Predicting the future of neuroimaging predictive models in mental health.

Predictive modeling using neuroimaging data has the potential to improve our understanding of the neurobiology underlying psychiatric disorders and putatively information interventions. Accordingly, there is a plethora of literature reviewing published studies, the mathematics underlying machine learning, and the best practices for using these approaches. As our knowledge of mental health and machine learning continue to evolve, we instead aim to look forward and "predict" topics that we believe will be important in current and future studies. Some of the most discussed topics in machine learning, such as bias and fairness, the handling of dirty data, and interpretable models, may be less familiar to the broader community using neuroimaging-based predictive modeling in psychiatry. In a similar vein, transdiagnostic research and targeting brain-based features for psychiatric intervention are modern topics in psychiatry that predictive models are well-suited to tackle. In this work, we target an audience who is a researcher familiar with the fundamental procedures of machine learning and who wishes to increase their knowledge of ongoing topics in the field. We aim to accelerate the utility and applications of neuroimaging-based predictive models for psychiatric research by highlighting and considering these topics. Furthermore, though not a focus, these ideas generalize to neuroimaging-based predictive modeling in other clinical neurosciences and predictive modeling with different data types (e.g., digital health data).

Atypical Intrinsic Hemispheric Interaction Associated with Autism Spectrum Disorder Is Present within the First Year of Life.

Autism spectrum disorder (ASD) is characterized by atypical connectivity lateralization of functional networks. However, previous studies have not directly investigated if differences in specialization between ASD and typically developing (TD) peers are present in infancy, leaving the timing of onset of these differences relatively unknown. We studied the hemispheric asymmetries of connectivity in children with ASD and infants later meeting the diagnostic criteria for ASD. Analyses were performed in 733 children with ASD and TD peers and in 71 infants at high risk (HR) or normal risk (NR) for ASD, with data collected at 1 month and 9 months of age. Comparing children with ASD (n = 301) to TDs (n = 432), four regions demonstrated group differences in connectivity: posterior cingulate cortex (PCC), posterior superior temporal gyrus, extrastriate cortex, and anterior prefrontal cortex. At 1 month, none of these regions exhibited group differences between ASD (n = 10), HR-nonASD (n = 15), or NR (n = 18) infants. However, by 9 months, the PCC and extrastriate exhibited atypical connectivity in ASD (n = 11) and HR-nonASD infants (n = 24) compared to NR infants (n = 22). Connectivity did not correlate with symptoms in either sample. Our results demonstrate that differences in network asymmetries associated with ASD risk are observable prior to the age of a reliable clinical diagnosis.

Large-scale differences in functional organization of left- and right-handed individuals using whole-brain, data-driven analysis of connectivity.

Handedness influences differences in lateralization of language areas as well as dominance of motor and somatosensory cortices. However, differences in whole-brain functional connectivity (i.e., functional connectomes) due to handedness have been relatively understudied beyond pre-specified networks of interest. Here, we compared functional connectomes of left- and right-handed individuals at the whole brain level. We explored differences in functional connectivity of previously established regions of interest, and showed differences between primarily left- and primarily right-handed individuals in the motor, somatosensory, and language areas using functional connectivity. We then proceeded to investigate these differences in the whole brain and found that the functional connectivity of left- and right-handed individuals are not specific to networks of interest, but extend across every region of the brain. In particular, we found that connections between and within the cerebellum show distinct patterns of connectivity. To put these effects into context, we show that the effect sizes associated with handedness differences account for a similar amount of individual differences in the connectome as sex differences. Together these results shed light on regions of the brain beyond those traditionally explored that contribute to differences in the functional organization of left- and right-handed individuals and underscore that handedness effects are neurobiologically meaningful in addition to being statistically significant.

Normalization of Speech Processing After Whole-Vault Cranioplasty in Sagittal Synostosis.

BACKGROUND: Neurocognitive studies have found impairments in language-related abilities in nonsyndromic craniosynostosis, highlighting clinical importance of early language processing. In this study, neural response to speech sounds in infants with nonsyndromic sagittal craniosynostosis (NSC) is compared, preoperatively and postoperatively, using event-related potentials (ERPs) to objectively characterize development in language processing. METHODS: Electroencephalogram was recorded while 39 infants (12 NSC and 27 controls; ages 73-283 days) listened to the Hindi dental /(Equation is included in full-text article.)a/ and retroflex /da/ phonemes (non-native phonemic discrimination task). The mismatch negativity (MMN) ERP was extracted as the peak amplitude of the largest negative deflection in the difference wave over 80 to 300 milliseconds poststimulus. Differences in MMN were analyzed using repeated measures analysis of variance. RESULTS: The MMN amplitude was attenuated in the infants with NSC preoperatively compared with controls (P = 0.047). A significant region by group interaction (P = 0.045) was observed, and infants with NSC displayed attenuated MMN in the frontal electrodes compared with controls (P = 0.010). Comparing the preoperative and postoperative MMN, a time by group interaction trend (P = 0.070) was observed. Pair-wise comparisons showed a trend for increase in MMN amplitude from preoperatively to postoperatively in the infants with NSC (P = 0.059). At the postoperative time point, infants with NSC showed no significant difference in MMN from controls (P = 0.344). CONCLUSION: Infants with NSC demonstrated atypical neural response to language preoperatively. After undergoing surgery, infants with NSC showed increased MMN amplitude which was not significantly different from controls. These findings support the idea that whole vault cranioplasty may improve neurocognitive outcomes in sagittal craniosynostosis.

