Study 01 · Registered
EXCALIBUR-HFpEF
A two-centre study comparing standard HFpEF care with standard care plus a 12-week individual training programme and app-based personalised activity coaching.
Public health research · Heidelberg, Mainz & Munich
Researching heart failure across disciplines: multimodal AI, exercise training and robotics for innovative therapy
Multi-dimensionAI links different scales of health information to investigate patterns in HFpEF and differences between patients.
Funded by
Funding amount
Approximately
€5 million
€5 million
Project period
July 2024
to June 2030
to June 2030
Locations
Heidelberg, Mainz
& Munich
& Munich
01 · About the project
Linking scales of information to improve care for patients with heart failure
Heart failure with preserved ejection fraction (HFpEF) affects approximately half of all patients with heart failure and is associated with substantial morbidity and mortality. As a complex multisystem disease characterised by heterogeneous clinical phenotypes, HFpEF presents considerable challenges for diagnosis and treatment. Advances in artificial intelligence (AI) offer new opportunities to improve disease characterisation and develop personalised treatment strategies.
The Multi-dimensionAI project brings together interdisciplinary expertise across three interconnected subprojects. Its overarching innovation lies in integrating AI-driven disease phenotyping, prospective clinical intervention data, and stakeholder-centred evaluation within a unified framework.
By linking existing and newly collected molecular, clinical, imaging and wearable-derived data across different scales, the project aims to uncover previously unrecognised disease patterns and identify patient subgroups that may benefit from tailored interventions.
Patients and healthcare professionals are involved in the development process at an early stage. Their active participation is intended to ensure that the AI technologies developed are trustworthy, prove themselves in everyday clinical practice and can be used meaningfully in future care.
02 · The medical challenge
HFpEF can look very different from one person to another
HFpEF (heart failure with preserved ejection fraction) is a common form of chronic heart failure. Relaxation and filling of the left ventricle are often impaired, while the proportion of blood it ejects with each beat remains preserved.
Causes, accompanying conditions, disease progression and response to therapy can differ substantially between patients.
This heterogeneity makes prognosis and treatment decisions difficult. Multi-dimensionAI investigates how different sources of information might help researchers better understand these differences in the future.
03 · Multimodal AI
Combining different data gives research a comprehensive view
“Multimodal” means bringing together information from different sources for AI-supported research analysis. The aim is to search for recurring patterns and possible correlations, not to make an automated diagnosis.
How the research relates
- Bring approved sources together
Information from different scales and study contexts can be considered in combination.
- Investigate possible patterns
AI-supported analyses can help researchers explore recurring relationships and subgroups.
- Evaluate through research
Any findings require careful scientific validation before possible future clinical use.
04 · Studies
Two distinct research components
The exercise-and-coaching study and the robotics-assisted study ask related questions, but they are separate.
Study 02 · Details pending
Robotics-assisted exosuit study
An individually adapted exosuit or exoskeleton can support movement during supervised training. The project investigates which patients may benefit from exercise therapy based on combinations of characteristics.
Recruitment status not yet confirmed · study details to follow
Funded by