Epilepsy is the most common serious chronic neurological disorder, affecting individuals of all ages. Structural magnetic resonance imaging (MRI) plays a critical role as a first-line radiological procedure in assessing abnormalities associated with epilepsy. Drug-resistant epilepsy patients face an increased risk of injury and mortality. Subtle epileptic features can sometimes be missed by the initial reporting radiologist who may have less specialized experience in epilepsy. Thus, relying solely on radiology reports may be insufficient, as anomalies might be overlooked due to poor image quality or a lack of expertise in epilepsy neuroimaging. Therefore, improving MR image assessment in epilepsy is essential.

This project is part of the Lundbeck Foundation-funded BrainDrugs Center, a research alliance based at the Neurobiology Research Unit (NRU) - Rigshospitalet, led by Prof. Gitte M. Knudsen (NRU) which aims to improve precision medicine in the treatment of depressive disorder and epilepsy.

In this specific work package, led by Assoc. Prof. Melanie Ganz and Prof. Gitte M. Knudsen with Llucia Coll Benejam as postdoctoral researcher and Alice Schiavone as research assistant, we aim to develop transparent predictive computer-based models to aid in identifying structural MRI abnormalities in epileptic patients. This includes establishing a text-mining model to automatically label radiology reports, as well as comparing MRI features identified by our deep learning image-based model proposal with the outcomes of existing task-specific methods in epilepsy research. The project is supported by Clinical Professor Lars Pinborg and his group.

Project participants:

• Professor Gitte Moss Knudsen, Neurobiology Research Unit, Copenhagen University Hospital/Rigshospitalet
• Associate Professor Melanie Ganz-Benjaminsen, Neurobiology Research Unit, Copenhagen University Hospital/Rigshospitalet
• Clinical Professor Lars Pinborg, Neurobiology Research Unit, Copenhagen University Hospital/Rigshospitalet
• PostDoc Vincent Beliveau, Neurobiology Research Unit, Copenhagen University Hospital/Rigshospitalet
• PostDoc Llucia Coll Benejam, Neurobiology Research Unit, Copenhagen University Hospital/Rigshospitalet
• Research assistant Alice Schiavone, Neurobiology Research Unit, Copenhagen University Hospital/Rigshospitalet

Website

• https://braindrugs.nru.dk/

Aging is associated with a wide range of molecular and structural changes throughout an individual’s lifespan, resulting in the deterioration of physical abilities and increased risk of disease. In the brain, aging results in cognitive decline and predisposes to neurodegenerative disorders like Alzheimer’s disease and Parkinson’s disease. Therefore, developing biomarkers capable of robustly capturing age-related biological changes is necessary to understand the pathophysiology of various disorders and potentially assess the effects of interventions that target aging.

This project is a collaboration between the Data Analysis group at the Neurobiology Research Unit (NRU), and lead by the Group of Pontus Plavén-Sigray at the Karolinska Institutet. In particular, we aim to:

• Validate existing biomarkers of brain aging on large datasets, focusing on test-retest reliability and construct validity. Both are important and need to be established before application in clinical trials
• Develop new biomarkers of aging using machine learning and neuroimaging. Mainly, we are interested in whether brain imaging using positron emission tomography yields important information regarding aging.

Trustworthy artificial intelligence (AI) methods aim to increase the fairness and transparency of AI algorithms. While this is a noble goal, current approaches are not well understood and actually risk making matters worse: Enforcing algorithmic fairness in diagnostic AI where some groups are systematically under-diagnosed can lead to increased bias in algorithms. Unfortunately, mental health is notorious in the sense that correct diagnosis depends on individuals' willingness to seek care and openly describe their symptoms -- qualities that have a strong demographic and cultural dependency. Additionally, naive interpretation of explainable AI (XAI) feedback can lead to incorrect conclusions on which features are important. This is particularly problematic in applications where there is a demographic bias, and the algorithm might be using demographic information to recalibrate the algorithm rather than to actually discover disease. Such incorrect feedback risks both leading to incorrect scientific discovery about diseases, and decreasing the users' trust in the algorithm. This project aims to bring these problems out in the open: By using major depressive disorder (MDD) risk scoring algorithms as a use case, we will illustrate how different levels of label bias affect trustworthy AI algorithms; we will estimate the magnitude of the bias and illustrate how it can be accounted for.

This project is anchored in the group of Professor Aasa Feragen at DTU Compute, Denmark’s Technical University and carried out in collaboration with Melanie Ganz from the Neurobiological Research Unit at Rigshospitalet, Denmark.

Project participants:

• Professor Aasa Feragen (PI), DTU Compute, Denmark's Technical University (DTU)
• Associate Professor Melanie Ganz-Benjaminsen, Neurobiology Research Unit, Copenhagen University Hospital/Rigshospitalet
• Postdoc Olalekan Joseph Akintande, DTU Compute, Denmark's Technical University (DTU)
• PostDoc Helen Coupland, Neurobiology Research Unit, Copenhagen University Hospital/Rigshospitalet

​NIBS-CP (NeuroImaging of Babies during Natural Sleep to Investigate Typical Development and Cerebral Palsy) is a brain research project at Hvidovre Hospital, supported by the Elsass Foundation. In NIBS-CP, we investigate the development of the brain and motor function in infants. The project has been approved by the Research Ethics Committees for the Capital Region of Denmark (H-24008782), and has also been reported to the Danish Data Protection Agency via the Knowledge Center for Data Reviews in the Capital Region of Denmark.

