Transition to chaos separates learning regimes and relates to measure of consciousness in recurrent neural networks.

Recurrent neural networks exhibit chaotic dynamics when the variance in their connection strengths exceed a critical value. Recent work indicates connection variance also modulates learning strategies; networks learn "rich" representations when initialized with low coupling and "lazier" solutions with larger variance. Using Watts-Strogatz networks of varying sparsity, structure, and hidden weight variance, we find that the critical coupling strength dividing chaotic from ordered dynamics also differentiates rich and lazy learning strategies. Training moves both stable and chaotic networks closer to the edge of chaos, with networks learning richer representations before the transition to chaos. In contrast, biologically realistic connectivity structures foster stability over a wide range of variances. The transition to chaos is also reflected in a measure that clinically discriminates levels of consciousness, the perturbational complexity index (PCIst). Networks with high values of PCIst exhibit stable dynamics and rich learning, suggesting a consciousness prior may promote rich learning. The results suggest a clear relationship between critical dynamics, learning regimes and complexity-based measures of consciousness.

Modeling the cell-type-specific mesoscale murine connectome with anterograde tracing experiments.

The Allen Mouse Brain Connectivity Atlas consists of anterograde tracing experiments targeting diverse structures and classes of projecting neurons. Beyond regional anterograde tracing done in C57BL/6 wild-type mice, a large fraction of experiments are performed using transgenic Cre-lines. This allows access to cell-class-specific whole-brain connectivity information, with class defined by the transgenic lines. However, even though the number of experiments is large, it does not come close to covering all existing cell classes in every area where they exist. Here, we study how much we can fill in these gaps and estimate the cell-class-specific connectivity function given the simplifying assumptions that nearby voxels have smoothly varying projections, but that these projection tensors can change sharply depending on the region and class of the projecting cells. This paper describes the conversion of Cre-line tracer experiments into class-specific connectivity matrices representing the connection strengths between source and target structures. We introduce and validate a novel statistical model for creation of connectivity matrices. We extend the Nadaraya-Watson kernel learning method that we previously used to fill in spatial gaps to also fill in gaps in cell-class connectivity information. To do this, we construct a "cell-class space" based on class-specific averaged regionalized projections and combine smoothing in 3D space as well as in this abstract space to share information between similar neuron classes. Using this method, we construct a set of connectivity matrices using multiple levels of resolution at which discontinuities in connectivity are assumed. We show that the connectivities obtained from this model display expected cell-type- and structure-specific connectivities. We also show that the wild-type connectivity matrix can be factored using a sparse set of factors, and analyze the informativeness of this latent variable model.

Heterogeneity in Neuronal Dynamics Is Learned by Gradient Descent for Temporal Processing Tasks.

Individual neurons in the brain have complex intrinsic dynamics that are highly diverse. We hypothesize that the complex dynamics produced by networks of complex and heterogeneous neurons may contribute to the brain's ability to process and respond to temporally complex data. To study the role of complex and heterogeneous neuronal dynamics in network computation, we develop a rate-based neuronal model, the generalized-leaky-integrate-and-fire-rate (GLIFR) model, which is a rate equivalent of the generalized-leaky-integrate-and-fire model. The GLIFR model has multiple dynamical mechanisms, which add to the complexity of its activity while maintaining differentiability. We focus on the role of after-spike currents, currents induced or modulated by neuronal spikes, in producing rich temporal dynamics. We use machine learning techniques to learn both synaptic weights and parameters underlying intrinsic dynamics to solve temporal tasks. The GLIFR model allows the use of standard gradient descent techniques rather than surrogate gradient descent, which has been used in spiking neural networks. After establishing the ability to optimize parameters using gradient descent in single neurons, we ask how networks of GLIFR neurons learn and perform on temporally challenging tasks, such as sequential MNIST. We find that these networks learn diverse parameters, which gives rise to diversity in neuronal dynamics, as demonstrated by clustering of neuronal parameters. GLIFR networks have mixed performance when compared to vanilla recurrent neural networks, with higher performance in pixel-by-pixel MNIST but lower in line-by-line MNIST. However, they appear to be more robust to random silencing. We find that the ability to learn heterogeneity and the presence of after-spike currents contribute to these gains in performance. Our work demonstrates both the computational robustness of neuronal complexity and diversity in networks and a feasible method of training such models using exact gradients.

Correlates and predictors of the severity of suicidal ideation in adolescence: an examination of brain connectomics and psychosocial characteristics.

BACKGROUND: Suicidal ideation (SI) typically emerges during adolescence but is challenging to predict. Given the potentially lethal consequences of SI, it is important to identify neurobiological and psychosocial variables explaining the severity of SI in adolescents. METHODS: In 106 participants (59 female) recruited from the community, we assessed psychosocial characteristics and obtained resting-state fMRI data in early adolescence (baseline: aged 9-13 years). Across 250 brain regions, we assessed local graph theory-based properties of interconnectedness: local efficiency, eigenvector centrality, nodal degree, within-module z-score, and participation coefficient. Four years later (follow-up: ages 13-19 years), participants self-reported their SI severity. We used least absolute shrinkage and selection operator (LASSO) regressions to identify a linear combination of psychosocial and brain-based variables that best explain the severity of SI symptoms at follow-up. Nested-cross-validation yielded model performance statistics for all LASSO models. RESULTS: A combination of psychosocial and brain-based variables explained subsequent severity of SI (R2 = .55); the strongest was internalizing and externalizing symptom severity at follow-up. Follow-up LASSO regressions of psychosocial-only and brain-based-only variables indicated that psychosocial-only variables explained 55% of the variance in SI severity; in contrast, brain-based-only variables performed worse than the null model. CONCLUSIONS: A linear combination of baseline and follow-up psychosocial variables best explained the severity of SI. Follow-up analyses indicated that graph theory resting-state metrics did not increase the prediction of the severity of SI in adolescents. Attending to internalizing and externalizing symptoms is important in early adolescence; resting-state connectivity properties other than local graph theory metrics might yield a stronger prediction of the severity of SI.

