Ultrasound-based techniquesConventional B-mode ultrasound

Conventional B-mode ultrasound (US) remains one of the most widely used tools for liver imaging, largely because it is inexpensive, broadly accessible, and easily incorporated into routine clinical practice [8]. The technique generates two-dimensional grayscale images based on the reflection of US waves at tissue interfaces, allowing for visualization of liver parenchymal characteristics (Fig. 2) [9]. Steatosis is typically suggested by several well-recognized sonographic signs, including increased parenchymal echogenicity, increased contrast between the liver and the right renal cortex, and deep beam attenuation, leading to loss of sharpness and reduced visualization of the intrahepatic vasculature, particularly in the deep segments, as well as reduced visibility of the gallbladder wall [8, 9]. The presence of one or more of these features is commonly interpreted as an indication of intracellular lipid accumulation, i.e., hepatic steatosis [9].

Fig. 2

B-mode abdominal ultrasound. Longitudinal subcostal view of the right upper quadrant. The liver parenchyma appears diffusely hyperechoic with posterior acoustic attenuation, consistent with hepatic steatosis. The right kidney is visualized for comparison, demonstrating relative cortical hypoechogenicity compared with the liver

Although B-mode US is easy to perform, well tolerated, and acceptably reproducible, it has certain limitations [9]. It reliably detects moderate to severe steatosis (≥ 30% fat content) but performs far less effectively in the setting of mild steatosis (< 30%), in which it displays lower accuracy [9, 10]. Diagnostic accuracy is further diminished in patients living with obesity because it is affected by abdominal adiposity; excess subcutaneous fat attenuates the US beam, thus degrading image quality and diagnostic accuracy [10, 11]. Since hepatic steatosis is histologically defined by the presence of fat in ≥ 5% of the hepatocytes, conventional US often fails to identify mild steatosis or early-stage disease [12]. In order to improve diagnostic accuracy, semiquantitative grading systems have been developed (e.g., the Hamaguchi score [13] and the US-fatty liver indicator [14]). However, although these techniques have improved consistency, the qualitative nature of B-mode US remains a certain limitation. Operator skill, variations in technique, and subjective interpretation introduce substantial variability and limit its reproducibility; thus, B-mode US is regarded as a fundamentally subjective technique [11, 15].

Quantitative ultrasound techniques for hepatic steatosis evaluation

In order to overcome the limitations of conventional B-mode US, several quantitative US techniques have been developed aiming at providing more objective, quantitative estimates of hepatic lipid content [16]. These approaches include the attenuation coefficient (AC), the backscatter coefficient (BSC), and the speed-of-sound estimation (SEE) [17]. Among the attenuation-based methods, several have already been integrated into clinical practice, such as the controlled attenuation parameter (CAP), the ultrasound-guided attenuation parameter (UGAP), and attenuation imaging (ATI). These techniques estimate fat accumulation by analyzing the loss of US energy as radiofrequency (RF) signals pass through the liver. The signal attenuation varies and is notably increased in steatotic livers. In particular, the BSC measures the proportion of incident US energy scattered back to the transducer [17].

More recently, ultrasound-derived fat fraction (UDFF) has emerged as an advanced quantitative US technique that uses raw RF data to estimate hepatic fat content as a percentage, analyzing frequency-dependent attenuation and backscatter properties of liver tissue [16]. The following subsections summarize the principles, diagnostic performance, and limitations of the most clinically relevant quantitative US techniques.

Controlled attenuation parameter (CAP)

The CAP is obtained during the vibration-controlled transient elastography (VCTE) and provides a rapid, point-of-care estimate of hepatic US attenuation, which is a surrogate for steatosis [3]. Systematic reviews and meta-analyses, using magnetic resonance imaging-proton density fat fraction (MRI-PDFF) as reference standard, reported acceptable diagnostic performance of CAP, with area under the receiver operating characteristic curve (AUROC) of 0.85–0.88 for detecting ≥ 5–10% liver fat, thereby supporting its suitability for screening applications [18]. Commonly proposed thresholds are the following: 248 dB/m for any steatosis (S ≥ 1), 268 dB/m for moderate steatosis (S ≥ 2), and 280 dB/m for severe steatosis (S3); however, the optimal thresholds vary across studies depending on specific population characteristics, probe size (M vs. XL), and reference standards [19, 20]. Several parameters may influence CAP, including body mass index (BMI), metabolic status (i.e., hyperglycemia, hypertriglyceridemia, or low high-density lipoprotein-cholesterol), underlying liver disease, and measurement quality; regarding the quality, the expertise of the operator, the appropriate use of adjusted cutoffs and appropriate probe selection are considered to be highly important [19, 21, 22]. While CAP reliably identifies the presence of steatosis, its capability to differentiate among different grades of steatosis (i.e., S2 and S3) is inferior to that of MRI-PDFF or some newer quantitative US modalities [23]. In clinical practice, CAP is widely used for MASLD screening, population-level assessments, and longitudinal follow-up (e.g., CAP is decreased after interventions, such as weight loss or bariatric surgery) [24, 25]. Newer CAP-based composite scores and disease-specific algorithms have also emerged, although external validation and standardized thresholds are still evolving [26, 27].

