A project funded
by Novo Nordisk Foundation

Nordic
CCT-AI

Disrupting Cardiac CT-Imaging
through large-scale data, collaboration and open-source AI

about the project

The Nordic Collaboration on AI and Cardiac Imaging is a large-scale, Novo Nordisk Foundation–funded initiative designed to transform cardiac diagnostics through open‑source artificial intelligence. The project, started in January 2026, brings together leading clinical and technical groups across Denmark, Norway, and Sweden to develop advanced AI models for the analysis of cardiac CT scans. These models aim to support earlier, more accurate diagnosis of cardiovascular disease and promote consistent clinical decision‑making across the Nordic region.

At the heart of the project is the creation of one of the world’s largest datasets of cardiac CT scans, encompassing thousands of individuals — from healthy participants to patients with severe symptomatic atherosclerosis. This comprehensive cohort enables researchers to study coronary artery disease across its full spectrum and evaluate disease progression over time. All data used in the project are non‑identifiable and processed exclusively within secure research environments.

The project will establish a shared Nordic research platform where engineers and clinicians jointly develop, test, and validate AI models. By releasing these models as open source, the initiative directly addresses current barriers posed by proprietary, commercial systems, enabling transparent comparison, faster innovation, and broader clinical adoption. Ultimately, the project aims to deliver AI tools that can be integrated into daily clinical workflows to support patient care across the Nordic healthcare systems.

OUR CLINICAL COHORT

The clinical cohort includes thousands ( > 100'000) of cardiac CT scans collected across Denmark, Norway, and Sweden. It contains imaging and health data representing:

  • Healthy individuals

  • Patients with early signs of atherosclerosis

  • Patients with advanced, symptomatic coronary artery disease

This breadth makes the dataset uniquely suited to training robust AI models and evaluating their performance across different patient populations and clinical scenarios.The dataset is one of the largest ever assembled for cardiac AI research.

DATA INFRASTRUCTURE

Nordic‑CCT‑AI relies on a secure, scalable computing setup to manage its large cardiac CT cohort. All scans are stored in Aalborg University’s research PACS, with initial AI development performed on the GDPR‑compliant TAAURUS HPC platform. For full‑scale model training and CFD simulations, the project uses the Gefion AI supercomputer, requiring an estimated 315,470 GPU hours.Data flows securely through the Danish Health Data Network, enabling integration with Statistics Denmark and the Danish Health Data Authority. Non‑cardiac CT data are harmonized with REDCap, and international datasets are processed on Statistics Denmark’s research platform. All data handling follows strict GDPR and institutional data‑protection procedures.

our team

This project is driven by a highly multidisciplinary team that unites expertise from cardiology, radiology, biomedical engineering, and computational sciences.The collaboration is led by Dr. Simon Winther, cardiologist and clinical associate professor at Gødstrup Regional Hospital and Aarhus University, and Prof. Samuel E. Schmidt, expert in cardiotechnology at Aalborg University. The coordinating lead is Dr. Monika Colombo, PhD, from Aarhus University.The team includes additional clinicians, imaging specialists, and engineers from leading institutions across the Nordic countries.
Together, the team combines deep clinical insight with cutting‑edge AI expertise to build solutions designed not only to advance research, but also to directly improve patient outcomes in routine cardiac care.

NameCountryOrcID
Simon Winther, MDGødstrup Regionalhospital, DK0000-0001-8872-3681
Samuel E. SchimdtAalborg University, DK0000-0002-0917-634X
Monika ColomboAarhus University, DK0000-0002-6658-4438
Araz Rawshani, MDSahlgrenska Universitetssjukhuset, SW0000-0003-2066-3533
Vibha GuptaGothenburg University, SW0000-0002-3785-8380
Fredrik E. FossanSt. Olavs Hospital, Trondheim, NO0000-0003-0006-5165
Henrik P. LindAalborg University, DK0000-0001-6581-8899
Jesper O. KragballeAalborg University, DK0009-0002-8206-6611
Kavin VishnuAarhus University, DK0009-0004-9510-6162
Jonathan N. Dahl, MDHerning Regional Hospital, DK0000-0001-5574-3926
Laust D. Rasmussen, MDAalborg University Hospital, DK0000-0002-2790-2608
Josephine Therkildsen, MDAarhus University Hospital, DK0000-0003-4027-4631
Ida M. Bøhme, MDAarhus University, DK0009-0008-9767-930X
FORMER MEMBERS  
Gitte S. Brix, MDHerning Regional Hospital, DK0000-0002-7426-2818
iNTERNATIONAL ADVISORY BOARD 
Cardiologist: 
Prof. Juhani KnuutiTurku University Hospital, FI
Prof. Peter PeturssonSahlgrenska Hospital, SW
Prof. Rune WisethSt. Olavs University Hospital, NO
Data scientist: 
Prof. Magnús Ö. ÚlfarssonUniversity of Iceland, IS
Ass. Prof. Riku KlenTurku University, FI

Publications

2026

News

August, 2026
ESC 2026 | Munich | August 2026
Great participation by our team, showcasing some of the findings and methods developed in the first 6 months of the project. Some example titles:
Opportunistic volume assessment of LV volume from coronary CT angiography: agreement with cardiac MRI presented by Henrik P. Lind.Fully automatic CACS using CT and AI: external validation on a danish population presented by Jesper O. Kragballe.Photo: Some team members at ESC 2026

July, 2026
World Congress on Biomechanics | Vancouver | July 2026
Our Monika has presented the work entitled AUTOMATIC CT IMAGE QUALITY ANALYSIS FOR SELECTION OF OPTIMAL PHASE IN LARGE MACHINE LEARNING DATASETS.
A first milestone in our automated pipeline for cardiac prediction on 100'000 patients, greatly led by our Jesper.Photo: Monika presenting at WCB

Jun, 2026
Symposium | Aalborg-Gothenburg | Summer 2026
Our interdisciplinary working group recently met at a dedicated symposium where researchers exchanged ideas, discussed ongoing work, and presented preliminary findings.
The event provided a valuable space for collaboration, constructive feedback, and the development of new research directions.
Photo: core working team

Feb, 2026Millions from the Novo Nordisk Foundation enable Nordic research in AI for cardiac diagnostics.
The Novo Nordisk Foundation has chosen to support a project aimed at developing AI models for the analysis of cardiac CT scans.
Read our full press release here


We are open to collaborations.
Reach out to us or follow us on GitHub!