Recent learning-based approaches have made astonishing advances in calibrated medical imaging like computerized tomography (CT), yet they struggle to generalize in uncalibrated modalities -- notably magnetic resonance (MR) imaging, where performance is highly sensitive to the differences in MR contrast, resolution, and orientation. This prevents broad applicability to diverse real-world clinical protocols. We introduce Brain-ID, an anatomical representation learning model for brain imaging. With the proposed "mild-to-severe" intra-subject generation, Brain-ID is robust to the subject-specific brain anatomy regardless of the appearance of acquired images (e.g., contrast, deformation, resolution, artifacts). Trained entirely on synthetic data, Brain-ID readily adapts to various downstream tasks through only one layer. We present new metrics to validate the intra- and inter-subject robustness of Brain-ID features, and evaluate their performance on four downstream applications, covering contrast-independent (anatomy reconstruction/contrast synthesis, brain segmentation), and contrast-dependent (super-resolution, bias field estimation) tasks. Extensive experiments on six public datasets demonstrate that Brain-ID achieves state-of-the-art performance in all tasks on different MRI modalities and CT, and more importantly, preserves its performance on low-resolution and small datasets. Code is available at https://github.com/peirong26/Brain-ID.
Detailed computational anatomical models of the entire head are needed for accurate in silico modeling in a variety of transcranial stimulation applications. Models from different subjects help to understand and account for population variability. To this end, we have developed a new library of head models of 20 individuals, segmented from co-aligned multi-modal medical image data. The acquired image modalities allow to accurately model tissues with different material properties, such as electrical conductivity or spatially varying acoustic properties. The usefulness of the models is illustrated for two example applications.
Transcranial brain stimulation (TBS) techniques such as transcranial magnetic stimulation (TMS), transcranial direct current stimulation (tDCS) and others have seen a strong increase as tools in therapy and research within the last 20 years. In order to precisely target the stimulation, it is important to accurately model the individual head anatomy of a subject. Of particular importance is accurate reconstruction of the skull, as it has the strongest impact on the current pathways due to its low conductivity. Thus providing automated tools, which can reliably reconstruct the anatomy of the human head from magnetic resonance (MR) scans would be highly valuable for the application of transcranial stimulation methods. These head models can also be used to inform source localization methods such as EEG and MEG. Automated segmentation of the skull from MR images is, however, challenging as the skull emits very little signal in MR. In order to avoid topological defects, such as holes in the segmentations, a strong model of the skull shape is needed. In this paper we propose a new shape model for skull segmentation based on the so-called convolutional restricted Boltzmann machines (cRBMs). Compared to traditionally used lower-order shape models, such as pair-wise Markov random fields (MRFs), the cRBMs model local shapes in larger spatial neighborhoods while still allowing for efficient inference. We compare the skull segmentation accuracy of our approach to two previously published methods and show significant improvement.
Transcranial brain stimulation (TBS) has been established as a method for modulating and mapping the function of the human brain, and as a potential treatment tool in several brain disorders. Typically, the stimulation is applied using a one-size-fits-all approach with predetermined locations for the electrodes, in electric stimulation (TES), or the coil, in magnetic stimulation (TMS), which disregards anatomical variability between individuals. However, the induced electric field distribution in the head largely depends on anatomical features implying the need for individually tailored stimulation protocols for focal dosing. This requires detailed models of the individual head anatomy, combined with electric field simulations, to find an optimal stimulation protocol for a given cortical target. Considering the anatomical and functional complexity of different brain disorders and pathologies, it is crucial to account for the anatomical variability in order to translate TBS from a research tool into a viable option for treatment. In this article we present a new method, called CHARM, for automated segmentation of fifteen different head tissues from magnetic resonance (MR) scans. The new method compares favorably to two freely available software tools on a five-tissue segmentation task, while obtaining reasonable segmentation accuracy over all fifteen tissues. The method automatically adapts to variability in the input scans and can thus be directly applied to clinical or research scans acquired with different scanners, sequences or settings. We show that an increase in automated segmentation accuracy results in a lower relative error in electric field simulations when compared to anatomical head models constructed from reference segmentations. However, also the improved segmentations and, by implication, the electric field simulations are affected by systematic artifacts in the input MR scans. As long as the artifacts are unaccounted for, this can lead to local simulation differences up to 30% of the peak field strength on reference simulations. Finally, we exemplarily demonstrate the effect of including all fifteen tissue classes in the field simulations against the standard approach of using only five tissue classes and show that for specific stimulation configurations the local differences can reach 10% of the peak field strength.
During the past decade, it became clear that the effects of non-invasive brain stimulation (NIBS) techniques such as transcranial direct current stimulation (tDCS) and transcranial magnetic stimulation (TMS) are substantially influenced by variations in individual head and brain anatomy. In addition to structural variations in the healthy, several psychiatric disorders are characterized by anatomical alterations that are likely to further constrain the intracerebral effects of NIBS. Here, we present high-resolution realistic head models derived from structural magnetic resonance imaging data of 19 healthy adults and 19 patients diagnosed with major depressive disorder (MDD). By using a freely available software package for modelling the effects of different NIBS protocols, we show that our head models are well-suited for assessing inter-individual and between-group variability in the magnitude and focality of tDCS-induced electric fields for two protocols targeting the left dorsolateral prefrontal cortex.
