Group for Applied Imaging Applications of Artificial Intelligence in Medicine (GAIA)

The Group for Applied Imaging Applications of Artificial Intelligence (GAIA) is designed as an interdisciplinary team that combines new means and methods of image analysis with the clinical expertise of various medical disciplines. While the technical equipment and AI know-how are provided by the data science group affiliated with radiology, the applications are based on the expertise of various groups throughout the hospital. By joining forces, the full power of AI-based image analysis can be unleashed and used to improve patient care.

Prof. Dr. Dr.--Große Hokamp-Nils
Prof. Dr. Dr. Nils Große Hokamp, MBA, EDiR

Head Computed Tomography Research

Master of Business Administration
Q2-Zertifikat für mpMRT der Prostata (DRG)
Q2-Zertifikat für Abdominelle Radiologie - Leber und Pankreas (DRG)
Q2-Zertifikat für Lungenkrebsfrüherkennung (DRG)
Abdominelle und onkologische Bildgebung
Sektionsleitung Uroradiologie
mpMRT der Prostata
Abdominelle Bildgebung (Leber und Pankreas)
Lungenkrebsfrüherkennung
Q2-zertifzierter Radiologe für
Q1-zertifizerter Radiologe für Herz-CT, Herz-MRT und Muskuloskelettale Radiologie
Master of Oncologic Imaging (Universität Pisa in Kooperation mit der European Society for Oncologic Imaging, ESOI)

Dr.--Lourenco Caldeira-Liliana
Dr. Liliana Lourenco Caldeira

Data Science Group - Team Leader

Google Scholar Profile
Research Gate Profile
LinkedIn Profile

telephone icon +49 221 478-85731

Structure and methods

GAIA builds on existing collaborations with clinical partners in nephrology, nuclear medicine, neurology, urology, dermatology and neurosurgery, but is intended as an open structure and environment to accommodate new collaborators and/or projects. The data science side provides AI methods for image analysis, including Convolutional Neural Nets, U-Nets, Res-Nets, but also traditional machine learning concepts such as Radiomics and Support Vector Machines. In addition, new and up-to-date algorithms and frameworks are constantly being developed and made available for translational research (including generative adversarial networks, ensemble and unsupervised learning techniques). All algorithms are embedded in research workflows, including GDPR-compliant data curation and storage, to enable direct applicability to new research projects and/or ideas.

Promotion

Cologne Fortune research pool (funding type 3)