Zaixu Cui

PANDA: a pipeline toolbox for analyzing brain diffusion images

Diffusion magnetic resonance imaging (dMRI) is widely used in both scientific research and clinical practice in in-vivo studies of the human brain. While a number of post-processing packages have been developed, fully automated processing of dMRI datasets remains challenging. Here, we developed a MATLAB toolbox named “Pipeline for Analyzing braiN Diffusion imAges” (PANDA) for fully automated processing of brain diffusion images. The processing modules of a few established packages, including FMRIB Software Library (FSL), Pipeline System for Octave and Matlab (PSOM), Diffusion Toolkit and MRIcron, were employed in PANDA. Using any number of raw dMRI datasets from different subjects, in either DICOM or NIfTI format, PANDA can automatically perform a series of steps to process DICOM/NIfTI to diffusion metrics [e.g., fractional anisotropy (FA) and mean diffusivity (MD)] that are ready for statistical analysis at the voxel-level, the atlas-level and the Tract-Based Spatial Statistics (TBSS)-level and can finish the construction of anatomical brain networks for all subjects. In particular, PANDA can process different subjects in parallel, using multiple cores either in a single computer or in a distributed computing environment, thus greatly reducing the time cost when dealing with a large number of datasets. In addition, PANDA has a friendly graphical user interface (GUI), allowing the user to be interactive and to adjust the input/output settings, as well as the processing parameters. As an open-source package, PANDA is freely available at http://www.nitrc.org/projects/panda/. This novel toolbox is expected to substantially simplify the image processing of dMRI datasets and facilitate human structural connectome studies.

Disrupted white matter connectivity underlying developmental dyslexia: A machine learning approach

Developmental dyslexia has been hypothesized to result from multiple causes and exhibit multiple manifestations, implying a distributed multidimensional effect on human brain. The disruption of specific white‐matter (WM) tracts/regions has been observed in dyslexic children. However, it remains unknown if developmental dyslexia affects the human brain WM in a multidimensional manner. Being a natural tool for evaluating this hypothesis, the multivariate machine learning approach was applied in this study to compare 28 school‐aged dyslexic children with 33 age‐matched controls. Structural magnetic resonance imaging (MRI) and diffusion tensor imaging were acquired to extract five multitype WM features at a regional level: white matter volume, fractional anisotropy, mean diffusivity, axial diffusivity, and radial diffusivity. A linear support vector machine (LSVM) classifier achieved an accuracy of 83.61% using these MRI features to distinguish dyslexic children from controls. Notably, the most discriminative features that contributed to the classification were primarily associated with WM regions within the putative reading network/system (e.g., the superior longitudinal fasciculus, inferior fronto‐occipital fasciculus, thalamocortical projections, and corpus callosum), the limbic system (e.g., the cingulum and fornix), and the motor system (e.g., the cerebellar peduncle, corona radiata, and corticospinal tract). These results were well replicated using a logistic regression classifier. These findings provided direct evidence supporting a multidimensional effect of developmental dyslexia on WM connectivity of human brain, and highlighted the involvement of WM tracts/regions beyond the well‐recognized reading system in dyslexia. Finally, the discriminating results demonstrated a potential of WM neuroimaging features as imaging markers for identifying dyslexic individuals. Hum Brain Mapp 37:1443‐1458, 2016.

Deficiency of brain structural sub-network underlying post-ischaemic stroke apathy.

This study aimed to reveal the structural basis of post‐ischaemic stroke apathy, especially in relation to disruptions in structural connectivity.

Identification of Amnestic Mild Cognitive Impairment Using Multi-Modal Brain Features: A Combined Structural MRI and Diffusion Tensor Imaging Study

Identifying amnestic mild cognitive impairment (aMCI) is of great clinical importance because aMCI is a putative prodromal stage of Alzheimers disease. The present study aimed to explore the feasibility of accurately identifying aMCI with a magnetic resonance imaging (MRI) biomarker. We integrated measures of both gray matter (GM) abnormalities derived from structural MRI and white matter (WM) alterations acquired from diffusion tensor imaging at the voxel level across the entire brain. In particular, multi-modal brain features, including GM volume, WM fractional anisotropy, and mean diffusivity, were extracted from a relatively large sample of 64 Han Chinese aMCI patients and 64 matched controls. Then, support vector machine classifiers for GM volume, FA, and MD were fused to distinguish the aMCI patients from the controls. The fused classifier was evaluated with the leave-one-out and the 10-fold cross-validations, and the classifier had an accuracy of 83.59% and an area under the curve of 0.862. The most discriminative regions of GM were mainly located in the medial temporal lobe, temporal lobe, precuneus, cingulate gyrus, parietal lobe, and frontal lobe, whereas the most discriminative regions of WM were mainly located in the corpus callosum, cingulum, corona radiata, frontal lobe, and parietal lobe. Our findings suggest that aMCI is characterized by a distributed pattern of GM abnormalities and WM alterations that represent discriminative power and reflect relevant pathological changes in the brain, and these changes further highlight the advantage of multi-modal feature integration for identifying aMCI.

