Biology subjects

Dimitriadis, S. I.

Publications and source records attributed to Dimitriadis, S. I..

6 recordsLinked to original sources

bioRxiv2018

Complexity of Brain Activity and Connectivity in Functional Neuroimaging

Understanding the complexity of human brain dynamics and brain connectivity across the repertoire of functional neuroimaging and various conditions, is of paramount importance. Novel measures should be designed tailored to the input focusing on multichannel activity and dynamic functional brain connectivity (DFBC).\n\nHere, we defined a novel complexity index (CI) from the field of symbolic dynamics that quantifies patterns of different words up to a length from a symbolic sequence. The CI characterizes the complexity of the brain activity.\n\nWe analysed dynamic functional brain connectivity by adopting the sliding window approach using imaginary part of phase locking value (iPLV) for EEG/ECoG/MEG and wavelet coherence (WC) for fMRI. Both intra and cross-frequency couplings (CFC) namely phase-to-amplitude were estimated using iPLV/WC at every snapshot of the DFBC. Using proper surrogate analysis, we defined the dominant intrinsic coupling mode (DICM) per pair of regions-of-interest (ROI). The spatio-temporal probability distribution of DICM were reported to reveal the most prominent coupling modes per condition and modality. Finally, a novel flexibility index is defined that quantifies the transition of DICM per pair of ROIs between consecutive time-windows.\n\nThe whole methodology was demonstrated using four neuroimaging datasets (EEG/ECoG/MEG/fMRI).\n\nFinally, we succeeded to totally discriminate healthy controls from schizophrenic using FI and dynamic reconfiguration of DICM. Anesthesia independently of the drug caused a global decreased of complexity in all frequency bands with the exception in {delta} and alters the dynamic reconfiguration of DICM. CI and DICM of MEG/fMRI resting-state recordings in two spatial scales were high reliable.\n\nSignificant StatementIn the present study, we demonstrated novel indexes for the estimation of complexity in both raw brain activity and dynamic functional connectivity. To ort the universality of both indexes for the majority of functional neuroimaging modalities, we adopted open datasets from electro and magneto-encephalography, from electro-corticography and functional magnetic resonance imaging. Both indexes proved informative and reliable across repeat scan sessions. Moreover, we succeeded to totally discriminate with absolute accuracy healthy controls from schizophrenic patients. Both indexes proved sensitive to common anesthetic drugs effect in monkeys and reliable in MEG and fMRI repeat scan sessions. We first reported the notion of cross-frequency coupling in BOLD activity. Our analysis could be adapted it in any task and modality for any hypothesis driven study.

neuroscience↗

bioRxiv2018

Reliability of Static and Dynamic Network Metrics in the Resting-State: A MEG-beamformed Connectivity Analysis

The resting activity of the brain can be described by so-called intrinsic connectivity networks (ICNs), which consist of spatially and temporally distributed, but functionally connected, nodes. The coordinated activity of the resting state can be explored via magnetoencephalography (MEG) by studying frequency-dependent functional brain networks at the source level. Although many algorithms for the analysis of brain connectivity have been proposed, the reliability of network metrics derived from both static and dynamic functional connectivity is still unknown. This is a particular problem for studies of associations between ICN metrics and personality variables or other traits, and for studies of differences between patient and control groups, which both depend critically on the reliability of the metrics used. A detailed investigation of the reliability of metrics derived from resting-state MEG repeat scans is therefore a prerequisite for the development of connectomic biomarkers.\n\nHere, we first estimated both static (SFC) and dynamic functional connectivity (DFC) after beamforming source reconstruction using the imaginary part of the phase locking index (iPLV) and the correlation of the amplitude envelope (CorEnv). Using our approach, functional network microstates (FCstates) were derived from the DFC and chronnectomics were computed from the evolution of FCstates across experimental time. In both temporal scales, the reliability of network metrics (SFC), the FCstates and the related chronnectomics were evaluated for every frequency band.\n\nChronnectomic parameters and FCstates were generally more reliable than node-wise static network metrics. CorEnv-based network metrics were more reproducible at the static approach. This analysis encourages the analysis of MEG resting-state via DFC.

