RESUMEN
Neuroimaging research has uncovered a multitude of neural abnormalities associated with psychopathology, but few prediction-based studies have been conducted during adolescence, and even fewer used neurobiological features that were extracted across multiple neuroimaging modalities. This gap in the literature is critical, as deriving accurate brain-based models of psychopathology is an essential step towards understanding key neural mechanisms and identifying high-risk individuals. As such, we trained adaptive tree-boosting algorithms on multimodal neuroimaging features from the Lifespan Human Connectome Developmental (HCP-D) sample that contained 956 participants between the ages of 8 to 22 years old. Our feature space consisted of 1037 anatomical, 1090 functional, and 192 diffusion MRI features, which were used to derive models that separately predicted internalizing symptoms, externalizing symptoms, and the general psychopathology factor. We found that multimodal models were the most accurate, but all brain-based models of psychopathology yielded out-of-sample predictions that were weakly correlated with actual symptoms (r2 < 0.15). White matter microstructural properties, including orientation dispersion indices and intracellular volume fractions, were the most predictive of general psychopathology, followed by cortical thickness and functional connectivity. Spatially, the most predictive features of general psychopathology were primarily localized within the default mode and dorsal attention networks. These results were mostly consistent across all dimensions of psychopathology, except orientation dispersion indices and the default mode network were not as heavily weighted in the prediction of internalizing and externalizing symptoms. Taken with prior literature, it appears that neurobiological features are an important part of the equation for predicting psychopathology but relying exclusively on neural markers is clearly not sufficient, especially among adolescent samples with subclinical symptoms. Consequently, risk factor models of psychopathology may benefit from incorporating additional sources of information that have also been shown to explain individual differences, such as psychosocial factors, environmental stressors, and genetic vulnerabilities.
RESUMEN
BACKGROUND: Fetal exposure to maternal mood dysregulation influences child cognitive and emotional development, which may have long-lasting implications for mental health. However, the neurobiological alterations associated with this dimension of adversity have yet to be explored. Here, we tested the hypothesis that fetal exposure to entropy, a novel index of dysregulated maternal mood, would predict the integrity of the salience network, which is involved in emotional processing. METHODS: A sample of 138 child-mother pairs (70 females) participated in this prospective longitudinal study. Maternal negative mood level and entropy (an index of variable and unpredictable mood) were assessed 5 times during pregnancy. Adolescents engaged in a functional magnetic resonance imaging task that was acquired between 2 resting-state scans. Changes in network integrity were analyzed using mixed-effect and latent growth curve models. The amplitude of low frequency fluctuations was analyzed to corroborate findings. RESULTS: Prenatal maternal mood entropy, but not mood level, was associated with salience network integrity. Both prenatal negative mood level and entropy were associated with the amplitude of low frequency fluctuations of the salience network. Latent class analysis yielded 2 profiles based on changes in network integrity across all functional magnetic resonance imaging sequences. The profile that exhibited little variation in network connectivity (i.e., inflexibility) consisted of adolescents who were exposed to higher negative maternal mood levels and more entropy. CONCLUSIONS: These findings suggest that fetal exposure to maternal mood dysregulation is associated with a weakened and inflexible salience network. More broadly, they identify maternal mood entropy as a novel marker of early adversity that exhibits long-lasting associations with offspring brain development.
Asunto(s)
Efectos Tardíos de la Exposición Prenatal , Humanos , Adolescente , Embarazo , Femenino , Estudios Longitudinales , Entropía , Estudios Prospectivos , Encéfalo/fisiologíaRESUMEN
Brain aging is a complex, multifaceted process that can be challenging to model in ways that are accurate and clinically useful. One of the most common approaches has been to apply machine learning to neuroimaging data with the goal of predicting age in a data-driven manner. Building on initial brain age studies that were derived solely from T1-weighted scans (i.e., unimodal), recent studies have incorporated features across multiple imaging modalities (i.e., "multimodal"). In this systematic review, we show that unimodal and multimodal models have distinct advantages. Multimodal models are the most accurate and sensitive to differences in chronic brain disorders. In contrast, unimodal models from functional magnetic resonance imaging were most sensitive to differences across a broad array of phenotypes. Altogether, multimodal imaging has provided us valuable insight for improving the accuracy of brain age models, but there is still much untapped potential with regard to achieving widespread clinical utility.
RESUMEN
Machine learning has been increasingly applied to neuroimaging data to predict age, deriving a personalized biomarker with potential clinical applications. The scientific and clinical value of these models depends on their applicability to independently acquired scans from diverse sources. Accordingly, we evaluated the generalizability of two brain age models that were trained across the lifespan by applying them to three distinct early-life samples with participants aged 8-22 years. These models were chosen based on the size and diversity of their training data, but they also differed greatly in their processing methods and predictive algorithms. Specifically, one brain age model was built by applying gradient tree boosting (GTB) to extracted features of cortical thickness, surface area, and brain volume. The other model applied a 2D convolutional neural network (DBN) to minimally preprocessed slices of T1-weighted scans. Additional model variants were created to understand how generalizability changed when each model was trained with data that became more similar to the test samples in terms of age and acquisition protocols. Our results illustrated numerous trade-offs. The GTB predictions were relatively more accurate overall and yielded more reliable predictions when applied to lower quality scans. In contrast, the DBN displayed the most utility in detecting associations between brain age gaps and cognitive functioning. Broadly speaking, the largest limitations affecting generalizability were acquisition protocol differences and biased brain age estimates. If such confounds could eventually be removed without post-hoc corrections, brain age predictions may have greater utility as personalized biomarkers of healthy aging.
