Home • Publications • Software • Student offers • Funding • News • People
Ongoing and completed projects
Ongoing
KIRAL (BMFTR): KIRAL: KI-gestützte Resektionsplanung und Anpassung in der Leberchirurgie nach neoadjuvanter Therapie 11/25 - 11/28
PEACOCC - German Cancer Aid (Deutsche Krebshilfe): 04/25 - 04/28: PEACOCC - Innovative and patient-specific visualization techniques for pre-operative optimization in oncologic surgery for perihilar cholangio-carcinoma
AI4IA - Co-PI of DFG (SPP 2311) AI4IA project - Automated Segmentation and DIscrimination for Intracranial Aneurysms
DFG LIZARD 02/25 - 02/28: Liver Resection Zone Prediction Using Image-Based and Geometric Deep Learning
IDIR - Iniative for Digital Implant Research (financed by CAU Kiel University, University Hospital Schleswig-Holstein (UKSH) and Helmholtz Zentrum Hereon) 08/24 - 02/28; 2 phd positions;
Completed
DFG (SPP 2311) SCALE – Multi-scale coupling of vascular hemodynamics for AI-based standardized evaluation of neurological pathologies;
Duration 10/21-10/24; Project: https://www.spp2311.uni-stuttgart.de/en/
DFG GEPARD - GEfäßwandsimulation und -visualisierung zur Patientenindividualisierten Blutflussvorhersage für die intrakranielle Aneurysmamodellierung;
Duration 05/19 - 05/22; Project: https://gepris.dfg.de/gepris/projekt/399581926
Adaption einer virtuellen Aneurysmaexploration für den Einsatz in der klinischen Praxis;
Duration 10/21 - 03/22, Funder:EFRE - Land Sachsen-Anhalt, Europäischer Fonds für regionale Entwicklung
Deep-Learning basierte Extraktion von Fett- und Muskelmasse für die patientenspezifische onkologische Therapieplanung;
Duration 01/22-03/22, Funder: EFRE - Land Sachsen-Anhalt, Europäischer Fonds für regionale Entwicklung
Wahrnehmungsbasierte Blutflussvisualisierung für die patientenspezifische Behandlungsoptimierung multipler Aneurysmen;
Duration 04/17 - 09/18, Funder:EFRE - Land Sachsen-Anhalt, Europäischer Fonds für regionale Entwicklung
Sensitivitätsanalyse klinisch verwendeter Rekonstruktionskernel für Rotationsangiographien;
Duration 03/17 - 03/19, Funder: EFRE - Land Sachsen-Anhalt, Europäischer Fonds für regionale Entwicklung
Specific Information
Project AI4IA
Detailed Information:
In September 2024, an interdisciplinary research team from the Research Campus STIMULATE together with Otto-von-Guericke-University Magdeburg, Charité Universitätsmedizin Berlin and University Hospital Schleswig Holstein started joint research work in the AI4IA project (Automated Segmentation and DIscrimination for Intracranial Aneurysms). The main objective of the AI4IA project is to identify and reduce sources of uncertainty in the determination of morphological and hemodynamic parameters in order to enable translation into healthcare.
Ruptured brain aneurysms are the main cause of cranial hemorrhage, which can lead to severe disability or death. The treatment of aneurysms that have not yet ruptured, which occur comparatively frequently, presents a challenge, as there are numerous uncertainties regarding the individual prognosis, the optimal treatment strategy and possible treatment complications. Determining the shape of the aneurysm (morphology) and the blood flow in the aneurysm (hemodynamics) can contribute to risk assessment. However, the determination of these parameters has not yet been widely used clinically due to a lack of robustness.
The project is designed as a first step towards the development of robust rupture risk prediction systems and is planned to last 36 months as part of the second DFG priority program phase (SPP2311). It is planned as a cooperation project between the research groups in Berlin and Magdeburg specializing in neurovascular issues, with each research group being supported by a clinical partner site.
The research groups have many years of experience in modelling brain aneurysms, while the clinical partner sites have corresponding clinical expertise and provide geometries of real cases in pseudonymized form. In this context, the cooperation partners focusing on medical flows (PD Philipp Berg), neuroradiological evaluation (Prof Daniel Behme) and AI-based image analysis (me) will work closely together.