Biology subjects

Fonov, V.

Publications and source records attributed to Fonov, V..

7 recordsLinked to original sources

bioRxiv2018

Unbiased Age-Appropriate Structural Brain Atlases for Chinese Pediatrics

In magnetic resonance imaging (MRI) studies of children brain development, structural brain atlases usually serve as important references of pediatric population in which individual images are spatially normalized into a common or standard stereotactic space. However, the existing popular children brain atlases (e.g., National Institutes of Health pediatric atlases, NIH-PD atlases) are made mostly based on MR images from Western populations, and are thus insufficient to characterize the brains of Chinese children due to the neuroanatomical differences that are relevant to genetic and environmental factors. By collecting high-quality T1- and T2- weighted MR images from 328 typically developing Chinese children aged from 6 to 12 years old, we created a set of age-appropriate Chinese pediatric (CHN-PD) atlases using an unbiased template construction algorithm. The CHN-PD atlases included the head/brain templates, the symmetric brain template, the gender-specific brain templates and the corresponding tissue probability atlases. Moreover, the atlases contained multiple age-specific templates with a one-year interval. A direct comparison of the CHN-PD and the NIH-PD atlases revealed remarkable anatomical differences bilaterally in the lateral frontal and parietal regions and somatosensory cortex. While applying the CHN-PD atlases to two independent Chinese pediatric datasets (N = 114 and N = 71, respectively), machine-learning regression approaches revealed higher prediction accuracy on brain ages than the usage of NIH-PD atlases. These results suggest that the CHN-PD brain atlases are necessary and important for future typical and atypical developmental studies in Chinese pediatric population. Currently, the CHN-PD atlases have been released on the NITRC website (https://www.nitrc.org/projects/chn-pd).

neuroscience↗

bioRxiv2018

Deep learning of quality control for stereotaxic registration of human brain MRI

Linear registration to stereotaxic space is a common first step in many automated image-processing tools for analysis of human brain MRI scans. This step is crucial for the success of the following image-processing steps. Several well-established algorithms are commonly used in the field of neuroimaging for this task, but none of them has a 100% success rate. Manual assessment of the registration is commonly used as part of quality control.\n\nWe propose a completely automatic quality control method based on deep learning that replaces human rater and accurately performs quality control assessment for stereotaxic registration of T1w brain scans.\n\nIn a recently published study from our group comparing linear registration methods, we used a database of 9693 MRI scans from several publically available datasets and applied five linear registration tools. In this study, the resulting images that were assessed and labeled by a human rater are used to train a deep neural network to detect cases when registration failed.\n\nOur method was able to achieve 88% accuracy and 11% false positive rate in detecting scans that should pass quality control, better than a manual QC rater.

neuroscience↗

bioRxiv2018

NeuroMTL iSEG challenge methods

We propose a tissue classification method for MRI scans of 6-month old infants, and used it to process the iSeg-2017 challenge data. The method relies on a deep-learning 3D U-Net network, trained with MRI scans of 216 infants, aged 6-24 months, from the ACE-IBIS longitudinal database.

neuroscience↗

bioRxiv2018

Comparison of different methods for average anatomical templates creation: do we really gain anything from a diffeomorphic framework?

In the field of computation anatomy, the diffeomorphic framework is widely used to perform analysis of human brain anatomy in both healthy and diseased populations. While useful for analysis, the framework imposes certain implementation constraints that do not necessarily result in improved accuracy of inter-subject co-registration in case of average anatomical template (AAT) construction - a common technique used in large population studies. In this work, we evaluated several state-of-the-art non-diffeomorphic and diffeomorphic non-linear registration frameworks in terms of their ability to build AATs. While all methods generated well behaved transforms, we found that the diffeomorphic framework does not automatically guarantee an increase of accuracy in average anatomical template construction.

neuroscience↗

bioRxiv2018

Voxel-wise T2 relaxometry of Normal Pediatric Brain Development in 326 healthy infants and toddlers.

Quantitative T2 data from an NIH-sponsored multi-center study of Normal Brain Development was used to perform automatic voxel-wise analysis of the changes in T2 evolution in the brain in healthy children within the age range from birth to 5 years. All data were non-linearly registered into a common coordinate space. The T2 parameters were estimated by 2 point fitting from the PD-weighted and T2-weighted image data, or by least-squares fitting of 4 data points when addition intermediate weighting images were available. The main result of this study is voxel-level map of monoexponential evolution of T2 in this age range indicating the delay (in months) and the rate (in 1/months) of development. The automatic maps are compared to manual region-of-interest based estimates of T2 evolution.

neuroscience↗

bioRxiv2018

BIC Defacing Algorithm

Public distribution of imaging information from several MRI data processing projects at the BIC has lead to the development of this DEFACING algorithm that is used as part of anonymisation process. Key features of the algorithm include: the defacing should modify voxels associated with face of the subject, making a rendering of the face unrecognisable, it should work on data in the native scanner space and it should not significantly affect subsequent data processing outcome (model based registration, brain extraction, brain tissue classification and brain segmentation). The algorithm is implemented using the MINC library and source code is publicly available. The effect of the defacing algorithm on the data processing was verified using the ICBM database of 152 scans of young adults.

neuroscience↗

bioRxiv2017

Network Connectivity Predicts Cortical Thinning and Cognitive Decline in Early Parkinson’s Disease

Parkinsons Disease (PD) is a progressive neurodegenerative disorder characterized by motor and cognitive deficits. The neurodegenerative process is thought to move stereotypically from the brainstem up to the cerebral cortex, possibly reflecting the spread of toxic alpha-synuclein molecules. Using a large, longitudinal, multi-center database of de novo PD patients, we tested whether focal reductions in cortical thickness could be explained by disease spread from a subcortical \"disease reservoir\" along the brains connectome. PD patients (n=105) and matched controls (n=57) underwent T1-MRI at entry and one year later. Over this period, PD patients demonstrated significantly greater loss of cortical thickness than healthy controls in parts of the left occipital and bilateral frontal lobes and right somatomotor-sensory cortex. Cortical regions with greater connectivity (measured functionally or structurally) to a \"disease reservoir\" evaluated via MRI at baseline demonstrated greater atrophy one year later. The atrophy pattern in the ventral frontal lobes resembled one described in certain cases of Alzheimers disease. Moreover, a multiple linear regression model suggested that cortical thinning was associated with impaired cognitive function at follow-up. Our findings suggest that disease propagation to the cortex in PD follows neural connectivity, and that disease spread to the cortex may herald the onset of cognitive impairment.

neuroscience↗