neuroscience↗

bioRxiv2018

Aberrant resting-state functional brain networks in dyslexia: Symbolic mutual information analysis of neuromagnetic signals

Neuroimaging studies have identified a variety of structural and functional connectivity abnormalities in students experiencing reading difficulties. The present study adopted a novel approach to assess the dynamics of resting-state neuromagnetic recordings in the form of symbolic sequences (i.e., repeated patterns of neuromagnetic fluctuations within and/or between sensors).\n\nParticipants were 25 students experiencing severe reading difficulties (RD) and 27 age-matched non-impaired readers (NI) aged 7-14 years. Sensor-level data were first represented as symbolic sequences in eight conventional frequency bands. Next, dominant types of sensor-to-sensor interactions in the form of intra and cross-frequency coupling were computed and subjected to graph modeling to assess group differences in global network characteristics.\n\nAs a group RD students displayed predominantly within-frequency interactions between neighboring sensors which may reflect reduced overall global network efficiency and cost-efficiency of information transfer. In contrast, sensor networks among NI students featured a higher proportion of cross-frequency interactions. Brain-reading achievement associations highlighted the role of left hemisphere temporo-parietal functional networks, at rest, for reading acquisition and ability.\n\nHighlightsO_LISymbolic dynamics of MEG time series revealed aberrant Cross Frequency Coupling in RD students\nC_LIO_LIGlobal efficiency and strength of Cross Frequency Coupling could reliably identify RD students from age-matched controls\nC_LIO_LIGlobal Cost Efficiency, coupling strength, and the relative preponderance of cross-frequency interactions strongly correlated with reading achievement across groups.\nC_LI

neuroscience↗

bioRxiv2017

Random Forest Feature Selection, Fusion and Ensemble Strategy: Combining Multiple Morphological MRI Measures to Discriminate among healthy elderly, MCI, cMCI and Alzheimer's disease patients: from the Alzheimer’s disease neuroimaging initiative (ADNI) database

BackgroundIn the era of computer-assisted diagnostic tools for various brain diseases, Alzheimers disease (AD) covers a large percentage of neuroimaging research, with the main scope being its use in daily practice. However, there has been no study attempting to simultaneously discriminate among Healthy Controls (HC), early mild cognitive impairment (MCI), late MCI (cMCI) and stable AD, using features derived from a single modality, namely MRI.\n\nNew MethodBased on preprocessed MRI images from the organizers of a neuroimaging challenge2, we attempted to quantify the prediction accuracy of multiple morphological MRI features to simultaneously discriminate among HC, MCI, cMCI and AD. We explored the efficacy of a novel scheme that includes multiple feature selections via Random Forest from subsets of the whole set of features (e.g. whole set, left/right hemisphere etc.), Random Forest classification using a fusion approach and ensemble classification via majority voting.\n\nFrom the ADNI database, 60 HC, 60 MCI, 60 cMCI and 60 AD were used as a training set with known labels. An extra dataset of 160 subjects (HC: 40, MCI: 40, cMCI: 40 and AD: 40) was used as an external blind validation dataset to evaluate the proposed machine learning scheme.\n\nResultsIn the second blind dataset, we succeeded in a four-class classification of 61.9% by combining MRI-based features with a Random Forest-based Ensemble Strategy. We achieved the best classification accuracy of all teams that participated in this neuroimaging competition.\n\nComparison with Existing Method(s)The results demonstrate the effectiveness of the proposed scheme to simultaneously discriminate among four groups using morphological MRI features for the very first time in the literature.\n\nConclusionsHence, the proposed machine learning scheme can be used to define single and multi-modal biomarkers for AD.\n\nHIGHLIGHTSO_LI1st place in International Challenge for Automated Prediction of MCI from MRI Data\nC_LIO_LIMulti-class classification of normal control, MCI, converting MCI, and Alzheimers disease\nC_LIO_LIMorphometric measures from 3D T1 brain MRI images have been analysed (ADNI1 cohort).\nC_LIO_LIA Random Forest Feature Selection, Fusion and Ensemble Strategy was applied to classification and prediction of AD.\nC_LIO_LIAccuracy and robustness have been assessed in a blind dataset\nC_LI

