The functional analysis of the left atrium (LA) is important for evaluating cardiac health and understanding diseases like atrial fibrillation. Cine MRI is ideally placed for the detailed 3D characterization of LA motion and deformation but is lacking appropriate acquisition and analysis tools. Here, we propose tools for the Analysis of Left Atrial Displacements and DeformatIons using online learning neural Networks (Aladdin) and present a technical feasibility study on how Aladdin can characterize 3D LA function globally and regionally. Aladdin includes an online segmentation and image registration network, and a strain calculation pipeline tailored to the LA. We create maps of LA Displacement Vector Field (DVF) magnitude and LA principal strain values from images of 10 healthy volunteers and 8 patients with cardiovascular disease (CVD), of which 2 had large left ventricular ejection fraction (LVEF) impairment. We additionally create an atlas of these biomarkers using the data from the healthy volunteers. Results showed that Aladdin can accurately track the LA wall across the cardiac cycle and characterize its motion and deformation. Global LA function markers assessed with Aladdin agree well with estimates from 2D Cine MRI. A more marked active contraction phase was observed in the healthy cohort, while the CVD LVEF↓ group showed overall reduced LA function. Aladdin is uniquely able to identify LA regions with abnormal deformation metrics that may indicate focal pathology. We expect Aladdin to have important clinical applications as it can non-invasively characterize atrial pathophysiology. All source code and data are available at: https://github.com/cgalaz0 1/aladdin_cmr_la.
Purpose: field inhomogeneity is a common problem in high field brain MRI ( T). Parallel‐transmit methods that adjust the field channelwise often require valuable scan time. Group‐optimized phase shims are presented to increase or attenuate the field in specific brain regions, omitting personalized calibrations and potentially enabling reduced FOV acquisitions or artifact reduction.Methods: Channelwise maps were obtained for seven participants using an 8Tx/32Rx coil and a 7 T MRI scanner. Two regional shim settings ( shims) were calculated: one to increase the field in the cerebellum and the other to increase the field in the occipital lobe while attenuating the field in the frontal lobe. maps from five participants outside the design group were used to simulate the profiles, and seven were scanned to evaluate the implementation of the shims using maps, 3D EPI, GRE acquisitions, and a visual fMRI experiment.Results: Both regional shim settings successfully amplified the field in the selected ROIs resulting in improved yield and increased tSNR in the 3D EPI images and fMRI experiments compared to the circularly polarized shim mode. The attenuating shim decreased in the frontal ROI, decreasing fold‐over artifacts in a reduced FOV, lowering g‐factors in accelerated scans with high undersampling factors and resulted in improved BOLD responses in the visual fMRI experiment.Conclusion: Regional shim settings remove the need for time‐consuming, personalized measurements and calibrations. The attenuating shim allows for signal reduction within the power limits of the rf‐coil, reducing artifacts while improving the field in selected ROIs.
The cerebellum plays a crucial role in the control of hand movements, enabling fine motor skills such as clapping and writing. Neurological diseases can affect the cerebellum, often leading to motor impairment. However, the cerebellar organisation of specific motor and sensory tasks in humans is under-explored in vivo compared to the neocortex, due to a lack of acquisition and analysis methods that effectively portray cerebellar activation in-vivo due to the cerebellum’s thin and highly-foliated cortex. In the neocortex, by comparison, response differences between distinct motor and sensory tasks have been reported, implying an extensive sensorimotor organisation. Here, we studied the cerebellar functional responses during three distinct tasks: flexing, extending and stroking of digits 1, 3 and 5 using B1-shimmed 7T functional MRI. We analysed the data in the standard 3D-functional space and the surface space, respecting the dense foliation of the cerebellum. All tasks elicited individual digit responses, engaging the cerebellar cortex in distinct ways: Digit extension yielded larger, more bilateral activation clusters and less distinct progressions of digit representations in comparison to flexing and stroking tasks. The stroking responses were found more medial in the anterior lobe than the flexing and extending clusters. The anterior lobe clusters were larger for the extending and flexing tasks than for the stroking task. These results imply that the cerebellum is engaged differently when tasks with differing sensory/motor components are performed and that these differences exist on a (sub)millimetre scale, akin to the mesoscale organisation in the cerebral cortex.
Cerebellar function and structure remain less studied compared to the cerebrum, and while high-resolution 7T gives superior anatomical information, methods to visualise its microstructure in vivo are not widely available. However, improved understanding of cerebellar microstructure is highly clinically relevant: the granular layer of the cerebellum, located in the grey matter, contains the majority of the brain’s neurons and has been shown in post mortem studies to be affected by various diseases. The limited spatial resolution of MR does not resolve individual cell layers, but advanced diffusion techniques can infer properties of the microstructure. The aim of this work was to implement a combination of spherical b-tensor encoding using free waveforms, B1+ field shimming and super-resolution reconstruction (SRR) at 7T and evaluate the method in healthy volunteers. Diffusion-weighted signals and normalised diffusion signals were compared between the grey and white matter of the cerebellum and the occipital lobe as well as across 30 smaller ROIs in the cerebellum. The normalised diffusion signal obtained with spherical tensor encoding relates to cell density, and was found to be increased in the cerebellar cortex relative to the occipital cortex, consistent with previous findings. In the cerebellar cortex, the normalised diffusion signal followed a pattern similar to the thickness of the granular layer reported post mortem. Specifically, increased normalised diffusion signal values were found in lobules VII-VIII and Crus I/II, while lower values were found in lobules I-V. We successfully implemented spherical tensor encoding and SRR at 7T to image the human cerebellum, showing consistent results with those obtained at 3T, with an increased normalised diffusion signal in the cerebellum relative to that found in the cerebral cortex.
The human cerebellar cortex contains the highest density and number of neurons in the human brain, yet this thin and tightly-folded structure has remained largely inaccessible to in-vivo imaging. We introduce an imaging framework to enable high-resolution, comprehensive imaging of the human cerebellar cortex. We validated the in-vivo estimates of cortical morphology against post-mortem measures. Crucially, our findings challenge the commonly held view of the cerebellar cortex as being uniform. Our findings reveal interlobular heterogeneity in both cortical morphology and vascular organization. We demonstrate that this spatial heterogeneity correlates with granular layer cell density. This convergence of morphology, cytoarchitecture and vascularization offers new mechanistic insights and reframes how cerebellar structure and function should be interpreted in health and disease.