Ultrasound-guided attenuation parameter (UGAP)

The UGAP and other techniques to follow are imaging-based techniques that differ from CAP, as they provide real-time liver visualization, which allows for optimal positioning of regions-of-interest (ROI) when measuring hepatic steatosis. Pitfalls involved in measuring ascites or focal tumors can thus be avoided. More than one segment can be evaluated in terms of steatosis thanks to the imaging visualization available with these techniques. The UGAP is a quantitative B-mode-based technique that measures the AC within a defined ROI, providing a pixel-wise index of steatosis, applicable on standard clinical scanners (Fig. 3) [28]. Validation studies using MRI-PDFF or liver biopsy as reference standards consistently report strong correlation and high accuracy for clinically relevant steatosis thresholds (e.g., a cutoff value of UGAP > 0.75 dB/cm/MHz provided AUROC ~ 0.8 for ≥S2) [29]. Proposed thresholds differ among different studies and devices but commonly fall around 0.60 for the detection of any steatosis (i.e., ≥S1), progressing to higher values for more severe disease stages [30, 31].

Fig. 3

UGAP imaging for quantitative liver fat assessment. Grayscale ultrasound overlaid with attenuation coefficient maps. (a) Multiple axial measurements within the predefined ROI demonstrate elevated attenuation values, supporting the presence of mild hepatic steatosis (S1). (b) Increased and heterogeneous attenuation values within the predefined ROI indicate higher hepatic fat content, compatible with S2 steatosis. (c) Markedly increased attenuation values within the predefined ROI reflect high hepatic fat content, supporting the diagnosis of S3 steatosis. Abbreviations: ROI, region of interest; UGAP, ultrasound-guided attenuation parameter

Reproducibility is generally high when acquisition steps are followed, though factors such as BMI, advanced fibrosis, scanner settings, and probe characteristics may adversely influence the performance [29, 32]. Overall, the UGAP provides a reliable, non-invasive assessment that outperforms qualitative US for moderate to severe steatosis and correlates well with MRI-PDFF. Its broader clinical implementation, however, will depend on its multicenter standardization and inter-vendor harmonization with different technologies, which should be based on further studies [30, 33].

Attenuation imaging (ATI)

Similarly to other attenuation-based methods, ATI quantifies frequency-dependent US attenuation in liver tissue and reports an AC that correlates well with MRI-PDFF and liver histology [34]. Relevant multicenter studies indicate that ATI may outperform CAP in detecting and grading steatosis; for example, AUROC values of 0.895 vs. 0.845 for ≥S1, 0.944 vs. 0.881 for ≥S2, and 0.928 vs. 0.860 for S3 have been reported in comparative studies for CAP vs. ATI, respectively [35]. Diagnostic thresholds, though, vary based on the methodology of the study and the device used. Typically, AC cutoffs for any steatosis (S > 0) vary 0.59–0.71 dB/cm/MHz, depending on the specific device and acquisition settings [36, 37]. Measurement may also be influenced by both technical and patient parameters, such as measurement depth, ROI characteristics (i.e., placement and size), the amount of subcutaneous fat, and confidence-map quality criteria. In this regard, some authors have reported that only three valid measurements may be sufficient when robust quality criteria are applied [38, 39]. Therefore, ATI is currently considered to be a reproducible, MRI-correlated technique for quantitative hepatic fat assessment when device-specific cutoffs and quality criteria are used. Broader implementation, however, requires additional standardization and validation across multiple vendors [37, 40].

Backscatter coefficient (BSC)

The BSC reflects the proportion of US energy scattered by tissue microarchitecture, providing information about the scattered density and composition. Animal studies combining AC and BSC measurements demonstrated high accuracy for steatosis grading, supporting BSC as a promising early diagnostic tool [41]. Likewise, clinical and ex vivo human studies indicate that hepatic steatosis is associated with substantial increase in BSC values compared with non-steatotic liver at similar frequencies, although results depend on measurement depth, ROI selection, and attenuation correction [42, 43]. Ongoing challenges include standardization for inter-platform variability, the development of standard protocols, as well as the need for accurate compensation for depth-dependent attenuation [41, 43]. Nevertheless, the World Federation for Ultrasound in Medicine and Biology (WFUMB) has recently recognized BSC as an emerging tool for non-invasive liver fat assessment alongside AC, reflecting its growing clinical interest [37]. BSC is currently incorporated into commercially available techniques, such as UDFF and ultrasound-guided fat fraction (UGFF).