Interactive social neuroscience to study autism spectrum disorder.

Individuals with autism spectrum disorder (ASD) demonstrate difficulty with social interactions and relationships, but the neural mechanisms underlying these difficulties remain largely unknown. While social difficulties in ASD are most apparent in the context of interactions with other people, most neuroscience research investigating ASD have provided limited insight into the complex dynamics of these interactions. The development of novel, innovative "interactive social neuroscience" methods to study the brain in contexts with two interacting humans is a necessary advance for ASD research. Studies applying an interactive neuroscience approach to study two brains engaging with one another have revealed significant differences in neural processes during interaction compared to observation in brain regions that are implicated in the neuropathology of ASD. Interactive social neuroscience methods are crucial in clarifying the mechanisms underlying the social and communication deficits that characterize ASD.

Machine Learning and Prediction in Fetal, Infant, and Toddler Neuroimaging: A Review and Primer.

Predictive models in neuroimaging are increasingly designed with the intent to improve risk stratification and support interventional efforts in psychiatry. Many of these models have been developed in samples of children school-aged or older. Nevertheless, despite growing evidence that altered brain maturation during the fetal, infant, and toddler (FIT) period modulates risk for poor mental health outcomes in childhood, these models are rarely implemented in FIT samples. Applications of predictive modeling in children of these ages provide an opportunity to develop powerful tools for improved characterization of the neural mechanisms underlying development. To facilitate the broader use of predictive models in FIT neuroimaging, we present a brief primer and systematic review on the methods used in current predictive modeling FIT studies. Reflecting on current practices in more than 100 studies conducted over the past decade, we provide an overview of topics, modalities, and methods commonly used in the field and under-researched areas. We then outline ethical and future considerations for neuroimaging researchers interested in predicting health outcomes in early life, including researchers who may be relatively new to either advanced machine learning methods or using FIT data. Altogether, the last decade of FIT research in machine learning has provided a foundation for accelerating the prediction of early-life trajectories across the full spectrum of illness and health.

Functional Connectome-Based Predictive Modeling in Autism.

Autism is a heterogeneous neurodevelopmental condition, and functional magnetic resonance imaging-based studies have helped advance our understanding of its effects on brain network activity. We review how predictive modeling, using measures of functional connectivity and symptoms, has helped reveal key insights into this condition. We discuss how different prediction frameworks can further our understanding of the brain-based features that underlie complex autism symptomatology and consider how predictive models may be used in clinical settings. Throughout, we highlight aspects of study interpretation, such as data decay and sampling biases, that require consideration within the context of this condition. We close by suggesting exciting future directions for predictive modeling in autism.

The N170 event-related potential reflects delayed neural response to faces when visual attention is directed to the eyes in youths with ASD.

Atypical neural response to faces is thought to contribute to social deficits in autism spectrum disorder (ASD). Compared to typically developing (TD) controls, individuals with ASD exhibit delayed brain responses to upright faces at a face-sensitive event-related potential (ERP), the N170. Given observed differences in patterns of visual attention to faces, it is not known whether slowed neural processing may simply reflect atypical looking to faces. The present study manipulated visual attention to facial features to examine whether directed attention to the eyes normalizes N170 latency in ASD. ERPs were recorded in 30 children and adolescents with ASD as well as 26 TD children and adolescents. Results replicated prior findings of shorter N170 latency to the eye region of the face in TD individuals. In contrast, those with ASD did not demonstrate modulation of N170 latency by point of regard to the face. Group differences in latency were most pronounced when attention was directed to the eyes. Results suggest that well-replicated findings of N170 delays in ASD do not simply reflect atypical patterns of visual engagement with experimental stimuli. These findings add to a body of evidence indicating that N170 delays are a promising marker of atypical neural response to social information in ASD. LAY SUMMARY: This study looks at how children's and adolescents' brains respond when looking at different parts of a face. Typically developing children and adolescents processed eyes faster than other parts of the face, whereas this pattern was not seen in ASD. Children and adolescents with ASD processed eyes more slowly than typically developing children. These findings suggest that observed inefficiencies in face processing in ASD are not simply reflective of failure to attend to the eyes.

Brief Report: Preliminary Evidence of the N170 as a Biomarker of Response to Treatment in Autism Spectrum Disorder.

