Features to Distinguish Metastasis From Tumor-Free Cervical Lymph Nodes in Patients With Papillary Thyroid Carcinoma

Ultrasound-Base Radiomics for Discerning Lymph Node Metastasis in Thyroid Cancer: A Systematic Review and Meta-analysis

PURPOSE

Ultrasound is the imaging modality of choice for preoperative diagnosis of lymph node metastasis (LNM) in thyroid cancer (TC), yet its efficacy remains suboptimal. As radiomics gains traction in tumor diagnosis, its integration with ultrasound for LNM differentiation in TC has emerged, but its diagnostic merit is debated. This study assesses the accuracy of ultrasound-integrated radiomics in preoperatively diagnosing LNM in TC.

METHODS

Literatures were searched in PubMed, Embase, Cochrane, and Web of Science until July 11, 2023. Quality of the studies was assessed by the radiomics quality score (RQS). A meta-analysis was executed using a bivariate mixed effects model, with a subgroup analysis based on modeling variables (clinical features, radiomics features, or their combination).

RESULTS

Among 27 articles (16,410 TC patients, 6356 with LNM), the average RQS was 16.5 (SD:5.47). Sensitivity of the models based on clinical features, radiomics features, and radiomics features plus clinical features were 0.64, 0.76 and 0.69. Specificities were 0.77, 0.78 and 0.82. SROC values were 0.76, 0.84 and 0.81.

CONCLUSION

Ultrasound-based radiomics effectively evaluates LNM in TC preoperatively. Adding clinical features does not notably enhance the model's performance. Some radiomics studies showed high bias, possibly due to the absence of standard application guidelines.

Prediction of cervical lymph node metastasis in differentiated thyroid cancer based on radiomics models

OBJECTIVE

The accurate clinical diagnosis of cervical lymph node metastasis plays an important role in the treatment of differentiated thyroid cancer (DTC). This study aimed to explore and summarize a more objective approach to detect cervical malignant lymph node metastasis of DTC via radiomics models.

METHODS

PubMed, Web of Science, MEDLINE, EMBASE, and Cochrane databases were searched for all eligible studies. Articles using radiomics models based on ultrasound, computed tomography, or magnetic resonance imaging to assess cervical lymph node metastasis preoperatively were included. Characteristics and diagnostic accuracy measures were extracted. Bias and applicability judgments were evaluated by the revised QUADAS-2 tool. The estimates were pooled using a random-effects model. Additionally, the leave-one-out method was conducted to assess the heterogeneity.

RESULTS

Twenty-nine radiomics studies with 6160 validation set patients were included in the qualitative analysis, and 11 studies with 3863 validation set patients were included in the meta-analysis. Four of them had an external independent validation set. The studies were heterogeneous, and a significant risk of bias was found in 29 studies. Meta-analysis showed that the pooled sensitivity and specificity for preoperative prediction of lymph node metastasis via US-based radiomics were 0.81 (95% CI, 0.73-0.86) and 0.87 (95% CI, 0.83-0.91), respectively.

CONCLUSIONS

Although radiomics-based models for cervical lymphatic metastasis in DTC have been demonstrated to have moderate diagnostic capabilities, broader data, standardized radiomics features, robust feature selection, and model exploitation are still needed in the future.

ADVANCES IN KNOWLEDGE

The radiomics models showed great potential in detecting malignant lymph nodes in thyroid cancer.

Ultrasound radiomics models based on multimodal imaging feature fusion of papillary thyroid carcinoma for predicting central lymph node metastasis

Objective

This retrospective study aimed to establish ultrasound radiomics models to predict central lymph node metastasis (CLNM) based on preoperative multimodal ultrasound imaging features fusion of primary papillary thyroid carcinoma (PTC).

