Brain connectivity is studied as a functionally connected network using statistical methods such as measuring correlation or covariance. The non-invasive neuroimaging techniques such as Electroencephalography (EEG) signals are converted to networks by transforming the signals into a Correlation Matrix and analyzing the resulting networks. Here, four learning models, namely, Logistic Regression, Random Forest, Support Vector Machine, and Recurrent Neural Networks (RNN), are implemented on two different types of correlation matrices: Correlation Matrix (static connectivity) and Time-resolved Correlation Matrix (dynamic connectivity), to classify them either on their psychometric assessment or the effect of therapy. These correlation matrices are different from traditional learning techniques in the sense that they incorporate theory-based graph features into the learning models, thus providing novelty to this study. The EEG data used in this study is trail-based/event-related from five different experimental paradigms, of which can be broadly classified as working memory tasks and assessment of emotional states (depression, anxiety, and stress). The classifications based on RNN provided higher accuracy (74–88%) than the other three models (50–78%). Instead of using individual graph features, a Correlation Matrix provides an initial test of the data. When compared with the Time-resolved Correlation Matrix, it offered a 4–5% higher accuracy. The Time-resolved Correlation Matrix is better suited for dynamic studies here; it provides lower accuracy when compared to the Correlation Matrix, a static feature.

Introduction: Network analysis allows investigators to explore the many facets of brain networks, particularly the proliferation of disease, using graph theory to model the disease movement. One of the hypotheses behind the disruption in brain networks in Alzheimer’s disease (AD) is the abnormal accumulation of beta-amyloid plaques and tau protein tangles. In this study, the potential use of percolation centrality to study the movement of beta-amyloids, as a feature of given PET image-based networks, is studied. The PET image-based network construction is possible using a public access database - Alzheimer’s Disease Neuroimaging Initiative, which provided 551 scans. For each image, the Julich atlas provides 121 regions of interest, which are the network nodes. Besides, the influential nodes for each scan are calculated using the collective influence algorithm.Results: Analysis of variance (p¡0.05) yields the region of interest GM Superior parietal lobule 7A L, for which percolation centrality is significant irrespective of the tracer type. Pairwise variance analysis between the clinical groups provides five and twelve Regions of Interest for AV45 and PiB. Multivariate linear regression between the percolation centrality values for nodes and psychometric assessment scores reveals Mini-Mental State Examination is a reliable metric. Finally, a ranking of the regions of interest is made based on the collective influence algorithm to indicate the anatomical areas strongly influencing the beta-amyloid network. Through this study, it is possible to use percolation centrality values to indicate the regions of interest that reflect the disease’s spread.

Functional Connectivity analysis using Electroencephalography signals is common. The EEG signals are converted to networks by transforming the signals into a correlation matrix and analyzing the resulting networks. Here, four learning models, namely, Logistic Regression, Random Forest, Support Vector Machine, and Recurrent Neural Networks, are implemented on the correlation matrix data to classify them either on their psychometric assessment or the effect of therapy; The EEG data is trail-based/event-related. The classifications based on RNN provided higher accuracy( 74-88%) than the other three models( 50-78%). Instead of using individual graph features, a correlation matrix provides an initial test of the data. When compared with the time-resolved correlation matrix, it offered a 4-5% higher accuracy. The time-resolved correlation matrix is better suited for dynamic studies here; it provides lower accuracy when compared to the correlation matrix, a static feature.

Substantial adolescence is spent in an academic environment where the student can experience varying intensities of depression, stress, and anxiety, which can be fatal. To address this concern, we utilized the Depression Anxiety and Stress Survey (DASS) 21 and Modified Sternberg working memory, thereby assessing the emotional states and assessing the impact on the cognitive ability of students (n=37, F=7) in terms of working memory. An intervention was provided (Art of Living YES+ program) for ten days. The assessment is carried out in the time window of two months before and after the intervention. F-test and T-test(p≤0.05) on the scores and reaction time are performed for hypothesis testing. This statistical analysis reveals that both the depression category and stress category reject the null hypothesis. Among the thirty-seven, only five students took part in the post-intervention assessment, the scores in 28% of the questions had lower scores, and 19 % did not have any change; however, there was an increase in the scores in 42% of the questions. No significant changes are observed in the working memory ability of the students. Based on reaction time analysis: 11.62%, 16.27%, and 25.58% are outliers for each type of question, respectively. Two participants showed significantly lower reaction times, indicating a faster reading ability than the rest. This study shows that the intervention can positively impact emotional states-depression, stress, and affect working memory abilities.

Network analysis allows investigators to explore the many facets of brain networks, particularly the proliferation of disease using graph theory to model the disease movement. The disruption in brain networks in Alzheimer's disease (AD) is due to the abnormal accumulation of beta-amyloid plaques and tau protein tangles. In this study, the potential use of percolation centrality to study the movement of beta-amyloid plaques, as a feature of given PET image-based networks, is studied. The PET image-based network construction is possible using the public access database - Alzheimer's Disease Neuroimaging Initiative, which provided 1522 scans, of which 429 are of AD patients, 583 of patients with mild cognitive impairment, and 510 of cognitively normal. For each image, the Julich atlas provides 121 regions of interest/network nodes. Additionally, the influential nodes for each scan are calculated using the collective influence algorithm. Through this study, it is possible to use percolation centrality values to indicate the regions of interest that reflect the disease's spread and show potential use for early AD diagnosis. Analysis of variance (ANOVA) shows the regions of interest for which percolation centrality is a valid measure, irrespective of the tracer type. A multivariate linear regression between the percolation centrality values for each of the nodes and psychometric assessment scores reveals that models Mini-Mental State Examination (MMSE) scores performed better than ones with Neuropsychiatric Inventory Questionnaire (NPIQ) scores as the target variable. Similar to ANOVA, the multivariate linear regression yields regions of interest for which percolation centrality is a good differentiator. Finally, a ranking of the regions of interest is made based on the collective influence algorithm to indicate the anatomical areas strongly influencing the beta-amyloid network.