Final Report Summary - HMRI (Non-Invasive In-Vivo Histology in Health and Disease Using Magnetic Resonance Imaging (MRI))
The hMRI project developed, optimized and applied novel non-invasive magnetic resonance imaging (MRI) methods to reliably characterize the detailed anatomical microstructure of the human cortex. The first phase of this project was focused on the development of MRI data acquisition, data analysis methods and biophysical models. The second phase was dedicated to the (clinical) neuroscience experiments and further refinements of the methods for these applications.
We achieved in-vivo imaging of the human cortex using quantitative multi-parameter maps (MPMs) with an unprecedented 400 µm isotropic resolution. This was only possible by developing and seamlessly integrating optical prospective motion correction, which addresses the unavoidable head motion artifacts at such high spatial resolution. The MPMs were combined with cutting edge diffusion imaging to estimate features of the cortical microstructure such as markers of fiber geometry and myelination. Different methods for functional MRI with 800 µm isotropic resolution were developed and compared for highest effective resolution taking into account physiological blurring effects (draining vein effects). They allowed for precise layer-specific functional activation studies of columnar cortical structures. The combination of MPM-based myelin markers with ultra-high resolution fMRI demonstrated different myelination levels in the different types of stripes in the secondary visual cortex (V2).
Extraordinary advances in diffusion imaging were achieved by leveraging the potential of the recently commissioned 3T Connectom MRI at the MPI for Human Cognitive and Brain Sciences, which offers the highest maximum gradient amplitude (300 mT/m) for human imaging and is one of just four systems worldwide (https://www.cbs.mpg.de/institute/technical-facilities). Data quality was further improved by a novel flexible radio-frequency (RF) surface coil and optimized data processing pipelines for the high-end hardware – for the first time demonstrating the structure-function relationship between U-fibers and retinotopy in the visual cortices.
We have developed novel data processing methods and biophysical models that integrate information from different MRI contrasts, ranging from relaxometry to diffusion and functional MRI. These models include the quantification of iron concentration, intracortical fibers, superficial white matter and description of tangential structures in the cortex such as ocular dominance columns and stripes in the primary and secondary visual cortex. They are the key link between non-invasive MRI measurements and the underlying microstructure.
Post-mortem MRI (with up to 25 µm resolution) and histology were further developed and applied to tissue specimen of controls and Alzheimer’s disease (AD) patients in order to inform model building and validate the models. We have pushed the limits of current post-mortem histology methods, which are limited by the vagaries of staining efficiency and their 2D view on the brain anatomy. Tissue clearing combined with advanced microscopy and image processing methods allowed for comprehensive imaging of cortical structures. By using proton-induced X-ray emission (PIXE), matrix-assisted laser desorption/ionization (MALDI), mass spectroscopy imaging we were able to estimate high-resolution maps (from macroscopic- down to cellular-resolution) of iron, myelination and variations in myelin properties in the brain.
The novel methods were used to study neurodegenerative changes and focal atrophy of the cortex in AD and posterior cortical atrophy (PCA), a variant of AD, demonstrating atrophy and microstructural disorganization in the cortex. Only the advances in modeling and acquisition techniques enabled the necessary ultra-high resolution imaging at 3T and 7T in these neurodegenerative diseases.
Several of the novel image processing, modeling and data analysis techniques were integrated into the open-source hMRI toolbox (http://hMRI.info) making these advances readily accessible to the community and supporting a community-driven toolbox development. The novel methods are widely used and also laid the foundation for novel imaging biomarkers used in (clinical) neuroscience (e.g. NISCI clinical trial on spinal cord injury; https://nisci-2020.eu/).
We achieved in-vivo imaging of the human cortex using quantitative multi-parameter maps (MPMs) with an unprecedented 400 µm isotropic resolution. This was only possible by developing and seamlessly integrating optical prospective motion correction, which addresses the unavoidable head motion artifacts at such high spatial resolution. The MPMs were combined with cutting edge diffusion imaging to estimate features of the cortical microstructure such as markers of fiber geometry and myelination. Different methods for functional MRI with 800 µm isotropic resolution were developed and compared for highest effective resolution taking into account physiological blurring effects (draining vein effects). They allowed for precise layer-specific functional activation studies of columnar cortical structures. The combination of MPM-based myelin markers with ultra-high resolution fMRI demonstrated different myelination levels in the different types of stripes in the secondary visual cortex (V2).
Extraordinary advances in diffusion imaging were achieved by leveraging the potential of the recently commissioned 3T Connectom MRI at the MPI for Human Cognitive and Brain Sciences, which offers the highest maximum gradient amplitude (300 mT/m) for human imaging and is one of just four systems worldwide (https://www.cbs.mpg.de/institute/technical-facilities). Data quality was further improved by a novel flexible radio-frequency (RF) surface coil and optimized data processing pipelines for the high-end hardware – for the first time demonstrating the structure-function relationship between U-fibers and retinotopy in the visual cortices.
We have developed novel data processing methods and biophysical models that integrate information from different MRI contrasts, ranging from relaxometry to diffusion and functional MRI. These models include the quantification of iron concentration, intracortical fibers, superficial white matter and description of tangential structures in the cortex such as ocular dominance columns and stripes in the primary and secondary visual cortex. They are the key link between non-invasive MRI measurements and the underlying microstructure.
Post-mortem MRI (with up to 25 µm resolution) and histology were further developed and applied to tissue specimen of controls and Alzheimer’s disease (AD) patients in order to inform model building and validate the models. We have pushed the limits of current post-mortem histology methods, which are limited by the vagaries of staining efficiency and their 2D view on the brain anatomy. Tissue clearing combined with advanced microscopy and image processing methods allowed for comprehensive imaging of cortical structures. By using proton-induced X-ray emission (PIXE), matrix-assisted laser desorption/ionization (MALDI), mass spectroscopy imaging we were able to estimate high-resolution maps (from macroscopic- down to cellular-resolution) of iron, myelination and variations in myelin properties in the brain.
The novel methods were used to study neurodegenerative changes and focal atrophy of the cortex in AD and posterior cortical atrophy (PCA), a variant of AD, demonstrating atrophy and microstructural disorganization in the cortex. Only the advances in modeling and acquisition techniques enabled the necessary ultra-high resolution imaging at 3T and 7T in these neurodegenerative diseases.
Several of the novel image processing, modeling and data analysis techniques were integrated into the open-source hMRI toolbox (http://hMRI.info) making these advances readily accessible to the community and supporting a community-driven toolbox development. The novel methods are widely used and also laid the foundation for novel imaging biomarkers used in (clinical) neuroscience (e.g. NISCI clinical trial on spinal cord injury; https://nisci-2020.eu/).