Ani Eloyan

Automated diagnoses of attention deficit hyperactive disorder using magnetic resonance imaging

Successful automated diagnoses of attention deficit hyperactive disorder (ADHD) using imaging and functional biomarkers would have fundamental consequences on the public health impact of the disease. In this work, we show results on the predictability of ADHD using imaging biomarkers and discuss the scientific and diagnostic impacts of the research. We created a prediction model using the landmark ADHD 200 data set focusing on resting state functional connectivity (rs-fc) and structural brain imaging. We predicted ADHD status and subtype, obtained by behavioral examination, using imaging data, intelligence quotients and other covariates. The novel contributions of this manuscript include a thorough exploration of prediction and image feature extraction methodology on this form of data, including the use of singular value decompositions (SVDs), CUR decompositions, random forest, gradient boosting, bagging, voxel-based morphometry, and support vector machines as well as important insights into the value, and potentially lack thereof, of imaging biomarkers of disease. The key results include the CUR-based decomposition of the rs-fc-fMRI along with gradient boosting and the prediction algorithm based on a motor network parcellation and random forest algorithm. We conjecture that the CUR decomposition is largely diagnosing common population directions of head motion. Of note, a byproduct of this research is a potential automated method for detecting subtle in-scanner motion. The final prediction algorithm, a weighted combination of several algorithms, had an external test set specificity of 94% with sensitivity of 21%. The most promising imaging biomarker was a correlation graph from a motor network parcellation. In summary, we have undertaken a large-scale statistical exploratory prediction exercise on the unique ADHD 200 data set. The exercise produced several potential leads for future scientific exploration of the neurological basis of ADHD.

Electrocorticographic functional mapping identifies human cortex critical for auditory and visual naming.

More comprehensive, and efficient, mapping strategies are needed to avoid post-operative language impairments in patients undergoing epilepsy surgery. Conservative resection of dominant anterior frontal or temporal cortex frequently results in post-operative naming deficits despite standard pre-operative electrocortical stimulation mapping of visual object (picture) naming. Naming to auditory description may better simulate word retrieval in human conversation but is not typically tested, in part due to the time demands of electrocortical stimulation mapping. Electrocorticographic high gamma (60-150 Hz) activity, recorded simultaneously through the same electrodes used for stimulation mapping, has recently been used to map brain function more efficiently, and has at times predicted deficits not anticipated based on stimulation mapping alone. The present study investigated electrocorticographic mapping of visual object naming and auditory descriptive naming within conservative dominant temporal or frontal lobe resection boundaries in 16 patients with 933 subdural electrodes implanted for epilepsy surgery planning. A logistic regression model showed that electrodes within traditional conservative dominant frontal or temporal lobe resection boundaries were significantly more likely to record high gamma activity during auditory descriptive naming than during visual object naming. Eleven patients ultimately underwent resection and 7 demonstrated post-operative language deficits not anticipated based on electrocortical stimulation mapping alone. Four patients with post-operative deficits underwent a resection that included sites where high gamma activity was observed during naming. These findings indicate that electrocorticographic mapping of auditory descriptive naming may reduce the risk of permanent post-operative language deficits following dominant temporal or frontal resection.

Intrinsic Visual-Motor Synchrony Correlates With Social Deficits in Autism.

BACKGROUND Imitation, which is impaired in children with autism spectrum disorder (ASD) and critically depends on the integration of visual input with motor output, likely impacts both motor and social skill acquisition in children with ASD; however, it is unclear what brain mechanisms contribute to this impairment. Children with ASD also exhibit what appears to be an ASD-specific bias against using visual feedback during motor learning. Does the temporal congruity of intrinsic activity, or functional connectivity, between motor and visual brain regions contribute to ASD-associated deficits in imitation, motor, and social skills? METHODS We acquired resting-state functional magnetic resonance imaging scans from 100 8- to 12-year-old children (50 ASD). Group independent component analysis was used to estimate functional connectivity between visual and motor systems. Brain-behavior relationships were assessed by regressing functional connectivity measures with social deficit severity, imitation, and gesture performance scores. RESULTS We observed increased intrinsic asynchrony between visual and motor systems in children with ASD and replicated this finding in an independent sample from the Autism Brain Imaging Data Exchange. Moreover, children with more out-of-sync intrinsic visual-motor activity displayed more severe autistic traits, while children with greater intrinsic visual-motor synchrony were better imitators. CONCLUSIONS Our twice replicated findings confirm that visual-motor functional connectivity is disrupted in ASD. Furthermore, the observed temporal incongruity between visual and motor systems, which may reflect diminished integration of visual consequences with motor output, was predictive of the severity of social deficits and may contribute to impaired social-communicative skill development in children with ASD.