Methods

In total, 498 cases of unifocal PTC were randomly divided into two sets which comprised 348 cases (training set) and 150 cases (validition set). In addition, the testing set contained 120 cases of PTC at different times. Post-operative histopathology was the gold standard for CLNM. The following steps were used to build models: the regions of interest were segmented in PTC ultrasound images, multimodal ultrasound image features were then extracted by the deep learning residual neural network with 50-layer network, followed by feature selection and fusion; subsequently, classification was performed using three classical classifiers-adaptive boosting (AB), linear discriminant analysis (LDA), and support vector machine (SVM). The performances of the unimodal models (Unimodal-AB, Unimodal-LDA, and Unimodal-SVM) and the multimodal models (Multimodal-AB, Multimodal-LDA, and Multimodal-SVM) were evaluated and compared.

Results

The Multimodal-SVM model achieved the best predictive performance than the other models (P < 0.05). For the Multimodal-SVM model validation and testing sets, the areas under the receiver operating characteristic curves (AUCs) were 0.910 (95% CI, 0.894-0.926) and 0.851 (95% CI, 0.833-0.869), respectively. The AUCs of the Multimodal-SVM model were 0.920 (95% CI, 0.881-0.959) in the cN0 subgroup-1 cases and 0.828 (95% CI, 0.769-0.887) in the cN0 subgroup-2 cases.

Conclusion

The ultrasound radiomics model only based on the PTC multimodal ultrasound image have high clinical value in predicting CLNM and can provide a reference for treatment decisions.

Diagnosis of Metastatic Lymph Nodes in Patients With Papillary Thyroid Cancer: A Comparative Multi-Center Study of Semantic Features and Deep Learning-Based Models

OBJECTIVES

Deep learning algorithms have shown potential in streamlining difficult clinical decisions. In the present study, we report the diagnostic profile of a deep learning model in differentiating malignant and benign lymph nodes in patients with papillary thyroid cancer.

METHODS

An in-house deep learning-based model called "ClymphNet" was developed and tested using two datasets containing ultrasound images of 195 malignant and 178 benign lymph nodes. An expert radiologist also viewed these ultrasound images and extracted qualitative imaging features used in routine clinical practice. These signs were used to train three different machine learning algorithms. Then the deep learning model was compared with the machine learning models on internal and external validation datasets containing 22 and 82 malignant and 20 and 76 benign lymph nodes, respectively.

RESULTS

Among the three machine learning algorithms, the support vector machine model (SVM) outperformed the best, reaching a sensitivity of 91.35%, specificity of 88.54%, accuracy of 90.00%, and an area under the curve (AUC) of 0.925 in all cohorts. The ClymphNet performed better than the SVM protocol in internal and external validation, achieving a sensitivity of 93.27%, specificity of 92.71%, and an accuracy of 93.00%, and an AUC of 0.948 in all cohorts.

CONCLUSION

A deep learning model trained with ultrasound images outperformed three conventional machine learning algorithms fed with qualitative imaging features interpreted by radiologists. Our study provides evidence regarding the utility of ClymphNet in the early and accurate differentiation of benign and malignant lymphadenopathy.

Artificial intelligence in thyroid ultrasound

Artificial intelligence (AI), particularly deep learning (DL) algorithms, has demonstrated remarkable progress in image-recognition tasks, enabling the automatic quantitative assessment of complex medical images with increased accuracy and efficiency. AI is widely used and is becoming increasingly popular in the field of ultrasound. The rising incidence of thyroid cancer and the workload of physicians have driven the need to utilize AI to efficiently process thyroid ultrasound images. Therefore, leveraging AI in thyroid cancer ultrasound screening and diagnosis cannot only help radiologists achieve more accurate and efficient imaging diagnosis but also reduce their workload. In this paper, we aim to present a comprehensive overview of the technical knowledge of AI with a focus on traditional machine learning (ML) algorithms and DL algorithms. We will also discuss their clinical applications in the ultrasound imaging of thyroid diseases, particularly in differentiating between benign and malignant nodules and predicting cervical lymph node metastasis in thyroid cancer. Finally, we will conclude that AI technology holds great promise for improving the accuracy of thyroid disease ultrasound diagnosis and discuss the potential prospects of AI in this field.

