White matter fiber clustering is an important strategy for white matter parcellation, which enables quantitative analysis of brain connections in health and disease. In combination with expert neuroanatomical labeling, data-driven white matter fiber clustering is a powerful tool for creating atlases that can model white matter anatomy across individuals. While widely used fiber clustering approaches have shown good performance using classical unsupervised machine learning techniques, recent advances in deep learning reveal a promising direction toward fast and effective fiber clustering. In this work, we propose a novel deep learning framework for white matter fiber clustering, Deep Fiber Clustering (DFC), which solves the unsupervised clustering problem as a self-supervised learning task with a domain-specific pretext task to predict pairwise fiber distances. This process learns a high-dimensional embedding feature representation for each fiber, regardless of the order of fiber points reconstructed during tractography. We design a novel network architecture that represents input fibers as point clouds and allows the incorporation of additional sources of input information from gray matter parcellation. Thus, DFC makes use of combined information about white matter fiber geometry and gray matter anatomy to improve the anatomical coherence of fiber clusters. In addition, DFC conducts outlier removal naturally by rejecting fibers with low cluster assignment probability. We evaluate DFC on three independently acquired cohorts, including data from 220 individuals across genders, ages (young and elderly adults), and different health conditions (healthy control and multiple neuropsychiatric disorders). We compare DFC to several state-of-the-art white matter fiber clustering algorithms. Experimental results demonstrate superior performance of DFC in terms of cluster compactness, generalization ability, anatomical coherence, and computational efficiency.
Acupuncture point names written in Chinese Han characters often provide clinically useful information in both their literal and figurative meanings about location and therapeutic use. The World Health Organization (WHO) standard acupuncture nomenclature includes these names in Han characters in an unusual array that includes both "original" forms and, in parentheses, simplified forms. Construction of a multilingual table of acupuncture point names during development of a database revealed that the assumption that the "original" form in the WHO nomenclature was the traditional Chinese character was frequently false. The Han character forms in the pdf of the 2009 reprint of WHO Standard Acupuncture Point Locations were carefully compared with Han characters used in traditional and simplified Chinese, Japanese and Korean writing systems. This work utilized three online tools: UnicodePlus, Unihan Database Lookup, and Wiktionary. Only 48% of the "original" character forms were traditional Chinese characters. The Unicode number was correct in 99%, but in most cases the East Asian font used was not a traditional Chinese one. The issue about Han character forms was also found in all earlier versions of the WHO standard acupuncture nomenclature. Other detected problems included the use of wrong characters for an "original" character form in one name and for a simplified character form in another name. The WHO standard acupuncture nomenclature should be revised with a focus on accuracy in the usage of Han characters.
Measuring and understanding functional fetal brain development in utero is critical for the study of the developmental foundations of our cognitive abilities, possible early detection of disorders, and their prevention. Thalamocortical connections are an intricate component of shaping the cortical layout, but so far, only ex-vivo studies provide evidence of how axons enter the sub-plate and cortex during this highly dynamic phase. Evidence for normal in-utero development of the functional thalamocortical connectome in humans is missing. Here, we modeled fetal functional thalamocortical connectome development using in-utero functional magnetic resonance imaging in fetuses observed from 19th to 40th weeks of gestation (GW). We observed a peak increase of thalamocortical functional connectivity strength between 29th and 31st GW, right before axons establish synapses in the cortex. The cortico-cortical connectivity increases in a similar time window, and exhibits significant functional laterality in temporal-superior, -medial, and -inferior areas. Homologous regions exhibit overall similar mirrored connectivity profiles, but this similarity decreases during gestation giving way to a more diverse cortical interconnectedness. Our results complement the understanding of structural development of the human connectome and may serve as the basis for the investigation of disease and deviations from a normal developmental trajectory of connectivity development.
The human thalamus is a highly connected brain structure, which is key for the control of numerous functions and is involved in several neurological disorders. Recently, neuroimaging studies have increasingly focused on the volume and connectivity of the specific nuclei comprising this structure, rather than looking at the thalamus as a whole. However, accurate identification of cytoarchitectonically designed histological nuclei on standard in vivo structural MRI is hampered by the lack of image contrast that can be used to distinguish nuclei from each other and from surrounding white matter tracts. While diffusion MRI may offer such contrast, it has lower resolution and lacks some boundaries visible in structural imaging. In this work, we present a Bayesian segmentation algorithm for the thalamus. This algorithm combines prior information from a probabilistic atlas with likelihood models for both structural and diffusion MRI, allowing segmentation of 25 thalamic labels per hemisphere informed by both modalities. We present an improved probabilistic atlas, incorporating thalamic nuclei identified from histology and 45 white matter tracts surrounding the thalamus identified in ultra-high gradient strength diffusion imaging. We present a family of likelihood models for diffusion tensor imaging, ensuring compatibility with the vast majority of neuroimaging datasets that include diffusion MRI data. The use of these diffusion likelihood models greatly improves identification of nuclear groups versus segmentation based solely on structural MRI. Dice comparison of 5 manually identifiable groups of nuclei to ground truth segmentations show improvements of up to 10 percentage points. Additionally, our chosen model shows a high degree of reliability, with median test-retest Dice scores above 0.85 for four out of five nuclei groups, whilst also offering improved detection of differential thalamic involvement in Alzheimer's disease (AUROC 81.98%). The probabilistic atlas and segmentation tool will be made publicly available as part of the neuroimaging package FreeSurfer.