Precentral gyrus functional connectivity signatures of autism

Motor impairments are prevalent in children with autism spectrum disorders (ASD) and are perhaps the earliest symptoms to develop. In addition, motor skills relate to the communicative/social deficits at the core of ASD diagnosis, and these behavioral deficits may reflect abnormal connectivity within brain networks underlying motor control and learning. Despite the fact that motor abnormalities in ASD are well-characterized, there remains a fundamental disconnect between the complexity of the clinical presentation of ASD and the underlying neurobiological mechanisms. In this study, we examined connectivity within and between functional subregions of a key component of the motor control network, the precentral gyrus, using resting state functional Magnetic Resonance Imaging data collected from a large, heterogeneous sample of individuals with ASD as well as neurotypical controls. We found that the strength of connectivity within and between distinct functional subregions of the precentral gyrus was related to ASD diagnosis and to the severity of ASD traits. In particular, connectivity involving the dorsomedial (lower limb/trunk) subregion was abnormal in ASD individuals as predicted by models using a dichotomous variable coding for the presence of ASD, as well as models using symptom severity ratings. These findings provide further support for a link between motor and social/communicative abilities in ASD.

Shrinkage prediction of seed-voxel brain connectivity using resting state fMRI.

Resting-state functional magnetic resonance imaging (rs-fMRI) is used to investigate synchronous activations in spatially distinct regions of the brain, which are thought to reflect functional systems supporting cognitive processes. Analyses are often performed using seed-based correlation analysis, allowing researchers to explore functional connectivity between data in a seed region and the rest of the brain. Using scan-rescan rs-fMRI data, we investigate how well the subject-specific seed-based correlation map from the second replication of the study can be predicted using data from the first replication. We show that one can dramatically improve prediction of subject-specific connectivity by borrowing strength from the group correlation map computed using all other subjects in the study. Even more surprisingly, we found that the group correlation map provided a better prediction of a subjects connectivity than the individuals own data. While further discussion and experimentation are required to understand how this can be used in practice, results indicate that shrinkage-based methods that borrow strength from the population mean should play a role in rs-fMRI data analysis.

The impact of the modified Atkins diet on lipid profiles in adults with epilepsy

Objectives: The modified Atkins diet (MAD) is a high fat, low carbohydrate ketogenic diet used to treat intractable seizures in children and adults. The long-term impact on fasting lipid profiles (FLPs) remains unknown. This study was designed to detect significant lipid changes in adults on MAD. Methods: Patients were observed prospectively. A FLP was obtained in all patients at the first visit then serially. Patients were started on a 20 g per day net carbohydrate limit MAD. They were screened for risk for coronary heart disease and counseled to reduce saturated fats by a registered dietitian if deemed at risk. Patients that remained on MAD for 3 or more months with one or more follow-up FLP were included. Results: Thirty-seven patients (14 male), mean age 33 years (SD 13, range 18–59) met study criteria. Median diet duration was 16 months (range 3–41). Total cholesterol and low-density lipoprotein (LDL) increased significantly over the first 3 months of MAD (P = 0.01 and 0.008, respectively), but were not significantly different from baseline after 1 year of treatment (P = 0.2 and P = 0.5, respectively). High-density lipoprotein levels trended upward in the first 3 months (P = 0.05) and triglycerides remained unchanged (P = 0.5). In 12 patients followed for 3 or more years, no cardiovascular or cerebrovascular events were reported. Discussion: Although total cholesterol and LDL increased over the first 3 months of the MAD, these values normalized within a year of treatment, including in patients treated with MAD for more than 3 years.

Longitudinal relationships among posturography and gait measures in multiple sclerosis.

