MP-Ultrasound of Thyroid Nodule: Cancer Risk According to EFSUMB 2026

What the EFSUMB 2026 Document Changed

EFSUMB Guidelines on Multiparametric Ultrasound Thyroid Nodule Evaluation: Part I, 2026 marks a shift from isolated sonographic classification to multiparametric nodule evaluation. The basic level is grayscale ultrasound with TIRADS stratification. Additional levels include elastography, microvascular flow imaging, contrast-enhanced ultrasound, and AI algorithms. An important practical conclusion: no additional method should 'rewrite' an obviously suspicious B-mode as benign, but it can assist in intermediate and discordant cases.

Starting Point: B-mode and TIRADS

The primary protocol begins with describing the nodule's composition, echogenicity, shape, contours, orientation, echogenic foci, capsular contact, signs of extrathyroidal extension, and regional lymph nodes. These features form the initial risk category. Multiparametric methods should be interpreted only after B-mode: a stiff, hypoenhanced, or chaotically vascularized nodule has different significance in EU-TIRADS 3 and EU-TIRADS 5.

Where Elastography Fits In

Elastography assesses tissue stiffness and is used as an additional marker, especially in low and intermediate-risk nodules. Significant stiffness increase compared to surrounding parenchyma, heterogeneous stiffness of the solid component, and correlation of the stiff area with a suspicious zone in B-mode are considered suspicious. A soft nodule reduces cancer likelihood but does not negate FNA if grayscale features fall into a high-risk category.

EFSUMB 2026 emphasizes the absence of a universal stiffness threshold applicable to all devices, techniques, and nodule types. Therefore, it is better not to turn elastography into an independent diagnosis in the conclusion. A correct formulation: 'elastographically, the nodule is predominantly soft/stiff; the feature agrees/disagrees with the B-mode risk category.'

Technical Pitfalls of Elastography

The result depends on precompression by the probe, breathing and swallowing, depth, nodule size, calcifications, cystic and fibrous components. For strain elastography, compression stability and reference tissue selection are critical. For shear wave elastography, map quality, absence of artifacts, and measurement zone placement in a viable solid component are important. Macrocalcifications, coarse fibrous remodeling, and significant cystic degeneration limit interpretation.

Microvascular Flow-Imaging

Methods for visualizing slow microvascular flow are applied after conventional color or power Doppler. They allow visualization of fine vascular branches without contrast agent. Risk assessment considers not the mere presence of flow but its architecture: a regular peripheral rim often supports a benign scenario, whereas chaotic intranodular vessels, uneven branching, and vascular asymmetry increase suspicion. As with elastography, EFSUMB does not set a single numerical threshold that would classify a nodule as cancerous.

CEUS: What to Look for in a Nodule

Contrast-enhanced ultrasound evaluates nodule perfusion in real-time. The physician compares nodule enhancement with surrounding thyroid tissue, notes contrast arrival time, homogeneity, rim completeness, perfusion defects, and washout characteristics. Hypoenhancement, heterogeneous enhancement, irregular avascular zones, and disrupted vascular architecture are more suspicious for malignancy. Iso- or hyperenhancement with a smooth peripheral rim often supports a benign variant but is not an absolute criterion.

The guideline's key position: CEUS is not a standalone replacement for TIRADS and cytology. Its greatest benefit is clarifying doubtful nodules, selecting a viable area for biopsy in necrosis or cystic transformation, assessing nodules with ambiguous B-mode, and planning interventional procedures.

How to Integrate Features into Risk

A practical algorithm can be described as a sequential increase or decrease in confidence in the initial category. If EU-TIRADS 5 is combined with high stiffness, chaotic microvascularization, and hypoenhancement on CEUS, the risk is considered concordantly high, and the strategy should match the high-risk category. If EU-TIRADS 3 is soft, with regular peripheral blood flow and uniform isoenhancement, cancer likelihood is lower, and observation or FNA is determined by size and clinical context. If features diverge, priority remains with the most validated grayscale criteria and cytology indications.

What to Write in the Protocol

The optimal protocol should be modular: first B-mode with EU-TIRADS category and nodule size, then elastography with qualitative stiffness description, then vascular pattern, then CEUS pattern if the method is performed. In the conclusion, it is useful to indicate whether additional methods are concordant or discordant with TIRADS. Example: 'Left lobe nodule 16 mm, EU-TIRADS 4. Elastography: increased stiffness of the solid component. Microvascular imaging: irregular intranodular blood flow. CEUS: heterogeneous hypoenhancement. The multiparametric picture increases suspicion; FNA is indicated based on risk category and size.'

AI in the EFSUMB Algorithm

AI is considered a tool for standardization and decision support, not as an autonomous arbiter. Algorithms can assist in segmentation, radiomic feature extraction, and TIRADS assessment reproducibility. Limitations remain clinically significant: dependence on training dataset, device, image quality, and external validation. Therefore, AI results should be compared with B-mode, elastography, CEUS, and the clinical task.