Diffusion MRI tractography is an advanced imaging technique that enables in vivo mapping of the brain's white matter connections. White matter parcellation classifies tractography streamlines into clusters or anatomically meaningful tracts. It enables quantification and visualization of whole-brain tractography. Currently, most parcellation methods focus on the deep white matter (DWM), whereas fewer methods address the superficial white matter (SWM) due to its complexity. We propose a novel two-stage deep-learning-based framework, Superficial White Matter Analysis (SupWMA), that performs an efficient and consistent parcellation of 198 SWM clusters from whole-brain tractography. A point-cloud-based network is adapted to our SWM parcellation task, and supervised contrastive learning enables more discriminative representations between plausible streamlines and outliers for SWM. We train our model on a large-scale tractography dataset including streamline samples from labeled long- and medium-range (over 40 mm) SWM clusters and anatomically implausible streamline samples, and we perform testing on six independently acquired datasets of different ages and health conditions (including neonates and patients with space-occupying brain tumors). Compared to several state-of-the-art methods, SupWMA obtains highly consistent and accurate SWM parcellation results on all datasets, showing good generalization across the lifespan in health and disease. In addition, the computational speed of SupWMA is much faster than other methods.
The cerebellum is ontogenetically one of the first structures to develop in the central nervous system; nevertheless, it has been only recently reconsidered for its significant neurobiological, functional, and clinical relevance in humans. Thus, it has been a relatively under-studied compared to the cerebrum. Currently, non-invasive imaging modalities can barely reach the necessary resolution to unfold its entire, convoluted surface, while only histological analyses can reveal local information at the micrometer scale. Herein, we used the BigBrain dataset to generate area and point-wise thickness measurements for all layers of the cerebellar cortex and for each lobule in particular. We found that the overall surface area of the cerebellar granular layer (including Purkinje cells) was 1,732 cm2 and the molecular layer was 1,945 cm2. The average thickness of the granular layer is 0.88 mm (± 0.83) and that of the molecular layer is 0.32 mm (± 0.08). The cerebellum (both granular and molecular layers) is thicker at the depth of the sulci and thinner at the crowns of the gyri. Globally, the granular layer is thicker in the lateral-posterior-inferior region than the medial-superior regions. The characterization of individual layers in the cerebellum achieved herein represents a stepping-stone for investigations interrelating structural and functional connectivity with cerebellar architectonics using neuroimaging, which is a matter of considerable relevance in basic and clinical neuroscience. Furthermore, these data provide templates for the construction of cerebellar topographic maps and the precise localization of structural and functional alterations in diseases affecting the cerebellum.
Sleep disturbances are strongly associated with mild traumatic brain injury (mTBI) and post-traumatic stress disorder (PTSD). PTSD and mTBI have been linked to alterations in white matter (WM) microstructure, but whether poor sleep quality has a compounding effect on WM remains largely unknown. We evaluated sleep and diffusion magnetic resonance imaging (dMRI) data from 180 male post-9/11 veterans diagnosed with (1) PTSD (n = 38), (2) mTBI (n = 25), (3) comorbid PTSD+mTBI (n = 94), and (4) a control group with neither PTSD nor mTBI (n = 23). We compared sleep quality (Pittsburgh Sleep Quality Index, PSQI) between groups using ANCOVAs and calculated regression and mediation models to assess associations between PTSD, mTBI, sleep quality, and WM. Veterans with PTSD and comorbid PTSD+mTBI reported poorer sleep quality than those with mTBI or no history of PTSD or mTBI (p = 0.012 to <0.001). Poor sleep quality was associated with abnormal WM microstructure in veterans with comorbid PTSD+mTBI (p < 0.001). Most importantly, poor sleep quality fully mediated the association between greater PTSD symptom severity and impaired WM microstructure (p < 0.001). Our findings highlight the significant impact of sleep disturbances on brain health in veterans with PTSD+mTBI, calling for sleep-targeted interventions.
