Joseph D. Viviano

A novel DTI-QA tool: Automated metric extraction exploiting the sphericity of an agar filled phantom

PURPOSE To develop a quality assurance (QA) tool (acquisition guidelines and automated processing) for diffusion tensor imaging (DTI) data using a common agar-based phantom used for fMRI QA. The goal is to produce a comprehensive set of automated, sensitive and robust QA metrics. METHODS A readily available agar phantom was scanned with and without parallel imaging reconstruction. Other scanning parameters were matched to the human scans. A central slab made up of either a thick slice or an average of a few slices, was extracted and all processing was performed on that image. The proposed QA relies on the creation of two ROIs for processing: (i) a preset central circular region of interest (ccROI) and (ii) a signal mask for all images in the dataset. The ccROI enables computation of average signal for SNR calculations as well as average FA values. The production of the signal masks enables automated measurements of eddy current and B0 inhomogeneity induced distortions by exploiting the sphericity of the phantom. Also, the signal masks allow automated background localization to assess levels of Nyquist ghosting. RESULTS The proposed DTI-QA was shown to produce eleven metrics which are robust yet sensitive to image quality changes within site and differences across sites. It can be performed in a reasonable amount of scan time (~15min) and the code for automated processing has been made publicly available. CONCLUSIONS A novel DTI-QA tool has been proposed. It has been applied successfully on data from several scanners/platforms. The novelty lies in the exploitation of the sphericity of the phantom for distortion measurements. Other novel contributions are: the computation of an SNR value per gradient direction for the diffusion weighted images (DWIs) and an SNR value per non-DWI, an automated background detection for the Nyquist ghosting measurement and an error metric reflecting the contribution of EPI instability to the eddy current induced shape changes observed for DWIs.

Integration of routine QA data into mega-analysis may improve quality and sensitivity of multisite diffusion tensor imaging studies

A novel mega‐analytical approach that reduced methodological variance was evaluated using a multisite diffusion tensor imaging (DTI) fractional anisotropy (FA) data by comparing white matter integrity in people with schizophrenia to controls. Methodological variance was reduced through regression of variance captured from quality assurance (QA) and by using Marchenko–Pastur Principal Component Analysis (MP‐PCA) denoising. N = 192 (119 patients/73 controls) data sets were collected at three sites equipped with 3T MRI systems: GE MR750, GE HDx, and Siemens Trio. DTI protocol included five b = 0 and 60 diffusion‐sensitized gradient directions (b = 1,000 s/mm2). In‐house DTI QA protocol data was acquired weekly using a uniform phantom; factor analysis was used to distil into two orthogonal QA factors related to: SNR and FA. They were used as site‐specific covariates to perform mega‐analytic data aggregation. The effect size of patient‐control differences was compared to these reported by the enhancing neuro imaging genetics meta‐analysis (ENIGMA) consortium before and after regressing QA variance. Impact of MP‐PCA filtering was evaluated likewise. QA‐factors explained ∼3–4% variance in the whole‐brain average FA values per site. Regression of QA factors improved the effect size of schizophrenia on whole brain average FA values—from Cohens d = .53 to .57—and improved the agreement between the regional pattern of FA differences observed in this study versus ENIGMA from r = .54 to .70. Application of MP‐PCA‐denoising further improved the agreement to r = .81. Regression of methodological variances captured by routine QA and advanced denoising that led to a better agreement with a large mega‐analytic study.

BDNF-Dependent Effects on Amygdala–Cortical Circuitry and Depression Risk in Children and Youth

The brain-derived neurotrophic factor (BDNF) is critical for brain development, and the functional BDNF Val66Met polymorphism is implicated in risk for mood disorders. The objective of this study was to determine how the Val66Met polymorphism influences amygdala-cortical connectivity during neurodevelopment and assess the relevance for mood disorders. Age- and sex-specific effects of the BDNF Val66Met polymorphism on amygdala-cortical connectivity were assessed by examining covariance of amygdala volumes with thickness throughout the cortex in a sample of Caucasian youths ages 8-22 that were part of the Philadelphia Neurodevelopmental Cohort (n = 339). Follow-up analyses assessed corresponding BDNF genotype effects on resting-state functional connectivity (n = 186) and the association between BDNF genotype and major depressive disorder (MDD) (n = 2749). In adolescents, amygdala-cortical covariance was significantly stronger in Met allele carriers compared with Val/Val homozygotes in amygdala-cortical networks implicated in depression; these differences were driven by females. In follow-up analyses, the Met allele was also associated with stronger resting-state functional connectivity in adolescents and increased likelihood of MDD in adolescent females. The BDNF Val66Met polymorphism may confer risk for mood disorders in females through effects on amygdala-cortical connectivity during adolescence, coinciding with a period in the lifespan when onset of depression often occurs, more commonly in females.