Objective: Gait and balance dysfunction frequently occurs early in the multiple sclerosis (MS) disease course. Hence, we sought to determine the longitudinal relationships among quantitative measures of gait and balance in individuals with MS. Methods: Fifty-seven ambulatory individuals with MS (28 relapsing-remitting, 29 progressive) were evaluated using posturography, quantitative sensorimotor and gait measures, and overall MS disability with the Expanded Disability Status Scale at each session. Results: Our cohorts age was 45.8 ± 10.4 years (mean ± SD), follow-up time 32.8 ± 15.4 months, median Expanded Disability Status Scale score 3.5, and 56% were women. Poorer performance on balance measures was related to slower walking velocity. Two posturography measures, the anterior-posterior sway and sway during static eyes open, feet apart conditions, were significant contributors to walk velocity over time (approximate R2 = 0.95), such that poorer performance on the posturography measures was related to slower walking velocity. Similarly, the anterior-posterior sway and sway during static eyes closed, feet together conditions were also significant contributors to the Timed 25-Foot Walk performance over time (approximate R2 = 0.83). Conclusions: This longitudinal cohort study establishes a strong relationship between clinical gait measures and posturography. The data show that increases in static posturography and reductions in dynamic posturography are associated with a decline in walk velocity and Timed 25-Foot Walk performance over time. Furthermore, longitudinal balance measures predict future walking performance. Quantitative walking and balance measures are important additions to clinical testing to explore longitudinal change and understand fall risk in this progressive disease population.

Health effects of lesion localization in multiple sclerosis: spatial registration and confounding adjustment.

Brain lesion localization in multiple sclerosis (MS) is thought to be associated with the type and severity of adverse health effects. However, several factors hinder statistical analyses of such associations using large MRI datasets: 1) spatial registration algorithms developed for healthy individuals may be less effective on diseased brains and lead to different spatial distributions of lesions; 2) interpretation of results requires the careful selection of confounders; and 3) most approaches have focused on voxel-wise regression approaches. In this paper, we evaluated the performance of five registration algorithms and observed that conclusions regarding lesion localization can vary substantially with the choice of registration algorithm. Methods for dealing with confounding factors due to differences in disease duration and local lesion volume are introduced. Voxel-wise regression is then extended by the introduction of a metric that measures the distance between a patient-specific lesion mask and the population prevalence map.

LIKELIHOOD BASED POPULATION INDEPENDENT COMPONENT ANALYSIS

Independent component analysis (ICA) is a widely used technique for blind source separation, used heavily in several scientific research areas including acoustics, electrophysiology, and functional neuroimaging. We propose a scalable two-stage iterative true group ICA methodology for analyzing population level functional magnetic resonance imaging (fMRI) data where the number of subjects is very large. The method is based on likelihood estimators of the underlying source densities and the mixing matrix. As opposed to many commonly used group ICA algorithms, the proposed method does not require significant data reduction by a 2-fold singular value decomposition. In addition, the method can be applied to a large group of subjects since the memory requirements are not restrictive. The performance of our approach is compared with a commonly used group ICA algorithm via simulation studies. Furthermore, the proposed method is applied to a large collection of resting state fMRI datasets. The results show that established brain networks are well recovered by the proposed algorithm.

Analysis of group ICA-based connectivity measures from fMRI: application to Alzheimer's disease.

Functional magnetic resonance imaging (fMRI) is a powerful tool for the in vivo study of the pathophysiology of brain disorders and disease. In this manuscript, we propose an analysis stream for fMRI functional connectivity data and apply it to a novel study of Alzheimers disease. In the first stage, spatial independent component analysis is applied to group fMRI data to obtain common brain networks (spatial maps) and subject-specific mixing matrices (time courses). In the second stage, functional principal component analysis is utilized to decompose the mixing matrices into population-level eigenvectors and subject-specific loadings. Inference is performed using permutation-based exact logistic regression for matched pairs data. The method is applied to a novel fMRI study of Alzheimers disease risk under a verbal paired associates task. We found empirical evidence of alternative ICA-based metrics of connectivity when comparing subjects evidencing mild cognitive impairment relative to carefully matched controls.