White Matter Microstructural Properties of the Cerebellar Peduncles Predict Change in Symptoms of Psychopathology in Adolescent Girls.

Internalizing symptoms typically emerge in adolescence and are more prevalent in females than in males; in contrast, externalizing symptoms typically emerge in childhood and are more commonly observed in males. Previous research has implicated aspects of white matter organization, including fractional anisotropy (FA), of cerebral tracts in both internalizing and externalizing symptoms. Although the cerebellum has been posited to integrate limbic and cortical regions, its role in psychopathology is not well understood. In this longitudinal study, we investigated whether FA of the superior (SCP), middle (MCP), and inferior cerebellar peduncles (ICP) predict change in symptoms and whether sex moderates this association. One hundred eleven adolescents completed the Youth Self-Report, assessing symptoms at baseline (ages 9-13 years) and again 2 years later. Participants also underwent diffusion-weighted imaging at baseline. We used deterministic tractography to segment and compute mean FA of the cerebellar peduncles. Lower FA of the right SCP at baseline predicted increases in internalizing symptoms in females only. Lower FA in the right SCP and ICP also predicted increases in externalizing symptoms in females, but these associations did not survive multiple comparison correction. There was no association between FA of the cerebellar peduncles and change in symptoms in males or between MCP FA and symptom changes in males or females. Organizational properties of the SCP may be a sex-specific marker of internalizing symptom changes in adolescence. The cerebellar peduncles should be explored further in future studies to elucidate sex differences in symptoms.

Searching for Imaging Biomarkers of Psychotic Dysconnectivity.

BACKGROUND: Progress in precision psychiatry is predicated on identifying reliable individual-level diagnostic biomarkers. For psychosis, measures of structural and functional connectivity could be promising biomarkers given consistent reports of dysconnectivity across psychotic disorders using magnetic resonance imaging. METHODS: We leveraged data from four independent cohorts of patients with psychosis and control subjects with observations from approximately 800 individuals. We used group-level analyses and two supervised machine learning algorithms (support vector machines and ridge regression) to test within-, between-, and across-sample classification performance of white matter and resting-state connectivity metrics. RESULTS: Although we replicated group-level differences in brain connectivity, individual-level classification was suboptimal. Classification performance within samples was variable across folds (highest area under the curve [AUC] range = 0.30) and across datasets (average support vector machine AUC range = 0.50; average ridge regression AUC range = 0.18). Classification performance between samples was similarly variable or resulted in AUC values of approximately 0.65, indicating a lack of model generalizability. Furthermore, collapsing across samples (resting-state functional magnetic resonance imaging, N = 888; diffusion tensor imaging, N = 860) did not improve model performance (maximal AUC = 0.67). Ridge regression models generally outperformed support vector machine models, although classification performance was still suboptimal in terms of clinical relevance. Adjusting for demographic covariates did not greatly affect results. CONCLUSIONS: Connectivity measures were not suitable as diagnostic biomarkers for psychosis as assessed in this study. Our results do not negate that other approaches may be more successful, although it is clear that a systematic approach to individual-level classification with large independent validation samples is necessary to properly vet neuroimaging features as diagnostic biomarkers.

Greater age-related changes in white matter morphometry following early life stress: Associations with internalizing problems in adolescence.

Early life stress (ELS) is associated with increased risk for internalizing disorders and variations in gray matter development. It is unclear, however, whether ELS affects normative age-related changes in white matter (WM) morphology, and if such maturational differences are associated with risk for internalizing psychopathology. We conducted comprehensive interviews in a cross-sectional sample of young adolescents (N = 156; 89 F; Ages 9-14) to assess lifetime exposure to stress and objective cumulative ELS severity. We used diffusion-weighted imaging to measure WM fixel-based morphometry and tested the effects of age and ELS on WM fiber density and cross-section (FDC), and associations between WM FDC and internalizing problems. Age was positively associated with FDC in all WM tracts; greater ELS severity was related to stronger age-WM associations in several association tracts connecting the frontal lobes with limbic, parietal, and occipital regions, including bilateral superior and inferior longitudinal and uncinate fasciculi (UF). Among older adolescents with greater ELS severity, a higher UF FDC was associated with fewer internalizing problems. Greater ELS severity predicted more mature WM morphometry in tracts implicated in emotion regulation and cognitive processing. More phenotypically mature UF WM may be adaptive against internalizing psychopathology in adolescents exposed to ELS.

Differences in atypical resting-state effective connectivity distinguish autism from schizophrenia.

Autism and schizophrenia share overlapping genetic etiology, common changes in brain structure and common cognitive deficits. A number of studies using resting state fMRI have shown that machine learning algorithms can distinguish between healthy controls and individuals diagnosed with either autism spectrum disorder or schizophrenia. However, it has not yet been determined whether machine learning algorithms can be used to distinguish between the two disorders. Using a linear support vector machine, we identify features that are most diagnostic for each disorder and successfully use them to classify an independent cohort of subjects. We find both common and divergent connectivity differences largely in the default mode network as well as in salience, and motor networks. Using divergent connectivity differences, we are able to distinguish autistic subjects from those with schizophrenia. Understanding the common and divergent connectivity changes associated with these disorders may provide a framework for understanding their shared cognitive deficits.