Statistical Approaches to Identify Pairwise and High-Order Brain Functional Connectivity Signatures on a Single-Subject Basis

Keeping up with the shift towards personalized neuroscience essentially requires the derivation of meaningful insights from individual brain signal recordings by analyzing the descriptive indexes of physio-pathological states through statistical methods that prioritize subject-specific differences under varying experimental conditions. Within this framework, the current study presents a methodology for assessing the value of the single-subject fingerprints of brain functional connectivity, assessed both by standard pairwise and novel high-order measures. Functional connectivity networks, which investigate the inter-relationships between pairs of brain regions, have long been a valuable tool for modeling the brain as a complex system. However, their usefulness is limited by their inability to detect high-order dependencies beyond pairwise correlations. In this study, by leveraging multivariate information theory, we confirm recent evidence suggesting that the brain contains a plethora of high-order, synergistic subsystems that would go unnoticed using a pairwise graph structure. The significance and variations across different conditions of functional pairwise and high-order interactions (HOIs) between groups of brain signals are statistically verified on an individual level through the utilization of surrogate and bootstrap data analyses. The approach is illustrated on the single-subject recordings of resting-state functional magnetic resonance imaging (rest-fMRI) signals acquired using a pediatric patient with hepatic encephalopathy associated with a portosystemic shunt and undergoing liver vascular shunt correction. Our results show that (i) the proposed single-subject analysis may have remarkable clinical relevance for subject-specific investigations and treatment planning, and (ii) the possibility of investigating brain connectivity and its post-treatment functional developments at a high-order level may be essential to fully capture the complexity and modalities of the recovery.

Personalized Lifestyle Intervention and Functional Evaluation Health Outcomes SurvEy: Presentation of the LIFEHOUSE Study Using N-of-One Tent-Umbrella-Bucket Design

The working definition of health is often the simple absence of diagnosed disease. This common standard is limiting given that changes in functional health status represent early warning signs of impending health declines. Longitudinal assessment of functional health status may foster prevention of disease occurrence and modify disease progression. The LIFEHOUSE (Lifestyle Intervention and Functional Evaluation-Health Outcomes SurvEy) longitudinal research project explores the impact of personalized lifestyle medicine approaches on functional health determinants. Utilizing an adaptive tent-umbrella-bucket design, the LIFEHOUSE study follows the functional health outcomes of adult participants recruited from a self-insured employee population. Participants were each allocated to the tent of an all-inclusive N-of-one case series. After assessing medical history, nutritional physical exam, baseline functional status (utilizing validated tools to measure metabolic, physical, cognitive, emotional and behavioral functional capacity), serum biomarkers, and genomic and microbiome markers, participants were assigned to applicable umbrellas and buckets. Personalized health programs were developed and implemented using systems biology formalism and functional medicine clinical approaches. The comprehensive database (currently 369 analyzable participants) will yield novel interdisciplinary big-health data and facilitate topological analyses focusing on the interactome among each participant's genomics, microbiome, diet, lifestyle and environment.

Differences in Motor Control Strategies of Jumping Tasks, as Revealed by Group and Individual Analysis

The aim of this study was to investigate the motor control strategies adopted when performing two jumping tasks with different task demands when analysed at an individual and group level. Twenty-two healthy individuals performed two jumping tasks: jumping without the use of an arm swing (CMJnas) and jumping starting in a plantar flexed position with the use of an arm swing (PF). Principal component analysis (PCA) was performed using hip, knee and ankle joint moment data on individual (PCAi) and group data (PCAc). The results demonstrate that a greater number of PCs are required to explain the majority of variance within the dataset in the PF condition at both an individual and group level, compared to CMJnas condition. Although common control strategies were observed between the two jumping conditions, differences in the organisation of the movement (PC loading coefficients) were observed. Results from the group analysis did not completely reflect the individual strategies used to perform each jumping task and highlight the value in performing individual analysis to determine emergent control strategies.