neuroscience↗

bioRxiv2017

A Novel, Fast and Efficient Single-Sensor Automatic Sleep-Stage Classification Based on Complementary Cross-Frequency Coupling Estimates

ObjectiveLimitations of the manual scoring of polysomnograms, which include data from electroencephalogram (EEG), electrooculogram (EOG), electrocardiogram (ECG) and electromyogram (EMG) channels, have long been recognised. Manual staging is resource-intensive and time-consuming and considerable efforts have to be spent to ensure inter-rater reliability. There is thus great interest in techniques based on signal processing and machine learning for a completely Automatic Sleep Stage Classification (ASSC).\n\nMethodsIn this paper, we present a single EEG-sensor ASSC technique based on dynamic reconfiguration of different aspects of cross-frequency coupling (CFC) estimated between predefined frequency pairs over 5s epoch lengths. The proposed analytic scheme is demonstrated using the PhysioNet Sleep European Data Format (EDF) Database using 20 healthy young adults with repeat recordings.\n\nResultsWe achieved very high classification sensitivity, specificity and accuracy of 96.2 {+/-} 2.2%, 94.2 {+/-} 2.3%, and 94.4 {+/-} 2.2% across 20 folds, respectively and high mean F1-score (92%, range 90-94%) when multi-class Bayes Naive classifier was applied.\n\nConclusionsOur method outperformed the accuracy of previous studies on different datasets but also on the same database.\n\nSignificanceSingle-sensor ASSC makes the whole methodology appropriate for longitudinal monitoring using wearable EEG in real world and lab-oriented environments.

neuroscience↗

bioRxiv2017

Reconfiguration Of Dominant Coupling Modes In Mild Traumatic Brain Injury Mediated By δ-Band Activity: A Resting State MEG Study

During the last few years, rich-club (RC) organization has been studied as a possible brain-connectivity organization model for large-scale brain networks. At the same time, empirical and simulated data of neurophysiological models have demonstrated the significant role of intra-frequency and inter-frequency coupling among distinct brain areas. The current study investigates further the importance of these couplings using recordings of resting-state magnetoencephalographic activity obtained from 30 mild traumatic brain injury (mTBI) subjects and 50 healthy controls. Intra-frequency and inter-frequency coupling modes are incorporated in a single graph to detect group differences within individual rich-club subnetworks (type I networks) and networks connecting RC nodes with the rest of the nodes (type II networks). Our results show a higher probability of inter-frequency coupling for ({delta}-{gamma}1), ({delta}-{gamma}2), ({theta}-{beta}), ({theta}-{gamma}2), (-{gamma}2), ({gamma}1-{gamma}2) and intra-frequency coupling for ({gamma}1-{gamma}1) and ({delta}-{delta}) for both type I and type II networks in the mTBI group. Additionally, mTBI and control subjects can be correctly classified with high accuracy (98.6%), whereas a general linear regression model can effectively predict the subject group using the ratio of type I and type II coupling in the ({delta}, {theta}), ({delta}, {beta}), ({delta}, {gamma}1), and ({delta}, {gamma}2) frequency pairs. These findings support the presence of an RC organization simultaneously with dominant frequency interactions within a single functional graph. Our results demonstrate a hyperactivation of intrinsic RC networks in mTBI subjects compared to controls, which can be seen as a plausible compensatory mechanism for alternative frequency-dependent routes of information flow in mTBI subjects.

neuroscience↗