Mode
Text Size
Log in / Sign up

CEUS combined with AI or elastography improves diagnostic performance for malignant thyroid nodulesContrast-Enhanced Ultrasound Shows Promise for Identifying Thyroid Nodules

AI-generated summary of the cited source, checked by automated accuracy review. How we work

Key Takeaway
Consider CEUS combined with AI or elastography to improve diagnostic performance in identifying malignant thyroid nodules.

This network meta-analysis evaluated the diagnostic performance of contrast-enhanced ultrasound (CEUS) when combined with three different modalities: ultrasound elastography (UE), artificial intelligence (AI), and conventional ultrasound (US) for differentiating benign from malignant thyroid nodules. The analysis included 5022 thyroid nodules from 4484 patients.

The study found that CEUS-UE demonstrated a superiority index (SI) of 3.47, with a pooled sensitivity of 0.92 and a pooled specificity of 0.83. CEUS-AI showed an SI of 3.81, with a pooled sensitivity of 0.95 and a pooled specificity of 0.81, which may offer value in identifying high-risk nodules. CEUS-US showed an SI of 5.29, with a pooled sensitivity of 0.89 and a pooled specificity of 0.88, which is effective for non-risk-stratified nodules.

Authors noted that these subgroup findings require further validation in larger, diverse external cohorts. Clinically, CEUS-UE shows favorable overall performance, while CEUS-AI and CEUS-US may offer specific utility depending on the clinical need for risk stratification or general diagnosis. The evidence is currently limited by the need for broader cohort validation.

How this fits prior evidence

This network meta-analysis extends prior evidence showing that contrast-enhanced ultrasound (CEUS) has 81.6% diagnostic accuracy for differentiating benign from malignant thyroid nodules. The current findings provide more specific performance metrics when CEUS is combined with other modalities, such as AI or elastography, to improve diagnostic accuracy or risk stratification.

Researchers analyzed data from over 5,000 thyroid nodules to see how different imaging techniques perform. They specifically looked at how contrast-enhanced ultrasound (CEUS) works when combined with other methods like ultrasound elastography (UE), artificial intelligence (AI), and conventional ultrasound (US). The goal was to see how well these combinations could tell the difference between benign and malignant nodules.

The study found that combining CEUS with ultrasound elastography showed favorable results for overall diagnosis. When CEUS was paired with artificial intelligence, it showed potential for identifying high-risk nodules. Additionally, combining CEUS with conventional ultrasound was found to be effective for identifying nodules that do not require specific risk stratification.

Because this was a network meta-analysis, the results are based on existing data rather than a new clinical trial. The findings are promising but need more validation in larger, diverse groups of patients before they can change standard medical practice. Patients should discuss these imaging options with their doctors to determine the best approach for their specific situation.

What this means for you:
Combining contrast-enhanced ultrasound with other imaging tools may help identify thyroid nodules more accurately.

Common questions

How does contrast-enhanced ultrasound help with thyroid nodules?

When combined with ultrasound elastography, contrast-enhanced ultrasound (CEUS) showed a pooled sensitivity of 0.92 and a specificity of 0.83. When paired with artificial intelligence, it showed a sensitivity of 0.95. These combinations help doctors better distinguish between benign and malignant nodules.

Is CEUS combined with AI effective for high-risk cases?

The study suggests that combining CEUS with artificial intelligence (AI) may offer value in identifying high-risk thyroid nodules. This specific combination showed a pooled sensitivity of 0.95 and a specificity of 0.81 in the data analyzed.

What is the difference between using CEUS with elastography versus conventional ultrasound?

CEUS combined with ultrasound elastography (UE) showed favorable overall diagnostic performance. In contrast, CEUS combined with conventional ultrasound (US) was found to be effective specifically for identifying non-risk-stratified thyroid nodules.

Study Details

Study typeMeta analysis
EvidenceLevel 1
PublishedSep 2026
View Original Abstract ↓
BackgroundThis network meta-analysis (NMA) intends to fully assess the value of contrast-enhanced ultrasound (CEUS) combined with other imaging techniques in the diagnosis of thyroid nodules (TNs), and provide clinical guidance.MethodsWe systematically searched PubMed, Web of Science, Embase, and Cochrane up to October 15, 2025, and assessed study quality by the QUADAS-2 tool. The combination of CEUS and CEUS with other techniques was assessed for accuracy by R4.5.1 and ANOVA-based NMA.ResultsTwenty-six original studies were included, encompassing 5022 TNs from 4484 patients. CEUS-ultrasound elastography (UE) demonstrated favorable diagnostic performance in differentiating benign and malignant TNs (superiority index [SI]: 3.47, 95%CI 0.2–7). The pooled sensitivity (SE) and specificity (SP) of CEUS-UE were 0.92 (95%CI 0.88–0.95, SUCRA 80.0%) and 0.83 (95%CI 0.75–0.89, SUCRA 60.7%), respectively. CEUS-artificial intelligence (AI) was generally effective in identifying high-risk TNs (SI: 3.81, 95%CI 0.33–9), with pooled SE and SP of 0.95 (95%CI 0.91–0.98, SUCRA 90.0%) and 0.81 (95%CI 0.68–0.91, SUCRA 49.6%), respectively. CEUS-conventional ultrasound (US) showed favorable diagnostic performance in identifying non-risk-stratified TNs (SI: 5.29, 95%CI 0.33–9), with pooled SE and SP of 0.89 (95% CI 0.74–0.97, SUCRA 61.4%) and 0.88 (95%CI 0.69–0.97, SUCRA 76.9%), respectively.ConclusionCEUS-UE demonstrates favorable overall diagnostic performance in differentiating benign and malignant TNs. CEUS-AI may offer additional value in identifying high-risk TNs with multiple suspicious features. CEUS-US is effective in identifying non-risk-stratified TNs. However, these subgroup findings require further validation in larger, diverse external cohorts.Systematic Review Registrationhttps://www.crd.york.ac.uk/PROSPERO/myprospero, identifier CRD420251169562.
Free Newsletter

Clinical research that matters. Delivered to your inbox.

Join thousands of clinicians and researchers. No spam, unsubscribe anytime.