Background: Autism Spectrum Disorder (ASD) is a neurodevelopmental disorder characterized by primary difficulties in social function. Individuals with ASD display slowed neural processing of faces, as indexed by the latency of the N170, a face-sensitive event-related potential. Currently, there are no objective biomarkers of ASD useful in clinical care or research. Efficacy of behavioral treatment is currently evaluated through subjective clinical impressions. To explore whether the N170 might have utility as an objective index of treatment response, we examined N170 before and after receipt of an empirically validated behavioral treatment in children with ASD. Method: Electroencephalography (EEG) data were obtained on a preliminary cohort of preschool-aged children with ASD before and after a 16-week course of PRT and in a subset of participants in waitlist control (16-weeks before the start of PRT) and follow-up (16-weeks after the end of PRT). EEG was recorded while participants viewed computer-generated faces with neutral and fearful affect. Results: Significant reductions in N170 latency to faces were observed following 16 weeks of PRT intervention. Change in N170 latency was not observed in the waitlist-control condition. Conclusions: This exploratory study offers suggestive evidence that N170 latency may index response to behavioral treatment. Future, more rigorous, studies in larger samples are indicated to evaluate whether the N170 may be useful as a biomarker of treatment response.

The Presence of Another Person Influences Oscillatory Cortical Dynamics During Dual Brain EEG Recording.

Humans are innately social creatures and the social environment strongly influences brain development. As such, the human brain is primed for and sensitive to social information even in the absence of explicit task or instruction. In this study, we examined the influence of different levels of interpersonal proximity on resting state brain activity and its association with social cognition. We measured EEG in pairs of 13 typically developing (TD) adults seated in separate rooms, in the same room back-to-back, and in the same room facing each other. Interpersonal proximity modulated broadband EEG power from 4-55 Hz and individual differences in self-reported social cognition modulated these effects in the beta and gamma frequency bands. These findings provide novel insight into the influence of social environment on brain activity and its association with social cognition through dual-brain EEG recording and demonstrate the importance of using interactive methods to study the human brain.

Oxytocin Enhances the Neural Efficiency of Social Perception.

Face perception is a highly conserved process that directs our attention from infancy and is supported by specialized neural circuitry. Oxytocin (OT) can increase accuracy and detection of emotional faces, but these effects are mediated by valence, individual differences, and context. We investigated the temporal dynamics of OT's influence on the neural substrates of face perception using event related potentials (ERPs). In a double blind, placebo controlled within-subject design, 21 healthy male adults inhaled OT or placebo and underwent ERP imaging during two face processing tasks. Experiment 1 investigated effects of OT on neural correlates of fearful vs. neutral facial expressions, and Experiment 2 manipulated point-of-gaze to neutral faces. In Experiment 1, we found that OT reduced N170 latency to fearful faces. In Experiment 2, N170 latency was decreased when participant gaze was directed to the eyes of neutral faces; however, there were no OT-associated effects in response to different facial features. Findings suggest OT modulates early stages of social perception for socially complex information such as emotional faces relative to neutral. These results are consistent with models suggesting OT impacts the salience of socially informative cues during processing, which leads to downstream effects in behavior. Future work should examine how OT affects neural processes underlying basic components of social behavior (such as, face perception) while varying emotional expression of stimuli or comparing different characteristics of participants (e.g., gender, personality traits).

Modulation of reward in a live social context as revealed through interactive social neuroscience.

Social neuroscience research investigating autism spectrum disorder (ASD) has yielded inconsistent findings, despite ASD being well-characterized by difficulties in social interaction and communication through behavioral observation. In particular, specific etiologies and functional and structural assays of the brain in autism have not been consistently identified. To date, most social neuroscience research has focused on a single person viewing static images. Research utilizing interactive social neuroscience featuring dual-brain recording offers great promise for the study of neurodevelopmental disabilities. Reward processing has been implicated in the pathology of ASD, yet mixed findings have brought uncertainty about the role reward processing deficits may play in ASD. The current study employed dual-brain EEG recording to examine reward processing during live interaction and its relation to autistic traits. Sixteen typically developing (TD) adults played a competitive treasure-hunt game against a computer and against a human partner. EEG results revealed enhanced neural sensitivity to reward outcome during live interaction with a human competitor. Further, individuals with higher levels of autistic traits demonstrated reduced sensitivity to reward outcome during live interaction. These findings provide novel insight into reward processing mechanisms associated with autistic traits, as well as support the necessary utility of interactive social neuroscience techniques to study developmental disorders.

Social Media Use, Friendship Quality, and the Moderating Role of Anxiety in Adolescents with Autism Spectrum Disorder.

Social media holds promise as a technology to facilitate social engagement, but may displace offline social activities. Adolescents with ASD are well suited to capitalize on the unique features of social media, which requires less decoding of complex social information. In this cross-sectional study, we assessed social media use, anxiety and friendship quality in 44 adolescents with ASD, and 56 clinical comparison controls. Social media use was significantly associated with high friendship quality in adolescents with ASD, which was moderated by the adolescents' anxiety levels. No associations were founds between social media use, anxiety and friendship quality in the controls. Social media may be a way for adolescents with ASD without significant anxiety to improve the quality of their friendships.