Klinik und Poliklinik für Radiologie
  1. Startseite LMU Klinikum
  2. Radiologie
  3. Forschung
  4. Functional Neuroimaging

Functional Neuroimaging

  • The Functional Neuroimaging Group focuses on the assessment of the “individual brain connectome” based on resting-state functional magnetic resonance imaging (fMRI). We work on algorithms to detect and quantify changes in the functional connectivity architecture of individual patients in comparison to large reference cohorts, with the overarching aim to develop imaging biomarkers for diagnosis, treatment monitoring and therapy guidance of neurological and neuropsychiatric disorders. Our main focus comprises neuro-oncological and neuro-degenerative diseases. In order to better understand the adult brain, our group further uses advanced MRI techniques to study brain development perinatally, including fetal and neonatal imaging of the brain, and organ systems that impact brain development such as the cardio-respiratory system.

    Our interdisciplinary group is composed of radiologists, neurobiologists, psychologist, engineers, and computer scientists.

    Functional connectivity in neurological disorders

    New concepts of brain disease suggest that network disruption does not only occur as an epi-phenomenon of underlying brain pathology, but that the formation of aberrant connections can be a major contributor to the progression of brain diseases, including neuro-oncological and neurodegenerative diseases. The current role of MRI in such diseases is to provide structural information about tumors and their close vicinity and to describe patterns of brain atrophy and vascular damage. Here, assessment of functional connectivity alterations on the whole brain level might provide important additional information that is not yet available in vivo in individual patients.

    However, despite more than two decades of functional connectivity research no clinically applicable and validated imaging tool is available to assess network dysfunctions quantitatively in the individual patient. We have designed a novel imaging marker based on resting-state fMRI which assesses global functional connectivity in individual patients (Stoecklein et al., 2020), with the aim to detect and quantify alterations of the connectivity architecture of individual patients by comparing individual connectivity profiles to a normative distribution derived from large, healthy reference cohorts.

    Neuro-oncology

    Current imaging techniques provide information about the tumors itself and its close vicinity. However, e.g., in glioma, tumor cells are not restricted to the macroscopically visible tumor but are also disseminated throughout the entire brain. Currently, there is no tool to non-invasively assess the structurally normal appearing brain parenchyma in vivo.

    We successfully tested the dysconnectivty index (DCI) in a cohort of patients with newly diagnosed glioma and found that there was a strong association between DCI and WHO grade and cognitive functioning level.

    We are currently investigating the potential of our marker in assessing prognosis, therapy response, and recurrence.

    Gliom


    Functional dysconnectivity in individual patients with high-grade and low-grade tumors.

    Overlaying connectome maps and anatomical scans visualizes widespread damage to brain functional connectivity in patients with high-grade tumors. In contrast, in patients with low-grade tumors abnormal functional connectivity is largely confined to the lesion and to the lesional hemisphere. Figure adapted from Stoecklein et al., 2020.

    Neurodegeneration

    Recent advances in functional neuroimaging highlighted disruptions in brain network in neurodegenerative disorders such as Alzheimer’s Disease, and Parkinson’s Disease. We aim to develop a functional imaging marker that predicts disease progression and assesses treatment response. Recent publications suggest the propagation of neurodegenerative pathology, e.g., tau spread, through functional brain network disruptions.   

    Artificial intelligence for data analysis of large cohort studies

    In radiomics research, artificial intelligence (AI) has emerged as a powerful tool to efficiently handle large multidimensional datasets. In particular, graph auto-encoders have proven to be potent instruments for unsupervised anomaly detection. We aim to develop AI algorithms that detect abnormalities in functional connectivity not only based on 4D functional MRI data, but also based on enriched multidimensional input data including demographics (age, sex, gender) and technical parameters (scanner type and protocol specifications). AI-based abnormality detection has been applied to structural MRI data, however moving this field into the fMRI arena and leveraging large publicly available databases as reference samples offers great potential for developing a refined diagnostic tool that is perfectly tailored to each individual patient.

    Perinatal imaging: Advanced MR imaging for the exploration of brain development in health and disease

    To understand the structure and function of the adult brain, it is crucial to understand its earliest development, including potential disruptive factors. We investigate brain development in utero and in the perinatal period, focusing on physiological development but also on early pathology of brain development, and how it might be affected by other organs such as the cardiorespiratory system.
    MRT fetal