Ultrasound Assessment of Liver Fat: Attenuation Coefficient and UDFF
What the WFUMB 2024 Guideline Changed
WFUMB Guideline/Guidance on Liver Multiparametric Ultrasound: Part 2, Fat Quantification marks the transition from qualitative descriptions of a "bright liver" to quantitative fat assessment. The document emphasizes: B-mode is useful for initial suspicion of steatosis but insufficient for reproducible staging, especially in mild steatosis and concomitant fibrosis.
In practical ultrasound diagnostics, several quantitative families are now considered: measurement of the attenuation coefficient, backscatter assessment, speed of sound in tissue, and combined indicators. CAP remains a known method, but the guideline does not limit quantitative assessment to it alone: consoles offer attenuation imaging/attenuation coefficient and algorithms that provide ultrasound-derived fat fraction, including UDFF.
Clinical Scale: What to Base On
Any ultrasound metric should be compared with the biological definition of steatosis. In clinical studies and WFUMB recommendations, the benchmark is the proportion of hepatocytes with macrovesicular fat. Therefore, when interpreting, one should not mix histological percentages, MRI-PDFF, and ultrasound indicators without reservations: they describe similar but not identical phenomena.
| Steatosis Grade | Histological Benchmark | Practical Meaning for Ultrasound |
|---|---|---|
| S0 | <5% hepatocytes with fat | Steatosis not confirmed by histological criteria |
| S1 | 5–33% | Mild steatosis; the zone where B-mode is least reliable |
| S2 | 34–66% | Moderate steatosis; quantitative methods are usually more reproducible |
| S3 | >66% | Severe steatosis; important to assess concomitant fibrosis and inflammation |
Attenuation Coefficient: Physical Meaning
The attenuation coefficient reflects the loss of energy of the ultrasound signal as it passes through the liver. Fat infiltration increases attenuation, so a higher indicator is validation-linked with a greater likelihood of steatosis. The unit of measurement is usually dB/cm/MHz.
The method fundamentally differs from subjective echogenicity assessment: the system analyzes the signal behavior in the selected tissue volume and provides a numerical parameter. This increases reproducibility but does not eliminate all limitations. The result is influenced by depth, acoustic window, parenchyma heterogeneity, rib shadows, vessels, and near-field artifacts.
How to Measure Attenuation in a Routine Protocol
The optimal protocol logic is to first obtain a stable intercostal window of the right lobe, then place the region of interest in homogeneous parenchyma, avoiding large vessels, capsule, bile ducts, focal lesions, and pronounced artifacts. The patient is usually examined fasting, as postprandial changes in hemodynamics and gastrointestinal tract filling worsen acoustic conditions.
Measurement should be accompanied by quality control provided by the specific system: reliability map, stability indicators, measurement series statistics, or other built-in criteria. If the device reports that the measurement is unreliable, the numerical value should not be transferred to the conclusion as a diagnostic fact.
UDFF: What the Percent Fat Fraction Means
UDFF — ultrasound-derived fat fraction, i.e., an ultrasound-derived assessment of liver fat fraction. Unlike the "raw" attenuation coefficient, UDFF aims to present the result in a clinically familiar format of percent fat fraction. Usually, such an indicator is calculated by the manufacturer's algorithm based on quantitative acoustic parameters, primarily related to attenuation and scattering.
Key point: the UDFF percentage is not a histological percentage of hepatocytes with fat. It is a device-dependent assessment validated against a reference method in specific studies. Therefore, in the conclusion, it is correct to indicate the name of the technology and platform, not just the number.
Why Thresholds Cannot Be Transferred Between Devices
WFUMB 2024 separately emphasizes the issue of inter-vendor comparability. Methods may have similar names but differ in probe frequency, sound propagation model, depth compensation algorithm, region of interest size, and quality criteria. Therefore, the same numerical result of attenuation on different systems does not have to have the same diagnostic meaning.
Practical consequence: thresholds should be taken from the validation of a specific technology and manufacturer. If local protocol specifies cut-off values, it is advisable to record the source of these thresholds, the type of device, and the software version. A universal table "one cut-off for all attenuation methods" does not correspond to the WFUMB position.
What to Write in the Conclusion
The conclusion should be clinically readable and reproducible. It is optimal to indicate: method, units of measurement, median or final value, quality sign of measurements according to system data, study conditions, and limitations. For UDFF, the percentage fat fraction and algorithm name should be written. For attenuation, the attenuation coefficient with units dB/cm/MHz.
The wording should separate the fact of measurement from the diagnosis. For example: "Quantitative assessment of liver fat by attenuation coefficient method: indicator increased relative to the reference range of this system; data correspond to steatosis in the absence of significant technical limitations." If quality is low, it is better to indicate: "measurement unreliable, quantitative assessment of steatosis limited."
Typical Mistakes
- Using B-mode as the only method for staging mild steatosis.
- Transferring CAP cut-off to attenuation imaging or UDFF.
- Comparing dynamics between different devices without specifying technology.
- Measuring through vessels, capsule, focal changes, or pronounced acoustic shadow.
- Interpreting UDFF as a direct histological percentage of fat.
- Ignoring concomitant fibrosis, inflammation, cholestasis, and congestive changes that may alter liver acoustic properties.
Connection with Multiparametric Ultrasound
Quantitative fat assessment is only one part of liver multiparametric ultrasound. In a patient with metabolically associated fatty liver disease, it is simultaneously necessary to assess the risk of fibrosis, inflammatory activity, and portal hypertension using appropriate methods. The fat fraction itself does not replace elastography and clinical-laboratory risk stratification.
In practice, the most useful is a combined protocol: B-mode for anatomy and exclusion of focal pathology, quantitative fat assessment for steatosis, elastography for fibrosis, Doppler ultrasound as indicated. This approach corresponds to the WFUMB idea: not to seek one universal indicator, but to use several validated parameters.
Practical Algorithm for the Ultrasound Physician
- Check which technology is available on the device: attenuation coefficient, UDFF, or other QUS indicator.
- Use the manufacturer's protocol and built-in quality control.
- Save images with the region of interest and numerical report.
- In the conclusion, indicate the method and units, not just "steatosis present/absent."
- Do not apply thresholds from another technology; if in doubt, describe the result as a quantitative indicator without strict staging.
- Compare the result with laboratory data, metabolic risk factors, and fibrosis assessment.
The main practical value of attenuation and UDFF is the objectification of steatosis and monitoring in the same technological environment. The most correct are repeated studies on the same device or on a system with proven comparability.
Frequently asked questions
Can one attenuation threshold be used for all devices?
No. WFUMB 2024 emphasizes the device and algorithm dependency of quantitative ultrasound indicators. Thresholds should be taken from the validation of a specific technology and manufacturer.
Is UDFF equal to the histological percentage of fat?
No. UDFF is expressed in percent fat fraction, but it is an ultrasound-derived algorithm calculation, not a direct count of hepatocytes with fat. It should be interpreted as a device-dependent quantitative indicator.
Why is B-mode insufficient for mild steatosis?
Liver echogenicity is subjective and depends on settings, body habitus, fibrosis, and acoustic window. In S1, i.e., 5–33% hepatocytes with fat, qualitative assessment is particularly vulnerable, so quantitative methods are preferred.