An Ultrasound-based Prediction Model for Occult Contralateral Papillary Thyroid Carcinoma in Adolescents and Young Adults

RATIONALE AND OBJECTIVES

To investigate the occult contralateral papillary thyroid carcinoma (PTC)-associated ultrasound (US) and clinical characteristics and establish a US-based model for the prediction of occult contralateral carcinoma in adolescents and young adults (AYAs) who were diagnosed with unilateral thyroid carcinoma preoperatively.

MATERIALS AND METHODS

From January 2015 to December 2020, patients who were diagnosed with unilateral thyroid carcinoma by preoperative US examination and underwent total thyroidectomy or thyroid lobectomy with more than 60 months of US follow-up at our hospital were retrospectively collected. Univariate and multivariate analyses were applied to identify the independent risk factors associated with occult contralateral PTC in AYAs, on which a prediction model was developed. The performance of the model was evaluated with accuracy, sensitivity, specificity, and the area under the receiver operating characteristic curve.

RESULTS

Occult contralateral PTC was found in 91 of 365 (24.9%) PTC patients with a median age at diagnosis of 26 years (interquartile range, 24-29 years). The multivariate analysis indicated that the presence of contralateral benign nodule, intra-tumoral calcification, and intraglandular dissemination were significantly associated with occult contralateral PTC in AYAs. The prediction model, which incorporated all independent predictors, yielded an area under the receiver operating characteristic curve of .661 (95% CI: .602-.719). The accuracy, sensitivity and specificity were 67.9%, 54.9%, and 72.3%, respectively.

CONCLUSION

The US-based prediction model proposed here exhibited a favorable performance for predicting occult contralateral PTC, which might be used to determine the appropriate extent of surgery for AYAs who had a preoperative diagnosis of unilateral thyroid carcinoma.

Paediatric differentiated thyroid carcinoma: a UK National Clinical Practice Consensus Guideline

This guideline is written as a reference document for clinicians presented with the challenge of managing paediatric patients with differentiated thyroid carcinoma up to the age of 19 years. Care of paediatric patients with differentiated thyroid carcinoma differs in key aspects from that of adults, and there have been several recent developments in the care pathways for this condition; this guideline has sought to identify and attend to these areas. It addresses the presentation, clinical assessment, diagnosis, management (both surgical and medical), genetic counselling, follow-up and prognosis of affected patients. The guideline development group formed of a multi-disciplinary panel of sub-speciality experts carried out a systematic primary literature review and Delphi Consensus exercise. The guideline was developed in accordance with The Appraisal of Guidelines Research and Evaluation Instrument II criteria, with input from stakeholders including charities and patient groups. Based on scientific evidence and expert opinion, 58 recommendations have been collected to produce a clear, pragmatic set of management guidelines. It is intended as an evidence base for future optimal management and to improve the quality of clinical care of paediatric patients with differentiated thyroid carcinoma.

Radiomics Profiling Identifies the Value of CT Features for the Preoperative Evaluation of Lymph Node Metastasis in Papillary Thyroid Carcinoma

Background: The aim of this study was to identify the increased value of integrating computed tomography (CT) radiomics analysis with the radiologists’ diagnosis and clinical factors to preoperatively diagnose cervical lymph node metastasis (LNM) in papillary thyroid carcinoma (PTC) patients. Methods: A total of 178 PTC patients were randomly divided into a training (n = 125) and a test cohort (n = 53) with a 7:3 ratio. A total of 2553 radiomic features were extracted from noncontrast, arterial contrast-enhanced and venous contrast-enhanced CT images of each patient. Principal component analysis (PCA) and Pearson’s correlation coefficient (PCC) were used for feature selection. Logistic regression was employed to build clinical–radiological, radiomics and combined models. A nomogram was developed by combining the radiomics features, CT-reported lymph node status and clinical factors. Results: The radiomics model showed a predictive performance similar to that of the clinical–radiological model, with similar areas under the curve (AUC) and accuracy (ACC). The combined model showed an optimal predictive performance in both the training (AUC, 0.868; ACC, 86.83%) and test cohorts (AUC, 0.878; ACC, 83.02%). Decision curve analysis demonstrated that the combined model has good clinical application value. Conclusions: Embedding CT radiomics into the clinical diagnostic process improved the diagnostic accuracy. The developed nomogram provides a potential noninvasive tool for LNM evaluation in PTC patients.

