Feeding the machine: Challenges to reproducible predictive modeling in resting-state connectomics

Translational Connectomics: overview of machine learning in macroscale Connectomics for clinical insights

Connectomics is a neuroscience paradigm focused on noninvasively mapping highly intricate and organized networks of neurons. The advent of neuroimaging has led to extensive mapping of the brain functional and structural connectome on a macroscale level through modalities such as functional and diffusion MRI. In parallel, the healthcare field has witnessed a surge in the application of machine learning and artificial intelligence for diagnostics, especially in imaging. While reviews covering machine learn ing and macroscale connectomics exist for specific disorders, none provide an overview that captures their evolving role, especially through the lens of clinical application and translation. The applications include understanding disorders, classification, identifying neuroimaging biomarkers, assessing severity, predicting outcomes and intervention response, identifying potential targets for brain stimulation, and evaluating the effects of stimulation intervention on the brain and connectome mapping in patients before neurosurgery. The covered studies span neurodegenerative, neurodevelopmental, neuropsychiatric, and neurological disorders. Along with applications, the review provides a brief of common ML methods to set context. Conjointly, limitations in ML studies within connectomics and strategies to mitigate them have been covered.

A robust brain network for sustained attention from adolescence to adulthood that predicts later substance use

Substance use, including cigarettes and cannabis, is associated with poorer sustained attention in late adolescence and early adulthood. Previous studies were predominantly cross-sectional or under-powered and could not indicate if impairment in sustained attention was a predictor of substance use or a marker of the inclination to engage in such behavior. This study explored the relationship between sustained attention and substance use across a longitudinal span from ages 14 to 23 in over 1000 participants. Behaviors and brain connectivity associated with diminished sustained attention at age 14 predicted subsequent increases in cannabis and cigarette smoking, establishing sustained attention as a robust biomarker for vulnerability to substance use. Individual differences in network strength relevant to sustained attention were preserved across developmental stages and sustained attention networks generalized to participants in an external dataset. In summary, brain networks of sustained attention are robust, consistent, and able to predict aspects of later substance use.

A deep learning approach for mental health quality prediction using functional network connectivity and assessment data

While one can characterize mental health using questionnaires, such tools do not provide direct insight into the underlying biology. By linking approaches that visualize brain activity to questionnaires in the context of individualized prediction, we can gain new insights into the biology and behavioral aspects of brain health. Resting-state fMRI (rs-fMRI) can be used to identify biomarkers of these conditions and study patterns of abnormal connectivity. In this work, we estimate mental health quality for individual participants using static functional network connectivity (sFNC) data from rs-fMRI. The deep learning model uses the sFNC data as input to predict four categories of mental health quality and visualize the neural patterns indicative of each group. We used guided gradient class activation maps (guided Grad-CAM) to identify the most discriminative sFNC patterns. The effectiveness of this model was validated using the UK Biobank dataset, in which we showed that our approach outperformed four alternative models by 4-18% accuracy. The proposed model's performance evaluation yielded a classification accuracy of 76%, 78%, 88%, and 98% for the excellent, good, fair, and poor mental health categories, with poor mental health accuracy being the highest. The findings show distinct sFNC patterns across each group. The patterns associated with excellent mental health consist of the cerebellar-subcortical regions, whereas the most prominent areas in the poor mental health category are in the sensorimotor and visual domains. Thus the combination of rs-fMRI and deep learning opens a promising path for developing a comprehensive framework to evaluate and measure mental health. Moreover, this approach had the potential to guide the development of personalized interventions and enable the monitoring of treatment response. Overall this highlights the crucial role of advanced imaging modalities and deep learning algorithms in advancing our understanding and management of mental health.

Structural-functional brain network coupling predicts human cognitive ability

Individual differences in general cognitive ability (GCA) have a biological basis within the structure and function of the human brain. Network neuroscience investigations revealed neural correlates of GCA in structural as well as in functional brain networks. However, whether the relationship between structural and functional networks, the structural-functional brain network coupling (SC-FC coupling), is related to individual differences in GCA remains an open question. We used data from 1030 adults of the Human Connectome Project, derived structural connectivity from diffusion weighted imaging, functional connectivity from resting-state fMRI, and assessed GCA as a latent g-factor from 12 cognitive tasks. Two similarity measures and six communication measures were used to model possible functional interactions arising from structural brain networks. SC-FC coupling was estimated as the degree to which these measures align with the actual functional connectivity, providing insights into different neural communication strategies. At the whole-brain level, higher GCA was associated with higher SC-FC coupling, but only when considering path transitivity as neural communication strategy. Taking region-specific variations in the SC-FC coupling strategy into account and differentiating between positive and negative associations with GCA, allows for prediction of individual cognitive ability scores in a cross-validated prediction framework (correlation between predicted and observed scores: r = 0.25, p < .001). The same model also predicts GCA scores in a completely independent sample (N = 567, r = 0.19, p < .001). Our results propose structural-functional brain network coupling as a neurobiological correlate of GCA and suggest brain region-specific coupling strategies as neural basis of efficient information processing predictive of cognitive ability.

