Research
I am involved in several projects at NRU, DIKU and abroad, spanning medical image analysis, machine learning, and open neuroimaging data. Most current projects run under the Lab for Neuroimaging and Neuroinformatics (LNN) @ UCPH, which I co-lead with Martin Nørgaard. Click through to each project below for the full story.
Ongoing projects
TRUSTMIND Lundbeck Foundation, PI, 2025–2030
AI in Mental Health – Ensuring Evaluation, Trustworthiness, and Ethical Considerations. A 5-year, 20,000,000 DKK project developing AI algorithms for screening Major Depressive Disorder, with Aasa Feragen, Merete Osler, Sune Holm and Katharina Ó Cathaoir.
Biomarkers of Biological Brain Aging 2023–2027
Developing and validating biomarkers that robustly capture age-related biological changes in the brain, combining structural MRI and PET-derived features. Ruben P. Dörfel's PhD project, co-supervised with colleagues at the Karolinska Institutet.
NIBS-CP Elsass Fonden, Co-PI, 2022–2028
NeuroImaging of Babies during natural Sleep to assess typical development and Cerebral Palsy. Establishing procedures for clinical infant/toddler MRI during natural sleep, without sedation or general anesthesia. PI Kathrine Skak Madsen, Co-PI Melanie Ganz.
OpenNeuroPET Novo Nordisk Foundation, Co-PI, 2020–2027
The OpenNeuroPET Archive – a research infrastructure to share, curate and analyze brain PET data, building directly on the PET BIDS extension.
BrainDrugs — WP5 Lundbeck Foundation, 2019–2027
Mining structural MRI data in Epilepsy. Part of BrainDrugs, a thematic alliance in precision medicine for epilepsy and depression (PI Gitte Moos Knudsen); I lead WP5, mining neuroimaging data.
Previous projects
Trustworthy Healthcare AI for Mental Health Risk Prediction Lundbeck Experiment, 2024–2026
Measuring the reliability of trustworthy healthcare AI for mental health risk prediction – a pilot project that followed on from Bias and Fairness in Medicine and helped shape TRUSTMIND.
Bias and Fairness in Medicine DFF, Co-PI, 2020–2023
Examining the nature of bias in the diagnosis and treatment of disease, using major depressive disorder as an example, with Aasa Feragen and Sune Holm. Helped kick-start the FAIMI initiative and led directly to TRUSTMIND.
Imaging Children Without Anesthesia (MoCo project) Elsass Fonden, PI, 2019–2024
"How can you avoid putting children with cerebral palsy in general anesthesia while they are in a brain scanner?" 94% of children were successfully scanned without anesthesia – a result that changed clinical practice at Rigshospitalet.
Brain Imaging Data Structure (BIDS) Steering Group, 2019–2022
The BIDS standard for organizing and sharing neuroimaging data. I helped develop the PET BIDS extension and served on the BIDS Steering Group, representing the PET community.
Permutation Tests for Binary Classification Oberwolfach, 2014
A theoretical analysis of permutation tests for binary classification with small sample sizes, common in neuroimaging and genomics.
Imaging the Serotonin System 2013–2020
A high-resolution, multi-modal PET/MRI atlas of the human brain's serotonin system, and machine learning methods to identify disease-relevant regions within it.