Ultrasound Attenuation and Biomarkers for Hepatic Fat Fraction Assessment in MASLD

A study of 172 participants from a single-center Chinese health-examination cohort, conducted by Zejun Ma and last updated in May 2026. It evaluates models combining ultrasound attenuation and metabolic biomarkers for non-invasive categorization of hepatic fat fraction, using quantitative computed tomography as a reference.

Use Cases

• Train a binary classifier for detecting imaging-defined steatosis based on ultrasound attenuation values.
• Compare the performance of biomarker-only models against integrated models for liver fat assessment.
• Analyze subgroup performance differences based on demographic factors like sex, age, and BMI.
• Explore multiclass categorization challenges for intermediate hepatic fat fraction groups.

Strengths

• Includes data from 172 participants, enabling model training and validation.
• Model performance was assessed using fold-separated internal validation and Harrell’s optimism-corrected bootstrap analysis.
• Specific performance metrics are reported, such as an AUC of 0.847 for the USAT-only model.

Limitations

• Row count and column-level documentation are unknown, limiting suitability assessment.
• The dataset is small (77.3 KB), suggesting limited scope.
• The study is single-center and requires external validation for generalizability.

Provenance

Source

Zejun Ma via figshare.

Collection Method

Data collected from participants categorized by quantitative computed tomography, with ultrasound attenuation values and serum biomarker levels measured.

Freshness

Last updated 2026-05-29 04:53:49.

Geography

Single-center Chinese health-examination cohort.