​The project investigates two groups of infants - typically developing infants and infants at risk of cerebral palsy. The infants at risk of cerebral palsy are recruited from Professor Christina Høi-Hansen’s national CP-EDIT (Cerebral Palsy: Early Diagnosis and Intervention Trial) study at Rigshospitalet and Herlev Hospital. The typically developing infants (control participants) are recruited via the Danish center for Magentic Resonance Imaging (DRCMR) at Hvidovre Hospital, social media, mothers’ groups, etc. Control participants must be born at term (> week 37) without birth complications and be free of neurological disease to participate in the project. In addition, you must reside in the Capital Region of Denmark or Region Zealand. The children will be followed over time with three visits between the ages of 3-24 months. The visits consist of motor and neurological examinations and MRI scans at Hvidovre Hospital.

The project will enable a mapping of typical and atypical brain development, as well as how brain development is connected to and predicts motor development in children. In addition, a Danish normative material of brain development in the 0-2 year age range will be established.

Project participants:

• Senior Researcher Kathrine Skak Madsen (PI), Danish Center for Magnetic Resonance (DRCMR), Hvidovre Hospital
• Associate Professor Melanie Ganz-Benjaminsen (Co-PI), Neurobiology Research Unit, Copenhagen University Hospital/Rigshospitalet
• PostDoc Maddy Wyburd, Department of Computer Science, University of Copenhagen

Website

• https://www.hvidovrehospital.dk/afdelinger-og-klinikker/mr-forskningssektion/nibs-cp/Sider/default.aspx

Molecular neuroimaging, e.g., positron emission tomography (PET) and single photon emission tomography (SPECT), is an in vivo brain scanning modality that uniquely can be used to investigate neuropharmacological effects and brain diseases in vivo. In the past two decades, we have seen an exponential growth in the scale at which neuroimaging research in psychiatry and neurology takes place in terms of sample size, technology advancements and number of disorders investigated. This has led to a parallel increase in the demand on computing power and need for a standardized platform that can be used for storing, processing and sharing of neuroimaging data. In neuroscience, PET studies have set the stage for measuring brain function in disorders such as depression or Alzheimers Disease, with the goal of developing effective treatment strategies. However, in the past decade there has been a growing concern about the validity and reproducibility of outcomes from such studies, and this has largely been attributed to low statistical power, software errors and flexible data analysis strategies. This is important, because only research that can be trusted will have a real impact on the society. Furthermore, a PET study is extremely costly and it also involves that people are exposed to radiation. That is, for both financial and ethical reasons, data sharing is urgently needed. We here propose establishment of a PET data infrastructure called OpenNeuroPET. The proposed infrastructure will unite the scientific community to enable meta- and mega-analyses of brain imaging data for by creating an expertly labelled, shared-access data repository and processing platform. The main ethos of the OpenNeuroPET will be: open, inclusive, participatory, and democratic. Identifying neuroimaging biomarkers is fundamental to improving our understanding of brain disorders and obtaining objective measures to aid their early diagnosis, prevention and treatment.

The OpenNeuroPET project (DK) is funded by the Novo Nordisk Foundation as an infrastructure grant, led by Prof. Gitte Moos Knudsen and anchored at the Neurobiology Research Unit at the Copenhagen University Hospital/Rigshospitalet. We collaborate closely with the Molecular Imaging Branch at the National Institute of Mental Health (NIMH) on the project.

Project participants:

• Professor Gitte Moss Knudsen (PI), Neurobiology Research Unit, Copenhagen University Hospital/Rigshospitalet
• Associate Professor Melanie Ganz-Benjaminsen, Neurobiology Research Unit, Copenhagen University Hospital/Rigshospitalet
• Senior Researcher Cyril Pernet, Neurobiology Research Unit, Copenhagen University Hospital/Rigshospitalet
• Assistant Professor Martin Nørgaard, Department of Computer Science, University of Copenhagen

Website

• https://openneuropet.github.io/

The mental health system in Europe faces challenges, with extensive waiting times for diagnostic psychiatric appointments, and in Denmark, a national average waiting time of 92 weeks to see a specialist.

We aim to develop AI algorithms for mental health screening, specifically targeting Major Depressive Disorder (MDD), which affects up to 6.9% of the European population. The algorithms will help to reduce the waiting time for diagnosis and improve the overall efficiency of the mental health system.

The research project is based on a collaboration between three senior scientists from different disciplines—neuroscience, AI as well as medicine—as well as two collaborators from the field of ethics.

The TRUSTMIND project is anchored at the Image section of the Department of Computer Science and led by Assc. Prof. Melanie Ganz-Benjaminsen.

Project participants:

• Associate Professor Melanie Ganz-Benjaminsen (PI), Department of Computer Science, University of Copenhagen
• Professor Merete Osler (Co-PI),Center for Clinical Research and Prevention at Bispebjerg and Frederiksberg Hospitals (CCRP-RH) and Institute of Public Health at the University of Copenhagen
• Professor Aasa Feragen (Co-PI), DTU Compute at Denmark's Technical University (DTU)
• Associate Professor Sune Hannibal Holm, Department of Food and Resource Economics, University of Copenhagen
• Associate Professor Katharina O'Cathair, LAW, University of Copenhagen

Website

• https://lundbeckfonden.com/news/120-million-dkk-interdisciplinary-brain-research/melanie-ganz-benjaminsen