Granger causality test with nonlinear neural-network-based methods: Python package and simulation study

Rosoł, Maciej; Młyńczak, Marcel; Cybulski, Gerard

doi:10.1016/j.cmpb.2022.106669

“…Nonetheless, in those studies more heterogeneous groups of patients were present in terms of age or health status (patients after heart failure or individuals with low to moderate risk of cardiovascular diseases). Further improvement of the prediction of VO 2peak might be achieved by increasing sample size, and inclusion of other parameters based on the raw signals (especially ECG) like HRV and parameters from information and causal domain (56)(57)(58)(59).…”

Section: Discussionmentioning

confidence: 99%

Prediction of peak oxygen consumption using cardiorespiratory parameters from warm-up and submaximal stage of treadmill cardiopulmonary exercise test

Rosoł,

Petelczyc,

Gąsior

et al. 2023

Preprint

Self Cite

0

View full text Add to dashboard Cite

This study investigates the quality of peak oxygen consumption (VO2peak) prediction based on cardiac and respiratory parameters calculated from warmup and submaximal stages of treadmill cardiopulmonary exercise test (CPET) using machine learning (ML) techniques and assesses the importance of respiratory parameters for the prediction outcome. The database consists of the following parameters: heart rate (HR), respiratory rate (RespRate), pulmonary ventilation (VE), oxygen consumption (VO2) and carbon dioxide production (VCO2) obtained from 369 treadmill CPETs. Combinations of features calculated based on the HR, VE and RespRate time-series from different stages of CPET were used to create 11 datasets for VO2peak prediction. Thirteen ML algorithms were employed, and model performances were evaluated using cross-validation with mean absolute percentage error (MAPE), R2 score, mean absolute error (MAE), and root mean squared error (RMSE) calculated after each iteration of the validation. The results demonstrated that incorporating respiratory-based features improves the prediction of VO2peak. The best results in terms of R2 score (0.47) and RMSE (5.78) were obtained for the dataset which included both cardiac- and respiratory-based features from CPET up to 85% of age-predicted HRmax, while the best results in terms of MAPE (10.5%) and MAE (4.63) were obtained for the dataset containing cardiorespiratory features from the last 30 seconds of warmup. The study showed the potential of using ML models based on cardiorespiratory features from submaximal tests for prediction of VO2peak and highlights the importance of the monitoring of respiratory signals, enabling to include respiratory parameters into the analysis. Presented approach offers a feasible alternative to direct VO2peak measurement, especially when specialized equipment is limited or unavailable.

show abstract

“…The computational theory of GC is that for two time series X and Y , if the past information of X is useful for predicting the future of Y , it can be considered that “ X ” causes “ Y ”, denoted as

{F}_{X - &gt; Y}

. A conditional GC (CGC) is an extension of the GC in multivariate autoregressive models that aim to reduce false‐positive causal relationships 40 . The following computation steps were implemented using the GCCA toolbox, 41 as described below.…”

Section: Methodsmentioning

confidence: 99%

“…A conditional GC (CGC) is an extension of the GC in multivariate autoregressive models that aim to reduce false-positive causal relationships. 40 The following computation steps were implemented using the GCCA toolbox, 41 as described below.…”

Section: Effective Network Analysismentioning

confidence: 99%

FNIRS based study of brain network characteristics in children with cerebral palsy during bilateral lower limb movement