Brain atrophy and white matter hyperintensity (WMH) are critical neuroimaging features for ascertaining brain injury in cerebrovascular disease and multiple sclerosis. Automated segmentation and quantification is desirable but existing methods require high-resolution MRI with good signal-to-noise ratio (SNR). This precludes application to clinical and low-field portable MRI (pMRI) scans, thus hampering large-scale tracking of atrophy and WMH progression, especially in underserved areas where pMRI has huge potential. Here we present a method that segments white matter hyperintensity and 36 brain regions from scans of any resolution and contrast (including pMRI) without retraining. We show results on eight public datasets and on a private dataset with paired high- and low-field scans (3T and 64mT), where we attain strong correlation between the WMH ($\rho$=.85) and hippocampal volumes (r=.89) estimated at both fields. Our method is publicly available as part of FreeSurfer, at: http://surfer.nmr.mgh.harvard.edu/fswiki/WMH-SynthSeg.
Head and brain anatomy have been related to e-field strength induced by transcranial electrical stimulation (tES). Individualization based on anatomic factors require high-quality structural magnetic resonance images, which are not always available. Head circumference (HC) can serve as an alternative means, but its linkage to electric field strength has not yet been established.We simulated electric fields induced by tES based on individual T1w- and T2w-images of 47 healthy adults, for four conventional ("standard") and four corresponding focal ("4x1") electrode montages. Associations of electric field strength with individual HC were calculated using linear mixed models.Larger HC was associated with lower electric field strength across montages. We provide mathematical equations to estimate individual electric field strength based on the HC.HC can be used as an alternative to estimate interindividual differences of the tES-induced electric field strength and to prospectively individualize stimulation dose, e.g., in the clinical context.
Segmentation of brain structures on magnetic resonance imaging (MRI) is a highly relevant neuroimaging topic, as it is a prerequisite for different analyses such as volumetry or shape analysis. Automated segmentation facilitates the study of brain structures in larger cohorts when compared with manual segmentation, which is time-consuming. However, the development of most automated methods relies on large and manually annotated datasets, which limits the generalizability of these methods. Recently, new techniques using synthetic images have emerged, reducing the need for manual annotation. Here we provide a dataset composed of label maps built from publicly available ultra-high resolution ex vivo MRI from 10 whole hemispheres, which can be used to develop segmentation methods using synthetic data. The label maps are obtained with a combination of manual labels for the hypothalamic regions and automated segmentations for the rest of the brain, and mirrored to simulate entire brains. We also provide the pre-processed ex vivo scans, as this dataset can support future projects to include other structures after these are manually segmented.
Magnetic resonance current density imaging (MRCDI) of the human brain aims to reconstruct the current density distribution caused by transcranial electric stimulation from MR-based measurements of the current-induced magnetic fields. So far, the MRCDI data acquisition achieves only a low signal-to-noise ratio, does not provide a full volume coverage and lacks data from the scalp and skull regions. In addition, it is only sensitive to the component of the current-induced magnetic field parallel to the scanner field. The reconstruction problem thus involves coping with noisy and incomplete data, which makes it mathematically challenging. Most existing reconstruction methods have been validated using simulation studies and measurements in phantoms with simplified geometries. Only one reconstruction method, the projected current density algorithm, has been applied to human in-vivo data so far, however resulting in blurred current density estimates even when applied to noise-free simulated data. We analyze the underlying causes for the limited performance of the projected current density algorithm when applied to human brain data. In addition, we compare it with an approach that relies on the optimization of the conductivities of a small number of tissue compartments of anatomically detailed head models reconstructed from structural MR data. Both for simulated ground truth data and human in-vivo MRCDI data, our results indicate that the estimation of current densities benefits more from using a personalized volume conductor model than from applying the projected current density algorithm. In particular, we introduce a hierarchical statistical testing approach as a principled way to test and compare the quality of reconstructed current density images that accounts for the limited signal-to-noise ratio of the human in-vivo MRCDI data and the fact that the ground truth of the current density is unknown for measured data. Our results indicate that the statistical testing approach constitutes a valuable framework for the further development of accurate volume conductor models of the head. Our findings also highlight the importance of tailoring the reconstruction approaches to the quality and specific properties of the available data.
'%3e%3cpath fill='%23012169' d='M0 0v30h60V0z'/%3e%3cpath stroke='%23fff' stroke-width='6' d='m0 0 60 30m0-30L0 30'/%3e%3cpath stroke='%23C8102E' stroke-width='4' d='m0 0 60 30m0-30L0 30' clip-path='url(%23b)'/%3e%3cpath stroke='%23fff' stroke-width='10' d='M30 0v30M0 15h60'/%3e%3cpath stroke='%23C8102E' stroke-width='6' d='M30 0v30M0 15h60'/%3e%3c/g%3e%3c/svg%3e)