ONGOING
Modelling human pRFs using 9.4T fMRI
In 2024, Jonas Bause, Susanne Stoll and me received the DFG Grant "How is touch represented in the human brain?". In this project, we use human 9.4T fMRI and large-scale MR-compatible tactile stimulators to, for the first time, provide a comprehensive and detailed description of the tactile population receptive field (pRF) architecture of the human body. We will first focus on mapping the whole finger, provide the code and toolbox open source to replicate the analyses, and will then move to other body parts, such as the hand palm and the arm. This project will significantly enhance our understanding of the architecture of tactile representations in the human brain. More info
Modelling human pRFs using 9.4T fMRI
In 2024, Jonas Bause, Susanne Stoll and me received the DFG Grant "How is touch represented in the human brain?". In this project, we use human 9.4T fMRI and large-scale MR-compatible tactile stimulators to, for the first time, provide a comprehensive and detailed description of the tactile population receptive field (pRF) architecture of the human body. We will first focus on mapping the whole finger, provide the code and toolbox open source to replicate the analyses, and will then move to other body parts, such as the hand palm and the arm. This project will significantly enhance our understanding of the architecture of tactile representations in the human brain. More info
EVENT
Conference on Computational Modelling, Artificial Intelligence and Ultra-high Field MRI
The brain-in-depth (BID) Conference is an annual conference on ultra-high resolution imaging that my group organizes together with collaborators from the Otto-von-Guericke University Magdeburg, the DZNE, the Max Planck Institute for Human Cognitive and Brain Sciences and the Hertie Institute for Clinical Brain Research. BID-2024 took place 14/15 March 2024 in Tübingen and focused on combining computational modelling, artificial intelligence and ultra-high field MRI. More info
Conference on Computational Modelling, Artificial Intelligence and Ultra-high Field MRI
The brain-in-depth (BID) Conference is an annual conference on ultra-high resolution imaging that my group organizes together with collaborators from the Otto-von-Guericke University Magdeburg, the DZNE, the Max Planck Institute for Human Cognitive and Brain Sciences and the Hertie Institute for Clinical Brain Research. BID-2024 took place 14/15 March 2024 in Tübingen and focused on combining computational modelling, artificial intelligence and ultra-high field MRI. More info
METHOD
Shared Response Modeling (SRM) of 7T-fMRI Data
Shared response modeling (SRM) is a technique that allows group analyses by mapping individual stimulus-driven responses to a lower dimensional shared feature space. This can facilitate group analyses of ultra-high field imaging data, because no smoothing or normalization is required for group statistics. Here, we combine SRM with column-based decoding (C-SRM), and show that the number of columns that optimally describes finger maps in primary somatosensory cortex (SI) is higher in younger compared to older adults, indicating a greater columnar size in older adults’ SI. We provide first evidence that the columnar architecture of a functional area changes with increasing age. Read the full article: [link].
Shared Response Modeling (SRM) of 7T-fMRI Data
Shared response modeling (SRM) is a technique that allows group analyses by mapping individual stimulus-driven responses to a lower dimensional shared feature space. This can facilitate group analyses of ultra-high field imaging data, because no smoothing or normalization is required for group statistics. Here, we combine SRM with column-based decoding (C-SRM), and show that the number of columns that optimally describes finger maps in primary somatosensory cortex (SI) is higher in younger compared to older adults, indicating a greater columnar size in older adults’ SI. We provide first evidence that the columnar architecture of a functional area changes with increasing age. Read the full article: [link].
METHOD
The 3D Architecture of the Human Hand Area described with 7T MRI
Septal boundaries in the barrel cortical field are a key feature of SI organization in mice and rats. Many monkey species are equipped with myelin-poor septa between individual finger representations. Here, we combined functional and structural 7T MRI and tested whether similar layer-specific, myelin-poor boundaries also exist in the human brain. We acquired functional and structural 7T MRI data in young adults, and combined layer-specific myelin mapping with population receptive field (pRF) mapping. We show that even though individual fingers are represented as distinct units, the underlying structural architecture is essentially non-topographic, and does neither show myelin-poor septa between finger representations nor structural differences between individual fingers. Read the full article [link].
The 3D Architecture of the Human Hand Area described with 7T MRI
Septal boundaries in the barrel cortical field are a key feature of SI organization in mice and rats. Many monkey species are equipped with myelin-poor septa between individual finger representations. Here, we combined functional and structural 7T MRI and tested whether similar layer-specific, myelin-poor boundaries also exist in the human brain. We acquired functional and structural 7T MRI data in young adults, and combined layer-specific myelin mapping with population receptive field (pRF) mapping. We show that even though individual fingers are represented as distinct units, the underlying structural architecture is essentially non-topographic, and does neither show myelin-poor septa between finger representations nor structural differences between individual fingers. Read the full article [link].
