Overview
This page links to the tutorials of all applications for recognition of cardiovascular diseases by analysing medical and CFD data.
Tutorials
All tutorials are available here.
Content:
Diagnosis tutorial
The full tutorial is available in the Sphinx repository here:
- Repository: AI-Based Cardiac Diagnosis from Echocardiography
- Diagnosis Tutorial
- Documentation page
What the tutorial covers
- How to organise the echocardiography datasets.
- How to run normalisation.
- (Optional) how to apply HODMD pre-processing.
- How to pre-train Masked Autoencoder (MAE).
- How to train the ViT classifier.
- How to evaluate the performance on new echocardiography datasets and inspect the results.
Related links
- Notebooks
- Datasets: Confidential CNIC datasets.
- Application hub: Cardiac Pathology
Prognosis tutorial
The full tutorial is available in the Sphinx repository here:
- Repository: AI-Based Cardiac Prognosis from Echocardiography
- Prognosis Tutorial
- Documentation page
What the tutorial covers
- How to organise the echocardiography datasets.
- How to run normalisation.
- (Optional) how to apply HODMD pre-processing.
- How to pre-train Masked Autoencoder (MAE).
- How to train the ViT regression model.
- How to evaluate the performance on new echocardiography datasets and inspect the results.
Related links
- Notebooks
- Datasets: Confidential CNIC datasets.
- Application hub: Cardiac Pathology
CFD & High-Fidelity Simulations tutorial
The full tutorial is available in the Sphinx repository here:
- Repository: Left Ventricle CFD Simulations
What the tutorial covers
- How to perform geometry pre-processing to define ventricular wall motion required for the simulations.
- How to load the geometry, meshing, and set up the simulation in STAR-CCM+ to replicate left ventricle blood flow results.
- How to perform left ventricle blood flow simulations in Fluent.
Related links
- Necessary files here.
- Video:
- Application hub: Cardiac Pathology
Contributors
- Andrés Bell-Navas
- Ander Sánchez
- Zhuoqun Zhao