Asunto(s)
Benchmarking , Imagen por Resonancia Magnética , Humanos , Imagen por Resonancia Magnética/métodos , Encéfalo/diagnóstico por imagen , Neuroimagen/métodos , LongevidadRESUMEN
BACKGROUND: Early-life experiences have profound effects on functioning in adulthood. Altered cortical development may be one mechanism through which early-life experiences, including poverty and psychopathology symptoms, affect outcomes. However, there is little prospective research beginning early in development that combines clinician-rated psychopathology symptoms and multiwave magnetic resonance imaging to examine when these relationships emerge. METHODS: Children from the Preschool Depression Study who completed diagnostic interviews at three different developmental stages (preschool, school age, early adolescent) and up to three magnetic resonance imaging scans beginning in middle childhood participated in this study (N = 138). Multilevel models were used to calculate intercepts and slopes of cortical thickness within a priori cortical regions of interest. Linear regressions probed how early-life poverty and psychopathology (depression, anxiety, and externalizing symptoms at separate developmental periods) related to intercept/slope. RESULTS: Collectively, experiences during the preschool period predicted reduced cortical thickness, via either reduced intercept or accelerated thinning (slope). Early-life poverty predicted intercepts within sensory and sensory-motor integration regions. Beyond poverty, preschool anxiety symptoms predicted intercepts within the insula, subgenual cingulate, and inferior parietal cortex. Preschool externalizing symptoms predicted accelerated thinning within prefrontal and parietal cortices. Depression and anxiety/externalizing symptoms at later ages were not significant predictors. CONCLUSIONS: Early childhood is a critical period of risk; experiences at this developmental stage specifically have the potential for prolonged influence on brain development. Negative early experiences collectively predicted reduced cortical thickness, but the specific neural systems affected aligned with those typically implicated in these individual disorders/experiences.
Asunto(s)
Trastornos de Ansiedad , Psicopatología , Adolescente , Adulto , Niño , Preescolar , Humanos , Estudios Longitudinales , Imagen por Resonancia Magnética/métodos , Estudios ProspectivosRESUMEN
Irritability is an important dimension of psychopathology that spans multiple clinical diagnostic categories, yet its relationship to patterns of brain development remains sparsely explored. Here, we examined how transdiagnostic symptoms of irritability relate to the development of structural brain networks. All participants (n = 137, 83 females) completed structural brain imaging with 3 Tesla MRI at two timepoints (mean age at follow-up: 21.1 years, mean inter-scan interval: 5.2 years). Irritability at follow-up was assessed using the Affective Reactivity Index, and cortical thickness was quantified using Advanced Normalization Tools software. Structural covariance networks were delineated using non-negative matrix factorization, a multivariate analysis technique. Both cross-sectional and longitudinal associations with irritability at follow-up were evaluated using generalized additive models with penalized splines. The False Discovery Rate (q < 0.05) was used to correct for multiple comparisons. Cross-sectional analysis of follow-up data revealed that 11 of the 24 covariance networks were associated with irritability, with higher levels of irritability being associated with thinner cortex. Longitudinal analyses further revealed that accelerated cortical thinning within nine networks was related to irritability at follow-up. Effects were particularly prominent in brain regions implicated in emotion regulation, including the orbitofrontal, lateral temporal, and medial temporal cortex. Collectively, these findings suggest that irritability is associated with widespread reductions in cortical thickness and accelerated cortical thinning, particularly within the frontal and temporal cortex. Aberrant structural maturation of regions important for emotional regulation may in part underlie symptoms of irritability.
Asunto(s)
Encéfalo/anatomía & histología , Encéfalo/crecimiento & desarrollo , Genio Irritable/fisiología , Adolescente , Adulto , Corteza Cerebral/anatomía & histología , Corteza Cerebral/crecimiento & desarrollo , Niño , Estudios Transversales , Femenino , Humanos , Estudios Longitudinales , Imagen por Resonancia Magnética , Masculino , Vías Nerviosas/anatomía & histología , Vías Nerviosas/crecimiento & desarrollo , Adulto JovenRESUMEN
Cognitive deficits are a core feature of schizophrenia, but the neural mechanisms that contribute to these characteristics are not fully understood. This study investigated whether volume of the dorsal lateral prefrontal cortex (DLPFC), inferior frontal gyrus (IFG), hippocampus, and white matter were associated with impairment in specific cognitive domains, including executive functioning, working memory, verbal memory, verbal fluency, processing speed, versus global functioning. The multi-site data used in this study was collected from the Bipolar and Schizophrenia Network on Intermediate Phenotypes (B-SNIP), and consisted of 206 healthy controls and 247 individuals with either schizophrenia or schizoaffective disorder. The neuroimaging data was segmented based on the Destrieux atlas in FreeSurfer. Linear regression analyses revealed that global cognition, executive functioning, working memory, and processing speed were associated with all brain structures, except the DLPFC was only associated with executive fucntion. When controlling for the global cognitive deficit, executive function was trending significance with white matter, but continued to be associated with the DLPFC and IFG, as did the association between processing speed and the hippocampus. These findings suggest that volumes of the DLPFC, IFG, hippocampus, and white matter are associated with the global cognitive impairment seen in schizophrenia, but some brain structures may also be specifically related to domain-specific deficits (primarily executive function) over-and-beyond the global cognitive deficit.