OBJECTIVES: Approximately 50% of comatose patients after cardiac arrest never regain consciousness. Cerebral ischaemia may lead to cytotoxic and/or vasogenic oedema, which can be detected by diffusion tensor imaging (DTI). Here, we evaluate the potential value of free water corrected mean diffusivity (MD) and fractional anisotropy (FA) based on DTI, for the prediction of neurological recovery of comatose patients after cardiac arrest. METHODS: A total of 50 patients after cardiac arrest were included in this prospective cohort study in two Dutch hospitals. DTI was obtained 2-4 days after cardiac arrest. Outcome was assessed at 6 months, dichotomised as poor (cerebral performance category 3-5; n = 20) or good (n = 30) neurological outcome. We calculated the whole brain mean MD and FA and compared between patients with good and poor outcomes. In addition, we compared a preliminary prediction model based on clinical parameters with or without the addition of MD and FA. RESULTS: We found significant differences between patients with good and poor outcome of mean MD (good: 726 [702-740] × 10-6 mm2/s vs. poor: 663 [575-736] × 10-6 mm2/s; p = 0.01) and mean FA (0.30 ± 0.03 vs. 0.28 ± 0.03; p = 0.03). An exploratory prediction model combining clinical parameters, MD and FA increased the sensitivity for reliable prediction of poor outcome from 60 to 85%, compared to the model containing clinical parameters only, but confidence intervals are overlapping. CONCLUSIONS: Free water-corrected MD and FA discriminate between patients with good and poor outcomes after cardiac arrest and hold the potential to add to multimodal outcome prediction. KEY POINTS: • Whole brain mean MD and FA differ between patients with good and poor outcome after cardiac arrest. • Free water-corrected MD can better discriminate between patients with good and poor outcome than uncorrected MD. • A combination of free water-corrected MD (sensitive to grey matter abnormalities) and FA (sensitive to white matter abnormalities) holds potential to add to the prediction of outcome.
BACKGROUND AND OBJECTIVES: While chronological age is one of the most influential determinants of poststroke outcomes, little is known of the impact of neuroimaging-derived biological "brain age." We hypothesized that radiomics analyses of T2-FLAIR images texture would provide brain age estimates and that advanced brain age of patients with stroke will be associated with cardiovascular risk factors and worse functional outcomes.
METHODS: We extracted radiomics from T2-FLAIR images acquired during acute stroke clinical evaluation. Brain age was determined from brain parenchyma radiomics using an ElasticNet linear regression model. Subsequently, relative brain age (RBA), which expresses brain age in comparison with chronological age-matched peers, was estimated. Finally, we built a linear regression model of RBA using clinical cardiovascular characteristics as inputs and a logistic regression model of favorable functional outcomes taking RBA as input.
RESULTS: We reviewed 4,163 patients from a large multisite ischemic stroke cohort (mean age = 62.8 years, 42.0% female patients). T2-FLAIR radiomics predicted chronological ages (mean absolute error = 6.9 years, r = 0.81). After adjustment for covariates, RBA was higher and therefore described older-appearing brains in patients with hypertension, diabetes mellitus, a history of smoking, and a history of a prior stroke. In multivariate analyses, age, RBA, NIHSS, and a history of prior stroke were all significantly associated with functional outcome (respective adjusted odds ratios: 0.58, 0.76, 0.48, 0.55; all p-values < 0.001). Moreover, the negative effect of RBA on outcome was especially pronounced in minor strokes.
DISCUSSION: T2-FLAIR radiomics can be used to predict brain age and derive RBA. Older-appearing brains, characterized by a higher RBA, reflect cardiovascular risk factor accumulation and are linked to worse outcomes after stroke.
Monitoring time dependence with diffusion MRI yields observables sensitive to compartment sizes (restricted diffusion) and membrane permeability (water exchange). However, restricted diffusion and exchange have opposite effects on the diffusion-weighted signal, which can lead to errors in parameter estimates. In this work, we propose a signal representation that incorporates the effects of both restricted diffusion and exchange up to second order in b-value and is compatible with gradient waveforms of arbitrary shape. The representation features mappings from a gradient waveform to two scalars that separately control the sensitivity to restriction and exchange. We demonstrate that these scalars span a two-dimensional space that can be used to choose waveforms that selectively probe restricted diffusion or exchange, eliminating the correlation between the two phenomena. We found that waveforms with specific but unconventional shapes provide an advantage over conventional pulsed and oscillating gradient acquisitions. We also show that parametrization of waveforms into a two-dimensional space can be used to understand protocols from other approaches that probe restricted diffusion and exchange. For example, we found that the variation of mixing time in filter-exchange imaging corresponds to variation of our exchange-weighting scalar at a fixed value of the restriction-weighting scalar. The proposed signal representation was evaluated using Monte Carlo simulations in identical parallel cylinders with hexagonal and random packing as well as parallel cylinders with gamma-distributed radii. Results showed that the approach is sensitive to sizes in the interval 4-12 μm $$ \upmu \mathrmm $$ and exchange rates in the simulated range of 0 to 20 s - 1 $$ \mathrms^-1 $$ , but also that there is a sensitivity to the extracellular geometry. The presented theory constitutes a simple and intuitive description of how restricted diffusion and exchange influence the signal as well as a guide to protocol design capable of separating the two effects.