Flow starvation during square-flow assisted ventilation detected by supervised deep learning techniques.

BACKGROUND: Flow starvation is a type of patient-ventilator asynchrony that occurs when gas delivery does not fully meet the patients' ventilatory demand due to an insufficient airflow and/or a high inspiratory effort, and it is usually identified by visual inspection of airway pressure waveform. Clinical diagnosis is cumbersome and prone to underdiagnosis, being an opportunity for artificial intelligence. Our objective is to develop a supervised artificial intelligence algorithm for identifying airway pressure deformation during square-flow assisted ventilation and patient-triggered breaths. METHODS: Multicenter, observational study. Adult critically ill patients under mechanical ventilation > 24 h on square-flow assisted ventilation were included. As the reference, 5 intensive care experts classified airway pressure deformation severity. Convolutional neural network and recurrent neural network models were trained and evaluated using accuracy, precision, recall and F1 score. In a subgroup of patients with esophageal pressure measurement (ΔPes), we analyzed the association between the intensity of the inspiratory effort and the airway pressure deformation. RESULTS: 6428 breaths from 28 patients were analyzed, 42% were classified as having normal-mild, 23% moderate, and 34% severe airway pressure deformation. The accuracy of recurrent neural network algorithm and convolutional neural network were 87.9% [87.6-88.3], and 86.8% [86.6-87.4], respectively. Double triggering appeared in 8.8% of breaths, always in the presence of severe airway pressure deformation. The subgroup analysis demonstrated that 74.4% of breaths classified as severe airway pressure deformation had a ΔPes > 10 cmH2O and 37.2% a ΔPes > 15 cmH2O. CONCLUSIONS: Recurrent neural network model appears excellent to identify airway pressure deformation due to flow starvation. It could be used as a real-time, 24-h bedside monitoring tool to minimize unrecognized periods of inappropriate patient-ventilator interaction.

Modeling the Hemodynamic Response Function Using EEG-fMRI Data During Eyes-Open Resting-State Conditions and Motor Task Execution.

Being able to accurately quantify the hemodynamic response function (HRF) that links the blood oxygen level dependent functional magnetic resonance imaging (BOLD-fMRI) signal to the underlying neural activity is important both for elucidating neurovascular coupling mechanisms and improving the accuracy of fMRI-based functional connectivity analyses. In particular, HRF estimation using BOLD-fMRI is challenging particularly in the case of resting-state data, due to the absence of information about the underlying neuronal dynamics. To this end, using simultaneously recorded electroencephalography (EEG) and fMRI data is a promising approach, as EEG provides a more direct measure of neural activations. In the present work, we employ simultaneous EEG-fMRI to investigate the regional characteristics of the HRF using measurements acquired during resting conditions. We propose a novel methodological approach based on combining distributed EEG source space reconstruction, which improves the spatial resolution of HRF estimation and using block-structured linear and nonlinear models, which enables us to simultaneously obtain HRF estimates and the contribution of different EEG frequency bands. Our results suggest that the dynamics of the resting-state BOLD signal can be sufficiently described using linear models and that the contribution of each band is region specific. Specifically, it was found that sensory-motor cortices exhibit positive HRF shapes, whereas the lateral occipital cortex and areas in the parietal cortex, such as the inferior and superior parietal lobule exhibit negative HRF shapes. To validate the proposed method, we repeated the analysis using simultaneous EEG-fMRI measurements acquired during execution of a unimanual hand-grip task. Our results reveal significant associations between BOLD signal variations and electrophysiological power fluctuations in the ipsilateral primary motor cortex, particularly for the EEG beta band, in agreement with previous studies in the literature.

Physiological and motion signatures in static and time-varying functional connectivity and their subject identifiability.

Human brain connectivity yields significant potential as a noninvasive biomarker. Several studies have used fMRI-based connectivity fingerprinting to characterize individual patterns of brain activity. However, it is not clear whether these patterns mainly reflect neural activity or the effect of physiological and motion processes. To answer this question, we capitalize on a large data sample from the Human Connectome Project and rigorously investigate the contribution of the aforementioned processes on functional connectivity (FC) and time-varying FC, as well as their contribution to subject identifiability. We find that head motion, as well as heart rate and breathing fluctuations, induce artifactual connectivity within distinct resting-state networks and that they correlate with recurrent patterns in time-varying FC. Even though the spatiotemporal signatures of these processes yield above-chance levels in subject identifiability, removing their effects at the preprocessing stage improves identifiability, suggesting a neural component underpinning the inter-individual differences in connectivity.

Estimating brain age from structural MRI and MEG data: Insights from dimensionality reduction techniques.

