Non-Traditional Lipid Indices for CHD Classification in MASLD and T2DM Patients

A multicenter retrospective study of 1,823 patients with comorbid metabolic dysfunction-associated steatotic liver disease (MASLD) and type 2 diabetes mellitus (T2DM). The dataset was used to identify eight non-traditional lipid indicators and develop an interpretable machine learning framework for coronary heart disease (CHD) identification, with an independent validation cohort of 158 patients. The data was uploaded by Luo Lv in April 2026.

Use Cases

• Training machine learning models for CHD risk classification based on non-traditional lipid indices.
• Analyzing associations between lipid ratios (e.g., Castelli risk index-II) and coronary lesion severity.
• Developing interpretable clinical decision tools using SHAP and LIME explanations.
• Validating cardiovascular risk models in a specific patient population with MASLD and T2DM.

Strengths

• Includes data from 1,823 patients across multiple centers, providing a substantial clinical cohort.
• Uses an independent external validation cohort of 158 patients for model testing.
• The study evaluates eight specific non-traditional lipid indices, such as remnant cholesterol and atherogenic index of plasma.
• Model performance is reported with specific metrics (AUCs of 0.750 and 0.756, Brier scores).

Limitations

• Row count for the underlying data tables is unknown, which may limit suitability assessment.
• Column-level documentation is absent; field semantics must be inferred after download.
• Data may reflect temporal and geographic bias inherent to the multicenter retrospective study design.

Provenance

Source

figshare, uploaded by Luo Lv.

Collection Method

Multicenter retrospective study using propensity score matching and machine learning model development.

Time Range

null

Freshness

Last updated 2026-04-22 05:35:05; freshness should be verified.

Geography

null