Ultrasound-derived fat fraction (UDFF)

The UDFF, simultaneously incorporating both AC and BSC data, provides a percentage-based estimation of liver fat content (Fig. 4) [44]. Clinical studies have reported sufficient reproducibility and correlation with MRI-PDFF measurements; however, inter-platform standardization remains an area of active investigation [44,45,46]. Inter-vendor differences represent a current challenge, likely driven by variations in hardware, signal processing algorithms, attenuation and/or backscatter correction, and ROI characterization [47]. Despite these technical considerations, UDFF may emerge as a practical, user-friendly option for quantitatively assessing hepatic fat in clinical practice when standardized protocols are applied [44, 47, 48]. One of the latest advances in the field of these technologies is UGFF, which combines the following: (i) attenuation coefficient, (ii) backscatter coefficient, and (iii) signal-to-noise ratio, providing a quantitative value of fat fraction as a percentage. In a first publication in December 2025, Popa et al. concluded that UGFF is effective and consistent across different breathing conditions and that it exhibited strong correlation with CAP and high accuracy (AUC 0.89–0.91) for the detection of hepatic steatosis. The cutoffs of 4% and 11% were proposed for S1 and S3 [48].

Fig. 4

Gray-scale ultrasound with UDFF, avoiding vessels and biliary structures. (a) The measured UDFF value is 4% at a depth of 4.6 cm, consistent with absence of hepatic steatosis (S0). (b) The measured UDFF value is 10% at a depth of 5.1 cm, consistent with mild steatosis (S1). (c) The measured UDFF value is 20% at a depth of 5.7 cm, consistent with moderate steatosis (S2). (d) The measured UDFF value is 22% at a depth of 5.0 cm, consistent with severe steatosis (S3). Abbreviations: UDFF, ultrasound-derived fat fraction

Speed-of-sound estimation (SEE)

SSE is another emerging technique that measures US wave velocity in liver tissue [49]. As fat accumulation increases, the SSE decreases, thereby providing a continuous range of liver fat content. Validation studies have demonstrated acceptable correlation of SEE with both MRI-PDFF and liver histological, suggesting potential utility for detection and longitudinal monitoring of hepatic steatosis [50]. However, SSE to date remains mostly experimental and is not widely implemented. Further multicenter validation, standardization, and protocol harmonization are necessary before routine clinical adoption may be considered [51].

Ultrasound elastography

US elastography assesses the mechanical properties of liver tissue and provides liver stiffness measurements (LSMs), serving mainly as a non-invasive surrogate for hepatic fibrosis and, secondarily, for viscosity or hepatic steatosis, depending on the technique. As hepatic fibrosis is considered a major prognostic factor for advanced disease and related mortality and a definite target for the management of the disease [52], US elastography techniques are considered to be essential for risk stratification and longitudinal monitoring of patients with MASLD; these techniques include strain elastography, shear wave elastography (SWE), and vibration-controlled transient elastography (VCTE) [53].

Strain elastography (SE)

SE evaluates relative tissue displacement in response to externally applied compression of the ultrasound probe [54]. Since the applied force cannot be standardized, the technique does not yield quantitative stiffness values, which greatly limits its reproducibility and reliability. As a result, strain elastography is currently infrequently used in hepatic fibrosis assessment [54, 55].

Shear wave elastography (SWE)

SWE is widely incorporated into modern US systems and has become a key technique for non-invasive assessment of hepatic fibrosis. It generates localized tissue displacements via acoustic radiation force, producing transverse shear waves, whose propagation velocity corresponds to underlying tissue stiffness [54, 56]. Depending on the setup, SWE can be performed at a single focal point [i.e., point SWE (pSWE)] or across a broad area of up to several cm2 to produce a two-dimensional (2D) elastogram (i.e., 2D-SWE) [56].

pSWE measures shear wave velocity within a single ROI and reports stiffness in m/s or kPa [57]. It is widely available and easy to integrate into standard US examinations, while it provides good reproducibility when appropriate quality criteria are used. Relevant meta-analyses consistently show AUROC values > 0.80 for detecting significant fibrosis (F2), indicating acceptable diagnostic accuracy [58,59,60]. Nonetheless, pSWE performance may be affected by factors such as ROI selection, operator expertise, breathing maneuvers, patient body habitus (e.g., obesity, muscularity, and cachexia), and vendor-specific implementation, underscoring the importance of standardized acquisition protocols [61].

2D-SWE expands the sampling area to generate real-time color elastograms, thereby reducing the possibi