CTA Scans and Clinical Data for Predicting Abdominal Aortic Aneurysm Rupture

A multimodal dataset combines sequential CTA slices and six clinical biomarkers from 263 symptomatic abdominal aortic aneurysm patients. It was created by Jiaxin Cheng for developing and validating a deep learning model for rupture risk assessment. The dataset was last updated in April 2026.

Use Cases

• Train a multimodal model to predict impending AAA rupture by fusing sequential CTA slices with six clinical biomarkers.
• Benchmark rupture risk prediction models against clinical-rule and CTA-sign baselines using the provided AUC metrics.
• Validate model interpretability using Gradient-weighted Class Activation Mapping (Grad-CAM) on CTA sequences.
• Assess model robustness against domain shifts using the independent temporal test set of 33 patients.

Strengths

• Includes data from 263 symptomatic AAA patients with a matched development cohort of 150 patients.
• Model achieved an AUC of 0.898 in the matched test set and 0.880 in an independent temporal validation cohort of 33 patients.
• Grad-CAM visualizations were anatomically plausible in 78.8% of cases, supporting interpretability.

Limitations

• Small dataset scale at 6.5 MB, with a total of 263 patients and a validation set of only 33 cases.
• Retrospective cohort study design may introduce selection bias.
• Clinical biomarkers are limited to six key variables; other potentially relevant factors may be excluded.

Provenance

Source

Jiaxin Cheng via figshare.

Collection Method

Retrospective cohort study of symptomatic AAA patients.

Time Range

null

Freshness

Updated in April 2026.

Geography

null