Deep multimodal learning for lymph node metastasis prediction of primary thyroid cancer

Objective. The incidence of primary thyroid cancer has risen steadily over the past decades because of overdiagnosis and overtreatment through the improvement in imaging techniques for screening, especially in ultrasound examination. Metastatic status of lymph nodes is important for staging the type of primary thyroid cancer. Deep learning algorithms based on ultrasound images were thus developed to assist radiologists on the diagnosis of lymph node metastasis. The objective of this study is to integrate more clinical context (e.g., health records and various image modalities) into, and explore more interpretable patterns discovered by, deep learning algorithms for the prediction of lymph node metastasis in primary thyroid cancer patients.Approach. A deep multimodal learning network was developed in this study with a novel index proposed to compare the contribution of different modalities when making the predictions.Main results. The proposed multimodal network achieved an average F1 score of 0.888 and an average area under the receiver operating characteristic curve (AUC) value of 0.973 in two independent validation sets, and the performance was significantly better than that of three single-modality deep learning networks. Moreover, among three modalities used in this study, the deep multimodal learning network relied generally more on image modalities than the data modality of clinic records when making the predictions.Significance. Our work is beneficial to prospective clinic trials of radiologists on the diagnosis of lymph node metastasis in primary thyroid cancer, and will better help them understand how the predictions are made in deep multimodal learning algorithms.

Application of Machine Learning Methods to Improve the Performance of Ultrasound in Head and Neck Oncology: A Literature Review

Simple Summary

Ultrasound (US) is a non-invasive imaging method that is routinely utilized in head and neck cancer patients to assess the anatomic extent of tumors, nodal and non-nodal neck masses and for imaging the salivary glands. In this review, we summarize the present evidence on whether the application of machine learning (ML) methods can potentially improve the performance of US in head and neck cancer patients. We found that published clinical literature on ML methods applied to US datasets was limited but showed evidence of improved diagnostic and prognostic performance. However, a majority of these studies were based on retrospective evaluation and conducted at a single center with a limited number of datasets. The conduct of multi-center studies could help better validate the performance of ML-based US radiomics and facilitate the integration of these approaches into routine clinical practice.

Abstract

Radiomics is a rapidly growing area of research within radiology that involves the extraction and modeling of high-dimensional quantitative imaging features using machine learning/artificial intelligence (ML/AI) methods. In this review, we describe the published clinical evidence on the application of ML methods to improve the performance of ultrasound (US) in head and neck oncology. A systematic search of electronic databases (MEDLINE, PubMed, clinicaltrials.gov) was conducted according to Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) guidelines. Of 15,080 initial articles identified, 34 studies were selected for in-depth analysis. Twenty-five out of 34 studies (74%) focused on the diagnostic application of US radiomics while 6 (18%) studies focused on response assessment and 3 (8%) studies utilized US radiomics for modeling normal tissue toxicity. Support vector machine (SVM) was the most commonly employed ML method (47%) followed by multivariate logistic regression (24%) and k-nearest neighbor analysis (21%). Only 11/34 (~32%) of the studies included an independent validation set. A majority of studies were retrospective in nature (76%) and based on single-center evaluation (85%) with variable numbers of patients (12–1609) and imaging datasets (32–1624). Despite these limitations, the application of ML methods resulted in improved diagnostic and prognostic performance of US highlighting the potential clinical utility of this approach.

A Proposed Heterogeneous Ensemble Algorithm Model for Predicting Central Lymph Node Metastasis in Papillary Thyroid Cancer

Purpose

To develop a heterogeneous ensemble algorithm model to precisely predict central lymph node metastasis (CLNM), which can provide a reference value on controversial topics of performing prophylactic central lymph node dissection for patients with papillary thyroid cancer (PTC).