Xie,

Nie,

Zhang

et al. 2024

Medical Physics

0

View full text Add to dashboard Cite

BackgroundMotor dysfunctions in children with cerebral palsy (CP) are caused by nonprogressive brain damage. Understanding the functional characteristics of the brain is important for rehabilitation.PurposeThis paper aimed to study the brain networks of children with CP during bilateral lower limb movement using functional near‐infrared spectroscopy (fNIRS) and to explore effective fNIRS indices for reflecting functional brain activity.MethodsUsing fNIRS, cerebral oxygenation signals in the bilateral prefrontal cortex (LPFC/RPFC) and motor cortex (LMC/RMC) were recorded from fifteen children with spastic CP and seventeen children with typical development (CTDs) in the resting state and during bilateral lower limb movement. Functional connectivity matrices based on phase‐locking values (PLVs) were calculated using Hilbert transformation, and binary networks were constructed at different sparsity levels. Network metrics such as the clustering coefficient, global efficiency, local efficiency, and transitivity were calculated. Furthermore, the time‐varying curves of network metrics during movement were obtained by dividing the time window and using sparse inverse covariance matrices. Finally, conditional Granger causality (GC) was used to explore the causal relationships between different brain regions.ResultsCompared to CTDs, the connectivity between RMC‐RPFC (p = 0.017) and RMC‐LMC (p = 0.002) in the brain network was decreased in children with CP, and the clustering coefficient (p = 0.003), global efficiency (p = 0.034), local efficiency (p = 0.015), and transitivity (p = 0.009) were significantly lower. The standard deviation of the changes in global efficiency of children with CP during motion was also greater than that of CTDs. Using GC, it was found that there was a significant increase in causal strength from the RMC to the RPFC (p = 0.04) and from the RMC to the LMC (p = 0.042) in children with CP during motion. Additionally, there were significant negative correlations between the PLV of LMC‐RMC (p = 0.002) and the Gross Motor Function Classification System (GMFCS) and between the GMFCS and the clustering coefficient (p = 0.01).ConclusionsDuring rehabilitation training of the lower limbs, there were significant differences in brain network indices between children with CP and CTDs. The indicators proposed in this paper are effective at evaluating motor function and the real‐time impact of rehabilitation training on the brain network and have great potential for application in guiding clinical motor function assessment and planning rehabilitation strategies.

show abstract

“…Nonetheless, in those studies more heterogeneous groups of patients were present in terms of age or health status (patients after heart failure or individuals with low to moderate risk of cardiovascular diseases). Further improvement of the prediction of VO 2peak might be achieved by increasing sample size, and inclusion of other parameters based on the raw signals (especially ECG) like HRV and parameters from information and causal domain [57][58][59][60].…”

Section: Plos Onementioning

confidence: 99%

Prediction of peak oxygen consumption using cardiorespiratory parameters from warmup and submaximal stage of treadmill cardiopulmonary exercise test

Rosoł,

Petelczyc,

Gąsior

et al. 2024

PLoS ONE

0

View full text Add to dashboard Cite

This study investigates the quality of peak oxygen consumption (VO2peak) prediction based on cardiac and respiratory parameters calculated from warmup and submaximal stages of treadmill cardiopulmonary exercise test (CPET) using machine learning (ML) techniques and assesses the importance of respiratory parameters for the prediction outcome. The database consists of the following parameters: heart rate (HR), respiratory rate (RespRate), pulmonary ventilation (VE), oxygen consumption (VO2) and carbon dioxide production (VCO2) obtained from 369 treadmill CPETs. Combinations of features calculated based on the HR, VE and RespRate time-series from different stages of CPET were used to create 11 datasets for VO2peak prediction. Thirteen ML algorithms were employed, and model performances were evaluated using cross-validation with mean absolute percentage error (MAPE), R2 score, mean absolute error (MAE), and root mean squared error (RMSE) calculated after each iteration of the validation. The results demonstrated that incorporating respiratory-based features improves the prediction of VO2peak. The best results in terms of R2 score (0.47) and RMSE (5.78) were obtained for the dataset which included both cardiac- and respiratory-based features from CPET up to 85% of age-predicted HRmax, while the best results in terms of MAPE (10.5%) and MAE (4.63) were obtained for the dataset containing cardiorespiratory features from the last 30 seconds of warmup. The study showed the potential of using ML models based on cardiorespiratory features from submaximal tests for prediction of VO2peak and highlights the importance of the monitoring of respiratory signals, enabling to include respiratory parameters into the analysis. Presented approach offers a feasible alternative to direct VO2peak measurement, especially when specialized equipment is limited or unavailable.

show abstract

Granger causality test with nonlinear neural-network-based methods: Python package and simulation study

Cited by 30 publications

References 41 publications

Prediction of peak oxygen consumption using cardiorespiratory parameters from warm-up and submaximal stage of treadmill cardiopulmonary exercise test

Prediction of peak oxygen consumption using cardiorespiratory parameters from warm-up and submaximal stage of treadmill cardiopulmonary exercise test

FNIRS based study of brain network characteristics in children with cerebral palsy during bilateral lower limb movement

Prediction of peak oxygen consumption using cardiorespiratory parameters from warmup and submaximal stage of treadmill cardiopulmonary exercise test

Contact Info

Product

Resources

About