Abnormal topological organization of the white matter network in Mandarin speakers with congenital amusia

Congenital amusia is a neurogenetic disorder that mainly affects the processing of musical pitch. Brain imaging evidence indicates that it is associated with abnormal structural and functional connections in the fronto-temporal region. However, a holistic understanding of the anatomical topology underlying amusia is still lacking. Here, we used probabilistic diffusion tensor imaging tractography and graph theory to examine whole brain white matter structural connectivity in 31 Mandarin-speaking amusics and 24 age- and IQ-matched controls. Amusics showed significantly reduced global connectivity, as indicated by the abnormally decreased clustering coefficient (Cp) and increased normalized shortest path length (λ) compared to the controls. Moreover, amusics exhibited enhanced nodal strength in the right inferior parietal lobule relative to controls. The co-existence of the lexical tone deficits was associated with even more deteriorated global network efficiency in amusics, as suggested by the significant correlation between the increments in normalized shortest path length (λ) and the insensitivity in lexical tone perception. Our study is the first to reveal reduced global connectivity efficiency in amusics as well as an increase in the global connectivity cost due to the co-existed lexical tone deficits. Taken together these results provide a holistic perspective on the anatomical substrates underlying congenital amusia.

Parallel workflow tools to facilitate human brain MRI post-processing.

Multi-modal magnetic resonance imaging (MRI) techniques are widely applied in human brain studies. To obtain specific brain measures of interest from MRI datasets, a number of complex image post-processing steps are typically required. Parallel workflow tools have recently been developed, concatenating individual processing steps and enabling fully automated processing of raw MRI data to obtain the final results. These workflow tools are also designed to make optimal use of available computational resources and to support the parallel processing of different subjects or of independent processing steps for a single subject. Automated, parallel MRI post-processing tools can greatly facilitate relevant brain investigations and are being increasingly applied. In this review, we briefly summarize these parallel workflow tools and discuss relevant issues.

Network analysis reveals disrupted functional brain circuitry in drug-naive social anxiety disorder

ABSTRACT Social anxiety disorder (SAD) is a common and disabling condition characterized by excessive fear and avoidance of public scrutiny. Psychoradiology studies have suggested that the emotional and behavior deficits in SAD are associated with abnormalities in regional brain function and functional connectivity. However, little is known about whether intrinsic functional brain networks in patients with SAD are topologically disrupted. Here, we collected resting‐state fMRI data from 33 drug‐naive patients with SAD and 32 healthy controls (HC), constructed functional networks with 34 predefined regions based on previous meta‐analytic research with task‐based fMRI in SAD, and performed network‐based statistic and graph‐theory analyses. The network‐based statistic analysis revealed a single connected abnormal circuitry including the frontolimbic circuit (termed the “fear circuit”, including the dorsolateral prefrontal cortex, ventral medial prefrontal cortex and insula) and posterior cingulate/occipital areas supporting perceptual processing. In this single altered network, patients with SAD had higher functional connectivity than HC. At the global level, graph‐theory analysis revealed that the patients exhibited a lower normalized characteristic path length than HC, which suggests a disorder‐related shift of network topology toward randomized configurations. SAD‐related deficits in nodal degree, efficiency and participation coefficient were detected in the parahippocampal gyrus, posterior cingulate cortex, dorsolateral prefrontal cortex, insula and the calcarine sulcus. Aspects of abnormal connectivity were associated with anxiety symptoms. These findings highlight the aberrant topological organization of functional brain network organization in SAD, which provides insights into the neural mechanisms underlying excessive fear and avoidance of social interactions in patients with debilitating social anxiety. HighlightsWe defined 34 network nodes based on task‐based SAD fMRI meta‐analytic studies.SAD had higher functional connectivity in a single connected component.SAD had a shift of brain network topology toward randomized configurations.Abnormal connectivity in SAD was significantly associated with anxiety symptoms.

Cognitive impairment and gray/white matter volume abnormalities in pediatric patients with Turner syndrome presenting with various karyotypes