Neural Activity while Imitating Emotional Faces is Related to Both Lower and Higher-Level Social Cognitive Performance

Imitation and observation of actions and facial emotional expressions activates the human fronto-parietal mirror network. There is skepticism regarding the role of this low-level network in more complex high-level social behaviour. We sought to test whether neural activation during an observation/imitation task was related to both lower and higher level social cognition. We employed an established observe/imitate task of emotional faces during functional MRI in 28 healthy adults, with final analyses based on 20 individuals following extensive quality control. Partial least squares (PLS) identified patterns of relationships between spatial activation and a battery of objective out-of-scanner assessments that index lower and higher-level social cognitive performance, including the Penn emotion recognition task, reading the mind in the eyes, the awareness of social inference test (TASIT) parts 1, 2, and 3, and the relationships across domains (RAD) test. Strikingly, activity in limbic, right inferior frontal, and inferior parietal areas during imitation of emotional faces correlated with performance on emotion evaluation (TASIT1), social inference - minimal (TASIT2), social inference - enriched (TASIT3), and the RAD tests. These results show a role for this network in both lower-level and higher-level social cognitive processes which are collectively critical for social functioning in everyday life.

T155. SEPARABLE AND REPLICABLE NEURAL STRATEGIES DURING SOCIAL BRAIN FUNCTION IN PEOPLE WITH AND WITHOUT SEVERE MENTAL ILLNESS

Abstract Background The case-control design and disease heterogeneity may be major limiting factors impeding biomarker discovery in brain disorders, including serious mental illness such as schizophrenia spectrum disorder (SSD) or bipolar disorder (BPD). We propose that this heterogeneity represents an opportunity for discovery by uncovering relevant biologically driven sub-types within disorders. Individuals with schizophrenia spectrum disorder (SSD) have deficits in social cognition related to poor functional outcome. Methods A total of 109 SSD and 70 matched healthy controls (HC) were recruited across three sites. Participants performed an fMRI task in which they observed or imitated emotional faces. For each participant, an individual pattern of activity (Imitate > Observe for emotional faces) was identified. Hierarchical clustering (Ward’s method) identified clusters of individuals with similar patterns of activity. We then examined whether new data-driven groups of participants (based on patterns of brain activity) demonstrated performance differences on a batter of social and neuro cognitive tests completed out of the scanner. As a validation of the importance of cluster membership, Euclidean distance was compared between participants to members of their own cluster, diagnosis, or site. The clustering analysis was repeated on a replication sample consisting of 32 SSD, 37 euthymic BPD, and 39 HC. Results Three clusters with distinct patterns of neural activity were found. Cluster one (24 HC and 44 SSD) represented ‘typical activators’ (lateral frontal and parietal activity). Cluster two (21 HC and 31 SSD) were identified as ‘hyper-activators’, showing more intense and extended activity. This was interpreted as a ‘compensatory’ response of over-activation related to impaired neural circuits, such as is seen in aging. Interestingly, cluster three (25 Controls and 35 SSD) showed a very atypical pattern, including suppression of activity during imitation in regions involved in the default mode network and/or higher order social cognition (e.g. theory of mind). This group also had improved social cognitive performance relative to the other clusters. Participants were found to have more similar patterns of brain activity to members of their cluster rather than to members of their diagnostic group or scanning site. Importantly, when clustering was applied to the replication sample, the same three patterns (typical activators, hyper activators, and deactivators) were identified. Discussion In independently collected samples, our findings demonstrate different patterns of neural activity among individuals during a socio-emotional task that were independent of DSM-diagnosis or scan site. Our findings may provide objective neuroimaging endpoints (or biomarkers) for subgroups of individuals in target engagement research aimed at enhancing cognitive performance independent of diagnostic category.