METHOD
Topographic layer imaging as a tool to track Neurodegenerative Disease Spread in M1
One important feature that is often left out when diagnosing and analyzing disease spread in primary motor cortex (M1) is the inhomogeneous architecture of M1 with respect to (i) cortical layers and (ii) topographic units. We propose that combining 3D layer modelling with topographic mapping serves as ideal tool to understand which microstructural changes in M1 determine disease progression. Read the full article: [link]
Topographic layer imaging as a tool to track Neurodegenerative Disease Spread in M1
One important feature that is often left out when diagnosing and analyzing disease spread in primary motor cortex (M1) is the inhomogeneous architecture of M1 with respect to (i) cortical layers and (ii) topographic units. We propose that combining 3D layer modelling with topographic mapping serves as ideal tool to understand which microstructural changes in M1 determine disease progression. Read the full article: [link]
METHOD
In vivo layer definition in the Human Cortex
The layer-dependent myeloarchitecture of the human cortex is used in post mortem tissue to define cortical layers, but has only recently been investigated in the living human brain. We used 0.5 mm isotropic qT1 tissue contrast acquired using a 7 T MRI scanner in living individuals to differentiate between outer layers, middle layers and inner layers in human primary somatosensory cortex (read the full article [link]) and between superficial layers, layer 5a, layer 5b and layer 6 in human primary motor cortex (read the full article [link]). This paves the way for using individualized segmentation and analyses methods to investigate the layer architecture of the living human cortex.
In vivo layer definition in the Human Cortex
The layer-dependent myeloarchitecture of the human cortex is used in post mortem tissue to define cortical layers, but has only recently been investigated in the living human brain. We used 0.5 mm isotropic qT1 tissue contrast acquired using a 7 T MRI scanner in living individuals to differentiate between outer layers, middle layers and inner layers in human primary somatosensory cortex (read the full article [link]) and between superficial layers, layer 5a, layer 5b and layer 6 in human primary motor cortex (read the full article [link]). This paves the way for using individualized segmentation and analyses methods to investigate the layer architecture of the living human cortex.
METHOD
Model the human cortex in 3D
In cognitive neuroscience, brain-behaviour relationships are usually mapped onto a two-dimensional cortical sheet. Cortical layers are a critical but often ignored third dimension of human cortical function. We explain why modelling the human cortex in three dimensions allows novel and unprecedented insights into the encoding schemes of human cognition. Read the full article: [link]
Model the human cortex in 3D
In cognitive neuroscience, brain-behaviour relationships are usually mapped onto a two-dimensional cortical sheet. Cortical layers are a critical but often ignored third dimension of human cortical function. We explain why modelling the human cortex in three dimensions allows novel and unprecedented insights into the encoding schemes of human cognition. Read the full article: [link]
METHOD
A new computational framework for 7T fMRI
One of the principal goals in fMRI is the detection of local activation in the human brain. However, lack of statistical power and inflated false positive rates have recently been identified as major problems. Here, we introduce a novel non-parametric and threshold-free software package called LISA to address this demand. LISA uses a non-linear filter for incorporating spatial context without sacrificing spatial precision. Compared to widely used other methods (e.g., SPM, FLS), it shows a boost in statistical power and allows a more reliable detection of small activation areas. The spatial sensitivity of LISA makes it particularly suitable for the analysis of fMRI data acquired at ultrahigh field (≥7 Tesla). Read the full article: [link]
A new computational framework for 7T fMRI
One of the principal goals in fMRI is the detection of local activation in the human brain. However, lack of statistical power and inflated false positive rates have recently been identified as major problems. Here, we introduce a novel non-parametric and threshold-free software package called LISA to address this demand. LISA uses a non-linear filter for incorporating spatial context without sacrificing spatial precision. Compared to widely used other methods (e.g., SPM, FLS), it shows a boost in statistical power and allows a more reliable detection of small activation areas. The spatial sensitivity of LISA makes it particularly suitable for the analysis of fMRI data acquired at ultrahigh field (≥7 Tesla). Read the full article: [link]
FUNDING
DFG-Sachbeihilfe: "How is touch represented in the human brain?" (2024-2027)
SFB-1436: Central-Project "Human Imaging at Mesoscale" (2020-2024)
DFG-Sachbeihilfe: "How is touch represented in the human brain?" (2024-2027)
SFB-1436: Central-Project "Human Imaging at Mesoscale" (2020-2024)