Methods

The study included patients with PTC who underwent an initial thyroid resection in a single-center medical institution between January 2014 and December 2018. A total of 18 variables, including clinical features and ultrasound (US) features, were used in the univariate analysis, multivariate analysis, and feature selection and were also used to develop a heterogeneous ensemble model based on five basic machine learning models, including extreme gradient boosting, k-nearest neighbors, random forest, gradient boosting, and AdaBoost. Moreover, a partial dependent plot was used to explain the heterogeneous ensemble model.

Results

The area under the receiver operating characteristic curve of the heterogeneous ensemble algorithm model was 0.67, which is significantly better than that of the basic machine models in predicting CLNM. All machine learning models performed better than US. Based on multivariate analysis and receiver operating characteristic curve analysis, age ≤33 years, tumor size ≥0.8 cm, US-suspected CLNM, and microcalcification were risk factors for CLNM, and anti-thyroid peroxidase antibody and serum thyroglobulin levels were favorable factors for CLNM.

Conclusion

The proposed heterogeneous ensemble algorithm model may be optimal tool to predict CLNM by integrating clinical and US features.

Ultrasound-Based Radiomic Nomogram for Predicting Lateral Cervical Lymph Node Metastasis in Papillary Thyroid Carcinoma

RATIONALE AND OBJECTIVES

Accurate preoperative identification of lateral cervical lymph node metastasis (LNM) is important for decision-making and clinical management of patients with papillary thyroid carcinoma (PTC). The aim of this study was to develop an ultrasound (US)-based radiomic nomogram to preoperatively predict the lateral LNM in PTC patients.

METHODS

In this retrospective study, a total of 886 patients were enrolled and randomly divided into 2 groups. Radiomic features were extracted from the preoperative US images. A radiomic signature was constructed using the least absolute shrinkage and selection operator algorithm in the training set. Multivariate logistic regression was performed to develop the radiomic nomogram, which incorporating the radiomic signature and the selected clinical characteristics. The performance of the nomogram was assessed by its discrimination, calibration, and clinical usefulness in both the training and validation sets.

RESULTS

The radiomic signature was significantly associated with the lateral LNM in both cohorts (p< 0.001). The nomogram that consisted of radiomic signature, US-reported cervical lymph node (CLN) status, and CT-reported CLN status demonstrated good discrimination and calibration in the training and validation sets with an AUC of 0.946 and 0.914, respectively. The decision curve analysis indicated that the radiomic nomogram was worthy of clinical application.

CONCLUSION

The radiomic nomogram proposed here has good performance for noninvasively predicting the lateral LNM and might be used to facilitate clinical decision-making and potentially improve the survival outcome in selected patients.

Using ultrasound features and radiomics analysis to predict lymph node metastasis in patients with thyroid cancer

Background

Lymph node metastasis (LNM) is an important factor for thyroid cancer patients’ treatment and prognosis. The aim of this study was to explore the clinical value of ultrasound features and radiomics analysis in predicting LNM in thyroid cancer patients before surgery.

Methods

The characteristics of ultrasound images of 150 thyroid nodules were retrospectively analysed. All nodules were confirmed as thyroid cancer. Among the assessed patients, only one hundred and twenty-six patients underwent lymph node dissection. All patients underwent an ultrasound examination before surgery. In the radiomic analysis, the area of interest was identified from selected ultrasound images by using ITK-SNAP software. The radiomic features were extracted by using Ultrosomics software. Then, the data were classified into a training set and a validation set. Hypothetical tests and bagging were used to build the model. The diagnostic performance of different ultrasound features was assessed, a radiomic analysis was conducted, and a receiver operating characteristic (ROC) curve analysis was performed to explore the diagnostic accuracy.