F20. SEX-SPECIFIC STRUCTURAL AND FUNCTIONAL CIRCUIT DIFFERENCES IN YOUTH WITH PSYCHOSIS SPECTRUM SYMPTOMS

Abstract Background Functional connectivity differences in the cortico-thalamic-striatal-cortical (CTSC) circuit, as well as altered subcortical region volumes have been observed in schizophrenia. In this study, structural and functional magnetic resonance imaging (MRI) were used in a large child and youth sample aged 11–21 years (n=1134) including children with psychosis spectrum (PS) symptoms (n=312) to further understanding of these biomarkers in youth outside of high risk groups and with a wider range of symptom severity. Methods Structural subregions of the thalamus and striatum were identified using the segmentation tool MAGeT Brain. Functional subregions were segmented based on functional connectivity with the 7 functional networks identified in Yeo et al, 2011. Average time series from functional subregions were correlated vertex-wide with cortical surfaces and Fisher Z transformed. FSL’s PALM was used to examine differences and interactions between PS groups and sex. Age and in scanner motion (mean framewise displacement) were covaried for and a family wise error rate correction was applied. Structural subregion volume differences and interactions between PS groups and sex were investigated statistically using analyses of covariance (ANCOVA) with a false discovery rate of 5% correction for multiple testing. Age, intracranial volume, WRAT score and current medication use were covaried for. Results Sex-specific differences between PS and non-PS youth in structural subregion volumes were seen in both the striatum and thalamus. There was a persistent pattern of increased volumes in girls with PS symptoms, but decreased volumes in boys with PS symptoms compared to non-PS youth in the bilateral posterior putamen of the striatum (F=9.26, pFDR=0.006), higher order thalamic bilateral pulvinar (F=9.85, pFDR=0.004), left medial dorsal nuclei (F=7.42, pFDR=0.01), as well as first order thalamic left ventral posterior nucleus (F=6.47, pFDR=0.02), medial geniculate nucleus (F=10.03, pFDR=0.004) and bilateral lateral geniculate nuclei (F=5.7, pFDR=0.03). However, both PS girls and boys had increased nucleus accumbens volumes (t=2.66, pFDR=0.02). Decreased functional connectivity was found in PS youth between a striatal subregion in the right posterior putamen (corresponding to the dorsal attention network) and occipital areas (pFWE=0.005). This pattern was found to be driven by differences in specifically PS boys and not PS girls (pFWE=0.004). Discussion Multiple sex-specific structural differences between PS and non-PS youth were found in striatal and thalamic subregions. Hypo-connectivity between the striatal posterior putamen and occipital regions in PS boys overlap with structural increases in this subcortical volume in PS boys. Finding these early indicators is a key strategy to provide insight into neural mechanisms underlying the development of psychosis with the aim to improve and better target treatments.

Modeling and prediction of clinical symptom trajectories in Alzheimer’s disease using longitudinal data

Computational models predicting symptomatic progression at the individual level can be highly beneficial for early intervention and treatment planning for Alzheimer’s disease (AD). Individual prognosis is complicated by many factors including the definition of the prediction objective itself. In this work, we present a computational framework comprising machine-learning techniques for 1) modeling symptom trajectories and 2) prediction of symptom trajectories using multimodal and longitudinal data. We perform primary analyses on three cohorts from Alzheimer’s Disease Neuroimaging Initiative (ADNI), and a replication analysis using subjects from Australian Imaging, Biomarker & Lifestyle Flagship Study of Ageing (AIBL). We model the prototypical symptom trajectory classes using clinical assessment scores from mini-mental state exam (MMSE) and Alzheimer’s Disease Assessment Scale (ADAS-13) at nine timepoints spanned over six years based on a hierarchical clustering approach. Subsequently we predict these trajectory classes for a given subject using magnetic resonance (MR) imaging, genetic, and clinical variables from two timepoints (baseline + follow-up). For prediction, we present a longitudinal Siamese neural-network (LSN) with novel architectural modules for combining multimodal data from two timepoints. The trajectory modeling yields two (stable and decline) and three (stable, slow-decline, fast-decline) trajectory classes for MMSE and ADAS-13 assessments, respectively. For the predictive tasks, LSN offers highly accurate performance with 0.900 accuracy and 0.968 AUC for binary MMSE task and 0.760 accuracy for 3-way ADAS-13 task on ADNI datasets, as well as, 0.724 accuracy and 0.883 AUC for binary MMSE task on replication AIBL dataset.

Spread of activity following TMS is related to intrinsic resting connectivity to the salience network: A concurrent TMS-fMRI study

Transcranial magnetic stimulation (TMS) modulates activity at local and regions distal to the site of simulation. TMS has also been found to modulate brain networks, and it has been hypothesized that functional connectivity may predict the neuronal changes at local and distal sites in response to a TMS pulse. However, a direct relationship between resting connectivity and change in TMS-induced brain activation has yet to be demonstrated. Concurrent TMS-fMRI is a technique to directly measure this spread activity following TMS in real time. In twenty-two participants, resting-state fMRI scans were acquired, followed by four ten minute sessions of concurrent TMS-fMRI over the left dorsolateral prefrontal cortex (DLPFC). Seed-based functional connectivity to the individualized TMS target was examined using the baseline resting fMRI scan data, and the change of activity resulting from TMS was determined using a general linear model (High vs Low intensity TMS). While at the group level the spatial pattern of resting connectivity related to the pattern of TMS-induced cortical changes, there was substantial variability across individuals. This variability was further probed by examining individuals connectivity from the TMS target to six resting state networks. Only connectivity between the salience network (SN) and the TMS target site correlated with the RSC-TMS score. This suggests that resting state connectivity is correlated with TMS-induced changes in activity following DLPFC stimulation, particularly when the DLPFC target interacts with the SN. These results highlight the importance of examining such relationships at the individual level and may help to guide individual treatment in clinical populations.