Echocardiography-Based Right Ventricle Features for Pulmonary Hypertension Prognosis

166 patient records from a study on pulmonary arterial hypertension, with features derived from manually segmented echocardiographic images of the right ventricle. The dataset was created by figshare admin karger and last updated in May 2026. It supports the development of prognostic models for cardiovascular hospitalization and mortality.

Use Cases

• Training prognostic models for pulmonary arterial hypertension based on echocardiographic myocardial features.
• Comparing the performance of radiomics and deep learning-based radiomics models for medical image analysis.
• Developing combined nomograms for 1-, 3-, and 5-year survival prediction in PAH patients.
• Assessing the added prognostic value of deep learning features over established risk scores like COMPERA 2.0.

Strengths

• Includes data from 166 patients with a mean age of 35.6 years.
• Model performance metrics are provided, including an AUC of 0.875 for the test cohort in the deep learning-based radiomics model.
• The combined nomogram demonstrated AUCs of 0.763, 0.824, and 0.935 for 1-, 3-, and 5-year predictions.

Limitations

• Column-level documentation is absent; field semantics must be inferred after download.
• Row count is unknown, which may limit suitability assessment.
• The dataset is very small at 144.1 KB, indicating limited scope.

Provenance

Source

figshare

Collection Method

Data appears to be derived from a clinical study where right ventricle myocardium was manually segmented from echocardiograms using ITK-SNAP.

Freshness

Last updated 2026-05-05 11:55:14; freshness should be verified.