Results

Regarding the prediction of LNM, the ROC curves showed that the area under the curve (AUC) values of an irregular shape and microcalcification were 0.591 (P = 0.059) and 0.629 (P = 0.007), respectively. In the radiomics analysis, in the training set, the AUC value of LNM was 0.759, with a sensitivity of 0.90 and a specificity of 0.860. In the verification set, the AUC was 0.803, with a sensitivity of 0.727 and a specificity of 0.800.

Conclusions

Microcalcification and an irregular shape are predictors of LNM in thyroid carcinoma patients. In addition, radiomics analysis has promising value in screening meaningful ultrasound features in thyroid cancer patients with LNM. Therefore, the prediction of LNM based on ultrasound features and radiomic features is useful for making appropriate decisions regarding surgery and interventions before thyroid carcinoma surgery.

The Diagnostic Efficiency of Ultrasound Computer-Aided Diagnosis in Differentiating Thyroid Nodules: A Systematic Review and Narrative Synthesis

Computer-aided diagnosis (CAD) techniques have emerged to complement qualitative assessment in the diagnosis of benign and malignant thyroid nodules. The aim of this review was to summarize the current evidence on the diagnostic performance of various ultrasound CAD in characterizing thyroid nodules. PUBMED, EMBASE and Cochrane databases were searched for studies published until August 2019. The Quality Assessment of Studies of Diagnostic Accuracy included in Systematic Review 2 (QUADAS-2) tool was used to assess the methodological quality of the studies. Reported diagnostic performance data were analyzed and discussed. Fourteen studies with 2232 patients and 2675 thyroid nodules met the inclusion criteria. The study quality based on QUADAS-2 assessment was moderate. At best performance, grey scale CAD had a sensitivity of 96.7% while Doppler CAD was 90%. Combined techniques of qualitative grey scale features and Doppler CAD assessment resulted in overall increased sensitivity (92%) and optimal specificity (85.1%). The experience of the CAD user, nodule size and the thyroid malignancy risk stratification system used for interpretation were the main potential factors affecting diagnostic performance outcomes. The diagnostic performance of CAD of thyroid ultrasound is comparable to that of qualitative visual assessment; however, combined techniques have the potential for better optimized diagnostic accuracy.

CAD system based on B-mode and color Doppler sonographic features may predict if a thyroid nodule is hot or cold

OBJECTIVES

The aim of this study was to evaluate if the analysis of sonographic parameters could predict if a thyroid nodule was hot or cold.

METHODS

Overall, 102 thyroid nodules, including 51 hyperfunctioning (hot) and 51 hypofunctioning (cold) nodules, were evaluated in this study. Twelve sonographic features (i.e., seven B-mode and five Doppler features) were extracted for each nodule type. The isthmus thickness, nodule volume, echogenicity, margin, internal component, microcalcification, and halo sign features were obtained in the B-mode, while the vascularity pattern, resistive index (RI), peak systolic velocity, end diastolic velocity, and peak systolic/end diastolic velocity ratio (SDR) were determined, based on Doppler ultrasounds. All significant features were incorporated in the computer-aided diagnosis (CAD) system to classify hot and cold nodules.

RESULTS

Among all sonographic features, only isthmus thickness, nodule volume, echogenicity, RI, and SDR were significantly different between hot and cold nodules. Based on these features in the training dataset, the CAD system could classify hot and cold nodules with an area under the curve (AUC) of 0.898. Also, in the test dataset, hot and cold nodules were classified with an AUC of 0.833.

CONCLUSIONS

2D sonographic features could differentiate hot and cold thyroid nodules. The CAD system showed a great potential to achieve it automatically.

KEY POINTS

• Cold nodules represent higher volume (p = 0.005), isthmus thickness (p = 0.035), RI (p = 0.020), and SDR (p = 0.044) and appear hypoechogenic (p = 0.010) in US. • Nodule volume with an AUC of 0.685 and resistive index with an AUC of 0.628 showed the highest classification potential among all B-mode and Doppler features respectively. • The proposed CAD system could distinguish hot nodules from cold ones with an AUC of 0.833 (sensitivity 90.00%, specificity 70.00%, accuracy 80.00%, PPV 87.50%, and NPV 75.00%).