Sie befinden sich hier
Inhalt
CARE-IA at a glance
CARE-IA stands for "Care, Awareness, Risk detection & prediction, prevention strategies in Intracranial Aneurysms." The project pursues a simple goal: to provide people with a brain aneurysm with a more accurate understanding of their personal risk – ensuring they are neither unnecessarily treated nor unnecessarily left alone.
To achieve this, we combine clinical data, high-resolution imaging, blood biomarkers, and artificial intelligence methods into a single platform. CARE-IA is funded by the HW & J. Hector Foundation and is based at the Center for Brain Aneurysms at the University Medical Center Mannheim.
The five areas of work:
I. Digital infrastructure
A secure, GDPR-compliant data platform that combines imaging, clinical data and biomarkers (led by Biomedical Informatics, UKMA).
II. GUIDE-IA – Education, Information, Decision Aids
The first nationwide awareness campaign on brain aneurysms, an independent knowledge and second opinion platform, and the first German self-help structure for those affected. This website is part of it.
III.Early detection and risk prediction
AI-supported detection of aneurysms in imaging, validation of blood biomarkers and a multimodal model for predicting development, growth and rupture.
IV. Prevention and self-efficacy
PROTECT-U and the telemedicine support of conservatively managed patients – with the explicit goal of also improving anxiety and quality of life.
V. RESCUE-SAH
A coordinated regional acute care network for subarachnoid hemorrhage with digital triage and optimized transfer pathways.
Early detection and AI-supported risk prediction
At present, it is only possible to make a rough assessment of whether an aneurysm is dangerous. We are developing methods that use imaging data, blood test results and medical history to calculate a significantly more accurate, personalised risk profile – so that the decision to ‘treat or monitor’ can be made on a more sound basis than is currently the case.
Read more
- Voxel-precise annotation of over 500 aneurysm patients as well as the Mannheim PROTECT-U cohort; training of deep learning models (CNN, Vision Transformer, multimodal foundation models) with a particular focus on robustness against incomplete imaging reports.
- External validation using, amongst other sources, data from the National Cohort (NAKO).
- Multimodal fusion models for the identification of clinically relevant aneurysm biotypes (in collaboration with ZI Mannheim and the Hector Institute for AI in Psychiatry).
Biomarkers from the blood
At present, imaging is required for every follow-up examination. We are investigating whether substances can be detected in the blood that indicate whether the wall of an aneurysm is becoming unstable – a simple blood test that could one day complement imaging or make follow-up examinations more targeted.
Read more
- Background: the detection of circulating collagen degradation products as markers for the presence and instability of intracranial aneurysms (Hackenberg et al., European Stroke Journal 2024; Stroke 2020).
- Extension to include circulating endothelial cells and endothelial progenitor cells as markers of vascular damage and repair.
- Longitudinal validation in the CONTINUITY cohort.
Prevention: PROTECT-U
PROTECT-U is the world’s first large-scale study to investigate whether aneurysm haemorrhage can be prevented with medication – through consistent blood pressure reduction and low-dose acetylsalicylic acid. Until now, there have been only two options for unruptured aneurysms: intervention or watchful waiting. PROTECT-U is investigating a third option.
Read more
- Randomised, international Phase III trial; trial coordination in Mannheim/Utrecht; currently underway in Germany, the Netherlands, Finland and Canada.
- Currently over 470 patients enrolled, of whom over 170 have completed the 36-month follow-up. [V: Update figures at the time of publication]
- PROTECT-U 2.0: Establishment of the first prospective ten-year cohort of unruptured aneurysms managed conservatively – robust long-term data where, to date, only retrospective and selected cohorts have been available.
- A dedicated, constantly updated section entitled ‘Can I take part?’ with inclusion and exclusion criteria explained in everyday language and a direct contact channel to the study team.
