Angioscopic Features for Aortic Dissection Characterization via Machine Learning

Satoru Takahashi's study presents a dataset of 500 patients (56 with aortic dissection, 444 controls) analyzed via non-obstructive general angioscopy (NOGA). The data includes classifications of 13 types of spontaneously ruptured aortic plaques and injuries (SRAPIs) from the aorta and bilateral common iliac arteries. A Random Forest classification model with LASSO feature selection was developed, and SHAP and Permutation Importance were used for interpretation.

Use Cases

• Developing diagnostic classifiers for aortic dissection based on angioscopic features like intramural blood and puff sign.
• Investigating pathophysiological relationships in aortic injury using feature importance analysis (SHAP, Permutation Importance).
• Validating machine learning models for medical imaging interpretation in a single-center, cross-sectional study context.
• Conducting network analysis to explore structural relationships between different SRAPI types in patient cohorts.

Strengths

• Includes data from 500 patients, providing a basis for exploratory analysis.
• Machine learning pipeline is detailed, featuring Random Forest, LASSO regression, and model interpretation techniques.
• SRAPIs are classified into 13 distinct types, allowing for granular feature analysis.

Limitations

• Row count and column-level documentation are unknown, which may limit suitability assessment.
• Data is from a single-center, cross-sectional pilot study, limiting generalizability.
• The dataset is very small at 14.3 KB, indicating limited scope.

Provenance

Source

Satoru Takahashi via figshare.

Collection Method

Single-center, cross-sectional observational pilot study using non-obstructive general angioscopy (NOGA).

Freshness

Last updated 2026-05-07 05:44:24; freshness should be verified.