Deep-learning based Cardiac Disease classification, Interpretation and Outcome prediction

Deep-learning based Cardiac Disease classification, Interpretation and Outcome prediction - DeepCARDInO

Status

Recruiting

Phases

Unknown

Study type

Observational

Source

DRKS

Registry ID

DRKS00040908

Enrollment

20000

Registered

2026-07-15

Start date

2026-02-02

Completion date

Unknown

Last updated

2026-08-03

For informational purposes only — not medical advice. Sourced from public registries and may not reflect the latest updates. Terms

Conditions

I20-I25 I26-I28 I30-I52 I05-I09

Interventions

Group 1: A retrospective analysis of adult, non-pregnant patients who underwent cardiac magnetic resonance imaging (cardiac MRI) between 2006 and 2024 at the Department of Internal Medicine III – Card

Eligibility

Sex/Gender

All

Age

18 Years to No maximum

Inclusion criteria

Inclusion criteria: - Performance of cardiac magnetic resonance imaging (cardiac MRI) and/or echocardiography at the Department of Internal Medicine III – Cardiology, Angiology and Pneumology at Heidelberg University Hospital between 2006 and 2024, regardless of the underlying cardiovascular diagnosis or clinical issue

Exclusion criteria

Exclusion criteria: - An existing pregnancy at the time of the examination - Express objection by the patient to the scientific (secondary) use of their data, provided such an objection is documented - Image or findings data that cannot be analysed or are technically inadequate (e.g. significant motion artefacts, incomplete image sequences, insufficient temporal resolution for determining cardiac phases)

Design outcomes

Primary

Measure	Time frame
Accuracy of AI-assisted, automated cardiac phase determination in cardiac MRI images compared with manual reference determination, and correlation of the resulting cardiac phase-specific automated strain parameters (radial, circumferential, longitudinal) of the myocardium with conventional, manually recorded measurement parameters.	—

Secondary

Measure	Time frame
- Correlation and transferability of automatically determined strain parameters between cardiac MRI and echocardiography data - Diagnostic performance (sensitivity, specificity, AUC) of machine learning models based on strain and motion analyses for the phenotyping of cardiovascular diseases and patient stratification - Identification and characterisation of potential subtypes within the HFpEF population using unsupervised cluster analysis in the latent space of neural networks, incorporating image-based, clinical and electrocardiographic parameters	—

Measure

Time frame

- Correlation and transferability of automatically determined strain parameters between cardiac MRI and echocardiography data - Diagnostic performance (sensitivity, specificity, AUC) of machine learning models based on strain and motion analyses for the phenotyping of cardiovascular diseases and patient stratification - Identification and characterisation of potential subtypes within the HFpEF population using unsupervised cluster analysis in the latent space of neural networks, incorporating image-based, clinical and electrocardiographic parameters

—

Countries

Germany

Contacts

Public ContactSarah Kaye Mueller

Universitätsklinikum Heidelberg

SarahKaye.Mueller@med.uni-heidelberg.de+49 6221 56672

Outcome results

None listed

Source: DRKS (via WHO ICTRP) · Data processed: Aug 10, 2026