Abstract Aim: To investigate the association between cognitive impairment and gray/white matter volume abnormalities in pediatric patients with Turner syndrome (TS) presenting with various karyotypes. Methods: In the present study, 21 pediatric patients with TS and the 45,X karyotype, 24 pediatric patients with TS and other karyotypes, and 20 normal healthy controls, underwent the Wechsler intelligence test, behavioral testing, and a 3.0T magnetic resonance (MR) scan. Whole-brain high-resolution T1-weighted images were processed with SPM8 software and analyzed using voxel-based morphometry (VBM); differences in gray/white matter volume between the TS groups and healthy controls were compared using analysis of covariance. Results: Pediatric patients in both TS groups had significantly lower IQ scores compared to the normal controls (p<0.05). Furthermore, both TS groups scored significantly less than the normal controls in various composite tests of cognitive function, including verbal comprehension, perceptual reasoning, working memory, and processing speed (p<0.05). There were no significant differences between the two TS patient groups in terms of their scores for verbal comprehension, perceptual reasoning, working memory, and processing speed. However, they did display significant differences in the following tests: accuracy and reaction times in the executive control test, reaction times in the short-, middle-, and long-term attention test, and accuracy in the long-term attention test. Patients in the 45,X karyotype group displayed decreased gray matter volume in the bilateral cuneus, calcarine sulcus postcentral gyrus, right precuneus, superior parietal lobule, lingual gyrus, left precentral gyrus, and cingulate gyrus. However, gray matter volume was increased in the bilateral dorsal midbrain, orbital frontal gyrus, left insular lobe, superior temporal gyrus, inferior temporal gyrus, parahippocampal gyrus, cerebellum, posterior insular lobe, right caudate nucleus, putamen, and temporal pole. Patients with TS with other karyotypes exhibited decreased gray matter volume in the left precuneus, cingulate gyrus, right postcentral gyrus, supramarginal gyrus, angular gyrus, and cuneus; contrastingly, gray matter volume increased in both the epencephals, left caudate nucleus, superior temporal gyrus, right insular lobe, and temporal pole. All volume differences were statistically significant when compared with normal controls [familywise error (FWE)-corrected p<0.05]. With regard to the two TS groups, gray matter volume in the left hippocampus and left caudate nucleus was significantly decreased in the 45,X karyotype group compared to patients with TS with other karyotypes (FWE-corrected p<0.05); conversely, gray matter volume in the right supramarginal gyrus was increased in the 45,X karyotype group (FWE-corrected p<0.05). Conclusion: Pediatric patients with TS display a lower level of intelligence compared to healthy controls, this is complicated by verbal and non-verbal cognitive impairment. The neuropathological basis of such cognitive deficiencies may be as a result of abnormalities in gray matter development.

The effect of machine learning regression algorithms and sample size on individualized behavioral prediction with functional connectivity features

Abstract Individualized behavioral/cognitive prediction using machine learning (ML) regression approaches is becoming increasingly applied. The specific ML regression algorithm and sample size are two key factors that non‐trivially influence prediction accuracies. However, the effects of the ML regression algorithm and sample size on individualized behavioral/cognitive prediction performance have not been comprehensively assessed. To address this issue, the present study included six commonly used ML regression algorithms: ordinary least squares (OLS) regression, least absolute shrinkage and selection operator (LASSO) regression, ridge regression, elastic‐net regression, linear support vector regression (LSVR), and relevance vector regression (RVR), to perform specific behavioral/cognitive predictions based on different sample sizes. Specifically, the publicly available resting‐state functional MRI (rs‐fMRI) dataset from the Human Connectome Project (HCP) was used, and whole‐brain resting‐state functional connectivity (rsFC) or rsFC strength (rsFCS) were extracted as prediction features. Twenty‐five sample sizes (ranged from 20 to 700) were studied by sub‐sampling from the entire HCP cohort. The analyses showed that rsFC‐based LASSO regression performed remarkably worse than the other algorithms, and rsFCS‐based OLS regression performed markedly worse than the other algorithms. Regardless of the algorithm and feature type, both the prediction accuracy and its stability exponentially increased with increasing sample size. The specific patterns of the observed algorithm and sample size effects were well replicated in the prediction using re‐testing fMRI data, data processed by different imaging preprocessing schemes, and different behavioral/cognitive scores, thus indicating excellent robustness/generalization of the effects. The current findings provide critical insight into how the selected ML regression algorithm and sample size influence individualized predictions of behavior/cognition and offer important guidance for choosing the ML regression algorithm or sample size in relevant investigations. HighlightsIndividualized prediction is influenced by regression algorithm and sample size.LASSO regression performed worse than other algorithms using rsFC feature.OLS regression performed worse than other algorithms using rsFCS feature.Prediction accuracy and its stability exponentially increased with sample size.The observed algorithm and sample size effects are robust and generalizable.

The lateralized arcuate fasciculus in developmental pitch disorders among mandarin amusics: left for speech and right for music

The arcuate fasciculus (AF) is a neural fiber tract that is critical to speech and music development. Although the predominant role of the left AF in speech development is relatively clear, how the AF engages in music development is not understood. Congenital amusia is a special neurodevelopmental condition, which not only affects musical pitch but also speech tone processing. Using diffusion tensor tractography, we aimed at understanding the role of AF in music and speech processing by examining the neural connectivity characteristics of the bilateral AF among thirty Mandarin amusics. Compared to age- and intelligence quotient (IQ)-matched controls, amusics demonstrated increased connectivity as reflected by the increased fractional anisotropy in the right posterior AF but decreased connectivity as reflected by the decreased volume in the right anterior AF. Moreover, greater fractional anisotropy in the left direct AF was correlated with worse performance in speech tone perception among amusics. This study is the first to examine the neural connectivity of AF in the neurodevelopmental condition of amusia as a result of disrupted music pitch and speech tone processing. We found abnormal white matter structural connectivity in the right AF for the amusic individuals. Moreover, we demonstrated that the white matter microstructural properties of the left direct AF is modulated by lexical tone deficits among the amusic individuals. These data support the notion of distinctive pitch processing systems between music and speech.