Integrated Model for Hepatic Fat Fraction Assessment in a Chinese MASLD Cohort

A single-center Chinese health-examination cohort of 172 participants used to evaluate models for non-invasive hepatic fat fraction categorization. The dataset includes ultrasound attenuation values and serum levels of ALT, AST, and ferritin, referenced against quantitative computed tomography categories. It was authored by Zejun Ma and last updated in May 2026.

Use Cases

• Comparing model performance (AUC) for steatosis detection based on ultrasound attenuation and metabolic biomarkers.
• Investigating subgroup performance differences based on sex, age, and BMI as mentioned in the results.
• Training or benchmarking multiclass classification models for hepatic fat fraction categories, noting the limited sample size for Category 2.
• Analyzing the correlation between ultrasound attenuation values and increasing hepatic fat fraction categories.

Strengths

• Includes performance metrics (AUC with 95% CI) for three distinct models from an internal validation study.
• Dataset is small (12.8 KB) and focused, derived from a study with a defined cohort of 172 participants.
• Uses a pragmatic imaging comparator (QCT-derived hepatic fat fraction) as referenced labels.

Limitations

• Row count and column-level documentation are unknown, limiting suitability assessment.
• The study concludes discrimination for the intermediate fat fraction category was weak, with only 31 participants in that group.
• Data is from a single-center cohort, and the authors state external validation in more diverse populations is required.

Provenance

Source

figshare, authored by Zejun Ma.

Collection Method

Data collected from participant enrollment where USAT values and serum biomarkers (ALT, AST, ferritin) were measured, with QCT used for reference categorization.

Freshness

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

Geography

Single-center Chinese health-examination cohort.