ProfessorFernando Calamante

Keywords: MRI, Diffusion MRI, brain connectivity, machine learning, neuroscience.

Recent advances in MRI have revolutionised the way we investigate brain structure, brain function, and brain network connectivity, as well as how these are affected in disease. His lab's main research interests include: (1) the development and application of MRI methods to measure the blood delivery to brain tissue (with a technique known as Perfusion MRI), (2) the use of Diffusion MRI methods to map the structural connections in the brain (that is, the brain 'wiring'), and (3) how to use this 'wiring' information to achieve Super-resolution Imaging (that is, creating images at higher resolution than that of the acquired image data).

In particular, Professor Calamante specialise in the development of the two main Perfusion MRI techniques: Dynamic Susceptibility Contrast MRI (DSC-MRI) and Arterial Spin Labelling (ASL). The former is playing a key role in many clinical applications (such as in stroke and tumours), while the latter provides a powerful quantitative tool to characterise how the various brain regions are connected into functional networks. In turn, the technique of Diffusion MRI provides a unique means to reconstruct a representation of the white matter pathways in the brain (the 'wiring' of the brain), and thus to study non-invasively how the various brain regions are structurally connected into networks. Professor Calamante has also developed a super-resolution imaging approach, which provides a means to exploit the 'wiring' information obtained from Diffusion MRI to achieve image resolution not previously possible in the in vivo human brain non-invasively. This method not only can generate images with exquisite image detail, but also provides a new way to combine multi-modal imaging, such as structural and functional connectivity information, thus allowing an overall structural-functional description of the brain and its networks.

Novel connectivity methods can become essential tools in neuroscience, and should greatly contribute to further our understanding of brain networks in the healthy brain, and the effects of their disruptions in disease. These methods can provide new diagnostic markers, which are of great importance for understanding how the healthy brain functions, how healthy brain networks are perturbed by disease, and could help explain the effect of treatment.

Diffusion MRI methods to study brain structural connectivity

Professor Calamante's and his colleagues have shown that the technique of Diffusion MRI can be used to obtain an estimate of the macroscopic white matter fibre orientations at each location in the brain, even in the presence of multiple fibre populations, which in turn can be used with a fibre-tracking algorithm to reconstruct a representation of the white matter pathways in the brain. For the case of whole-brain fibre-tracking, several million tracks (also known as streamlines) are generated, thus providing an overall representation of white matter pathways throughout the brain. The advanced analysis methods for Diffusion MRI developed by Prof. Calamante and colleagues have been released as an open access software tool (MRtrix), which has been widely used internationally.

Figure legend: whole-brain fibre-tracking data. (a) Human brain. (b) Mouse brain. The colour-coding indicates the local fibre orientation (red left-right, green dorsal-ventral, blue cranial-caudal).

Super-resolution MRI methods for the brain connectome

Professor Calamante and his team have shown that whole-brain diffusion MRI fibre-tracking data can be used to generate images with resolution higher than the resolution of the acquired data (i.e. to achieve 'super-resolution'), with a technique they called super-resolution track-weighted imaging (TWI). This methodology provides a natural means to combine multi-modal imaging and, in particular, structural and functional connectivity information into a single image, and therefore can play a major role in the characterisation of the Brain Connectome (a comprehensive map of neural connections in the brain).

Novel MRI methods to study dynamic brain connectivity

Interest in the study of brain connectivity is growing, particularly in understanding the the structural/functional relation and how these two distinct forms of connectivity relate to each other. Recent studies demonstrated that dynamic characterizations of functional connectivity provide a more comprehensive description of functional brain networks. However, how these dynamic fluctuations influence the adaptive properties of neural information processing remains an open question that requires the development of new methods. One of the research interests of Calamante's group is in the development of novel ways to investigate dynamic brain connectivity. For example, they recently developed the technique of track-weighted dynamic functional connectivity (TW-dFC), which fuses structural/functional connectivity data into a 4D image, providing a new approach to investigate dynamic connectivity. The structural connectivity information effectively 'constrains' the extremely large number of possible connections in the functional connectivity data (i.e. each voxel's connection to each other), thus providing a way of reducing the problem's dimensionality, while still maintaining key data features.

These methods can not only be used to study brain dynamics, but the connectivity information can also be exploited to perform data-driven parcellation of the brain. For example, Calamante's group has recently demonstrated how the combination of structural connectivity, functional connectivity and dynamic connectivity can be exploited to parcellate the corpus callosum (one of the largest white matters structures in the human brain). This white matter parcellation can be subsequently used, in combination with Diffusion MRI fibre-tracking, to achieve cortical segregation. All these results are obtained in a data-driven way, without having to rely on a prior assumption of how the white matter or cerebral cortex is subdivided.

Figure legend: (a) Parcellation of the corpus callosum using TW-dFC and independent component analysis (mid-sagittal region); colours were arbitrarily chosen to aid visual differentiation. The cortical connections for the tracks traversing each of the white matter parcels are shown as a lateral sagittal slice (b) and viewed from above (c).

Perfusion MRI: novel methods to image cerebral blood flow and brain function

MRI provides a powerful non-invasive tool to measure the rate of blood delivery to brain tissue (also known as cerebral perfusion). Cerebral perfusion plays an essential role in tissue viability and function, and it is implicated in many diseases (such as stroke, epilepsy, and tumours). Professor Calamante is among the international leaders in the development of Perfusion MRI methods. He has been involved in the development of novel methods to measure and analyse Perfusion MRI data, with the aim to apply these methods to investigate brain disorders (including stroke, epilepsy and dementia). In particular, his work has had a direct impact on how Perfusion MRI is used for the clinical management of stroke. He has also shown that Perfusion MRI, using the technique known as Arterial Spin Labelling (ASL), can be a powerful tool to study brain functional connectivity and connectomics. Perfusion MRI has been shown to have a number of important advantages compared with more traditional BOLD fMRI methods to functional networks, including quantitation, reduced image distortions and signal drop-out, increased spatial specificity, and increased sensitivity to low task frequency paradigms.