Personalized functional brain network topography is associated with individual differences in youth cognition

Individual differences in cognition during childhood are associated with important social, physical, and mental health outcomes in adolescence and adulthood. Given that cortical surface arealization during development reflects the brain's functional prioritization, quantifying variation in the topography of functional brain networks across the developing cortex may provide insight regarding individual differences in cognition. We test this idea by defining personalized functional networks (PFNs) that account for interindividual heterogeneity in functional brain network topography in 9-10 year olds from the Adolescent Brain Cognitive Development℠ Study. Across matched discovery (n = 3525) and replication (n = 3447) samples, the total cortical representation of fronto-parietal PFNs positively correlates with general cognition. Cross-validated ridge regressions trained on PFN topography predict cognition in unseen data across domains, with prediction accuracy increasing along the cortex's sensorimotor-association organizational axis. These results establish that functional network topography heterogeneity is associated with individual differences in cognition before the critical transition into adolescence.

Enhancing precision in human neuroscience

Human neuroscience has always been pushing the boundary of what is measurable. During the last decade, concerns about statistical power and replicability - in science in general, but also specifically in human neuroscience - have fueled an extensive debate. One important insight from this discourse is the need for larger samples, which naturally increases statistical power. An alternative is to increase the precision of measurements, which is the focus of this review. This option is often overlooked, even though statistical power benefits from increasing precision as much as from increasing sample size. Nonetheless, precision has always been at the heart of good scientific practice in human neuroscience, with researchers relying on lab traditions or rules of thumb to ensure sufficient precision for their studies. In this review, we encourage a more systematic approach to precision. We start by introducing measurement precision and its importance for well-powered studies in human neuroscience. Then, determinants for precision in a range of neuroscientific methods (MRI, M/EEG, EDA, Eye-Tracking, and Endocrinology) are elaborated. We end by discussing how a more systematic evaluation of precision and the application of respective insights can lead to an increase in reproducibility in human neuroscience.

Multimodal Brain Signal Complexity Predicts Human Intelligence

Spontaneous brain activity builds the foundation for human cognitive processing during external demands. Neuroimaging studies based on functional magnetic resonance imaging (fMRI) identified specific characteristics of spontaneous (intrinsic) brain dynamics to be associated with individual differences in general cognitive ability, i.e., intelligence. However, fMRI research is inherently limited by low temporal resolution, thus, preventing conclusions about neural fluctuations within the range of milliseconds. Here, we used resting-state electroencephalographical (EEG) recordings from 144 healthy adults to test whether individual differences in intelligence (Raven's Advanced Progressive Matrices scores) can be predicted from the complexity of temporally highly resolved intrinsic brain signals. We compared different operationalizations of brain signal complexity (multiscale entropy, Shannon entropy, Fuzzy entropy, and specific characteristics of microstates) regarding their relation to intelligence. The results indicate that associations between brain signal complexity measures and intelligence are of small effect sizes (r ∼ 0.20) and vary across different spatial and temporal scales. Specifically, higher intelligence scores were associated with lower complexity in local aspects of neural processing, and less activity in task-negative brain regions belonging to the default-mode network. Finally, we combined multiple measures of brain signal complexity to show that individual intelligence scores can be significantly predicted with a multimodal model within the sample (10-fold cross-validation) as well as in an independent sample (external replication, N = 57). In sum, our results highlight the temporal and spatial dependency of associations between intelligence and intrinsic brain dynamics, proposing multimodal approaches as promising means for future neuroscientific research on complex human traits.

How failure to falsify in high-volume science contributes to the replication crisis

The number of scientific papers published every year continues to increase, but scientific knowledge is not progressing at the same rate. Here we argue that a greater emphasis on falsification - the direct testing of strong hypotheses - would lead to faster progress by allowing well-specified hypotheses to be eliminated. We describe an example from neuroscience where there has been little work to directly test two prominent but incompatible hypotheses related to traumatic brain injury. Based on this example, we discuss how building strong hypotheses and then setting out to falsify them can bring greater precision to the clinical neurosciences, and argue that this approach could be beneficial to all areas of science.