The correlation between jump height and mechanical power in a countermovement jump is artificially inflated

The countermovement jump is commonly used to assess an athlete's neuromuscular capacity. The aim of this study was to identify the mechanism behind the strong correlation between jump height and mechanical power in a countermovement jump. Three athletes each performed between 47 and 60 maximal-effort countermovement jumps on a force platform. For all three athletes, peak mechanical power and average mechanical power were strongly correlated with jump height (r = 0.54-0.90). The correlation between jump height and peak power was largely determined by the correlation between jump height and the velocity at peak power (r = 0.83-0.94) and was not related to the correlation between jump height and the ground reaction force at peak power (r = -0.20-0.18). These results confirm that the strong correlation between jump height and power is an artefact arising from how power is calculated. Power is a compound variable calculated from the product of instantaneous ground reaction force and instantaneous velocity, and application of statistical theory shows that the correlation between jump height and power is artificially inflated by the near-perfect correlation between jump height and the velocity at peak power. Despite this finding, mechanical power might still be useful in assessing the neuromuscular capacity of an athlete.

Joint- and subject-specific strategies in male basketball players across a range of countermovement jump heights

The purpose of this study was to investigate subject- and joint-specific strategies used by male basketball players as they increase their countermovement jump (CMJ) height from sub-maximal to maximal efforts. Lower extremity joint kinematics and kinetics were recorded as 11 male, NCAA Division I basketball players performed 8-10 CMJ across effort levels of approximately 25%, 50%, 75% and 100%. Simple correlation models were used to investigate the associations between effort levels (i.e., CMJ height) and joint mechanics (i.e., negative (eccentric) and positive (concentric) mechanical work performed at the hip, knee, and ankle joints) for each individual player and the entire group. Group-analyses showed that increases in all joint mechanical variables were associated with increases in CMJ height. In contrast, single-subject analyses revealed that players used individualised strategies, and selectively scaled the magnitude of mechanical work at none (n = 2), one (n = 2), two (n = 5), or all three (n = 2) joints as they increased CMJ efforts. In addition, individual players also appeared to selectively scale different combinations of eccentric or concentric joint work as they increased CMJ height. These results highlight that male basketball players use joint-specific strategies to increase CMJ height when progressively increasing CMJ effort.

Weighted Walking Influences Lower Extremity Coordination in Children on the Autism Spectrum

There is sparse quantitative research regarding gait coordination patterns of children on the autism spectrum, though previous studies, relying only upon observational data, have alluded to characteristically poor movement coordination. This study compared walking with a weighted vest, a backpack carriage, and an unloaded walking condition on lower extremity coordination among 10 male children (aged 8-17 years) on the autism spectrum. All participants completed 15 gait trials in the following three conditions: (a) unloaded, (b) wearing a backpack weighted with 15% body mass, and (c) wearing a vest weighted with 15% body mass. We used continuous relative phase analysis to quantify lower extremity coordination and analyzed data through both group and single-subject comparisons. We used the Model Statistic to test for statistical significance at each of the normalized data points for each segment couple (thigh-leg, leg-foot, and thigh-foot). The first 10 and last 10 stride blocks were tested for possible accommodation strategies. Group comparisons revealed no coordination changes among the three conditions (likely due to insufficient statistical power), while single-subject comparisons exposed significant decreased variability in gait coordination patterns ( p < .05) in both loaded conditions, relative to the unloaded condition. These participants exhibited variable coordination patterns during the unloaded gait. When walking with loads, coordination pattern variability of the lower extremities was decreased. This finding suggests that walking while carrying or wearing heavy objects may reduce the number of potential motor pattern choices and thus decrease the overall variability of lower extremity movement patterns. Additional research with a larger and more diverse participant sample is required to confirm this conclusion.

Applying fMRI complexity analyses to the single subject: a case study for proposed neurodiagnostics

Nonlinear dynamic tools have been statistically validated at the group level to identify subtle differences in system wide regulation of brain meso-circuits, often increasing clinical sensitivity over conventional analyses alone. We explored the feasibility of extracting information at the single-subject level, illustrating two pairs of healthy individuals with psychological differences in stress reactivity. We applied statistical and nonlinear dynamic tools to capture key characteristics of the prefrontal-limbic loop. We compared single subject results with statistical results for the larger group. We concluded that complexity analyses may identify important differences at the single-subject level, supporting their potential towards neurodiagnostic applications.