Brain age prediction studies aim at reliably estimating the difference between the chronological age of an individual and their predicted age based on neuroimaging data, which has been proposed as an informative measure of disease and cognitive decline. As most previous studies relied exclusively on magnetic resonance imaging (MRI) data, we hereby investigate whether combining structural MRI with functional magnetoencephalography (MEG) information improves age prediction using a large cohort of healthy subjects (N = 613, age 18-88 years) from the Cam-CAN repository. To this end, we examined the performance of dimensionality reduction and multivariate associative techniques, namely Principal Component Analysis (PCA) and Canonical Correlation Analysis (CCA), to tackle the high dimensionality of neuroimaging data. Using MEG features (mean absolute error (MAE) of 9.60 years) yielded worse performance when compared to using MRI features (MAE of 5.33 years), but a stacking model combining both feature sets improved age prediction performance (MAE of 4.88 years). Furthermore, we found that PCA resulted in inferior performance, whereas CCA in conjunction with Gaussian process regression models yielded the best prediction performance. Notably, CCA allowed us to visualize the features that significantly contributed to brain age prediction. We found that MRI features from subcortical structures were more reliable age predictors than cortical features, and that spectral MEG measures were more reliable than connectivity metrics. Our results provide an insight into the underlying processes that are reflective of brain aging, yielding promise for the identification of reliable biomarkers of neurodegenerative diseases that emerge later during the lifespan.

Older adults exhibit a more pronounced modulation of beta oscillations when performing sustained and dynamic handgrips.

Muscle contractions are associated with a decrease in beta oscillatory activity, known as movement-related beta desynchronization (MRBD). Older adults exhibit a MRBD of greater amplitude compared to their younger counterparts, even though their beta power remains higher both at rest and during muscle contractions. Further, a modulation in MRBD has been observed during sustained and dynamic pinch contractions, whereby beta activity during periods of steady contraction following a dynamic contraction is elevated. However, how the modulation of MRBD is affected by aging has remained an open question. In the present work, we investigated the effect of aging on the modulation of beta oscillations and their putative link with motor performance. We collected magnetoencephalography (MEG) data from younger and older adults during a resting-state period and motor handgrip paradigms, which included sustained and dynamic contractions, to quantify spontaneous and motor-related beta oscillatory activity. Beta power at rest was found to be significantly increased in the motor cortex of older adults. During dynamic hand contractions, MRBD was more pronounced in older participants in frontal, premotor and motor brain regions. These brain areas also exhibited age-related decreases in cortical thickness; however, the magnitude of MRBD and cortical thickness were not found to be associated after controlling for age. During sustained hand contractions, MRBD exhibited a decrease in magnitude compared to dynamic contraction periods in both groups and did not show age-related differences. This suggests that the amplitude change in MRBD between dynamic and sustained contractions is larger in older compared to younger adults. We further probed for a relationship between beta oscillations and motor behaviour and found that greater MRBD in primary motor cortices was related to degraded motor performance beyond age, but our results suggested that age-related differences in beta oscillations were not predictive of motor performance.

Functional and effective reorganization of the aging brain during unimanual and bimanual hand movements.

Motor performance decline observed during aging is linked to changes in brain structure and function, however, the precise neural reorganization associated with these changes remains largely unknown. We investigated the neurophysiological correlates of this reorganization by quantifying functional and effective brain network connectivity in elderly individuals (n = 11; mean age = 67.5 years), compared to young adults (n = 12; mean age = 23.7 years), while they performed visually-guided unimanual and bimanual handgrips inside the magnetoencephalography (MEG) scanner. Through a combination of principal component analysis and Granger causality, we observed age-related increases in functional and effective connectivity in whole-brain, task-related motor networks. Specifically, elderly individuals demonstrated (i) greater information flow from contralateral parietal and ipsilateral secondary motor regions to the left primary motor cortex during the unimanual task and (ii) decreased interhemispheric temporo-frontal communication during the bimanual task. Maintenance of motor performance and task accuracy in elderly was achieved by hyperactivation of the task-specific motor networks, reflecting a possible mechanism by which the aging brain recruits additional resources to counteract known myelo- and cytoarchitectural changes. Furthermore, resting-state sessions acquired before and after each motor task revealed that both older and younger adults maintain the capacity to adapt to task demands via network-wide increases in functional connectivity. Collectively, our study consolidates functional connectivity and directionality of information flow in systems-level cortical networks during aging and furthers our understanding of neuronal flexibility in motor processes.

Identification of Time-Varying Cortico-cortical and Cortico-Muscular Coherence during Motor Tasks with Multivariate Autoregressive Models.

Neural populations coordinate at fast subsecond time-scales during rest and task execution. As a result, functional brain connectivity assessed with different neuroimaging modalities (EEG, MEG, fMRI) may also change over different time scales. In addition to the more commonly used sliding window techniques, the General Linear Kalman Filter (GLFK) approach has been proposed to estimate time-varying brain connectivity. In the present work, we propose a modification of the GLFK approach to model timevarying connectivity. We also propose a systematic method to select the hyper-parameters of the model. We evaluate the performance of the method using MEG and EMG data collected from 12 young subjects performing two motor tasks (unimanual and bimanual hand grips), by quantifying time-varying cortico-cortical and corticomuscular coherence (CCC and CMC). The CMC results revealed patterns in accordance with earlier findings, as well as an improvement in both time and frequency resolution compared to sliding window approaches. These results suggest that the proposed methodology is able to unveil accurate time-varying connectivity patterns with an excellent time resolution.