François Paugam, B. Pinsard, Guillaume Lajoie, Pierre Bellec
Abstract Dense functional magnetic resonance imaging datasets open new avenues to create auto-regressive models of brain activity. Individual idiosyncrasies are obscured by group models, but can be captured by purely individual models given sufficient amounts of training data. In this study, we compared several deep and shallow individual models on the temporal auto-regression of BOLD time-series recorded during a natural video-watching task. The best performing models were then analyzed in terms of their data requirements and scaling, subject specificity, and the space-time structure of their predicted dynamics. We found the Chebnets, a type of graph convolutional neural network, to be best suited for temporal BOLD auto-regression, closely followed by linear models. Chebnets demonstrated an increase in performance with increasing amounts of data, with no complete saturation at 9 h of training data. Good generalization to other kinds of video stimuli and to resting-state data marked the Chebnets’ ability to capture intrinsic brain dynamics rather than only stimulus-specific autocorrelation patterns. Significant subject specificity was found at short prediction time lags. The Chebnets were found to capture lower frequencies at longer prediction time lags, and the spatial correlations in predicted dynamics were found to match traditional functional connectivity networks. Overall, these results demonstrate that large individual functional magnetic resonance imaging (fMRI) datasets can be used to efficiently train purely individual auto-regressive models of brain activity, and that massive amounts of individual data are required to do so. The excellent performance of the Chebnets likely reflects their ability to combine spatial and temporal interactions on large time scales at a low complexity cost. The non-linearities of the models did not appear as a key advantage. In fact, surprisingly, linear versions of the Chebnets appeared to outperform the original non-linear ones. Individual temporal auto-regressive models have the potential to improve the predictability of the BOLD signal. This study is based on a massive, publicly-available dataset, which can serve for future benchmarks of individual auto-regressive modeling.
{"title":"A benchmark of individual auto-regressive models in a massive fMRI dataset","authors":"François Paugam, B. Pinsard, Guillaume Lajoie, Pierre Bellec","doi":"10.1162/imag_a_00228","DOIUrl":"https://doi.org/10.1162/imag_a_00228","url":null,"abstract":"Abstract Dense functional magnetic resonance imaging datasets open new avenues to create auto-regressive models of brain activity. Individual idiosyncrasies are obscured by group models, but can be captured by purely individual models given sufficient amounts of training data. In this study, we compared several deep and shallow individual models on the temporal auto-regression of BOLD time-series recorded during a natural video-watching task. The best performing models were then analyzed in terms of their data requirements and scaling, subject specificity, and the space-time structure of their predicted dynamics. We found the Chebnets, a type of graph convolutional neural network, to be best suited for temporal BOLD auto-regression, closely followed by linear models. Chebnets demonstrated an increase in performance with increasing amounts of data, with no complete saturation at 9 h of training data. Good generalization to other kinds of video stimuli and to resting-state data marked the Chebnets’ ability to capture intrinsic brain dynamics rather than only stimulus-specific autocorrelation patterns. Significant subject specificity was found at short prediction time lags. The Chebnets were found to capture lower frequencies at longer prediction time lags, and the spatial correlations in predicted dynamics were found to match traditional functional connectivity networks. Overall, these results demonstrate that large individual functional magnetic resonance imaging (fMRI) datasets can be used to efficiently train purely individual auto-regressive models of brain activity, and that massive amounts of individual data are required to do so. The excellent performance of the Chebnets likely reflects their ability to combine spatial and temporal interactions on large time scales at a low complexity cost. The non-linearities of the models did not appear as a key advantage. In fact, surprisingly, linear versions of the Chebnets appeared to outperform the original non-linear ones. Individual temporal auto-regressive models have the potential to improve the predictability of the BOLD signal. This study is based on a massive, publicly-available dataset, which can serve for future benchmarks of individual auto-regressive modeling.","PeriodicalId":507939,"journal":{"name":"Imaging Neuroscience","volume":"2004 16","pages":"1-23"},"PeriodicalIF":0.0,"publicationDate":"2024-07-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"141707021","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
A. Harvey, Clara A Moreau, K. Kumar, Guillaume Huguet, S. Urchs, H. Sharmarke, K. Jizi, Charles-Olivier Martin, N. Younis, P. Tamer, J.-L. Martineau, P. Orban, Ana Isabel Silva, Jeremy Hall, M. B. van den Bree, Michael J. Owen, David E J Linden, Sarah Lippé, C. Bearden, Guillaume Dumas, Sébastien Jacquemont, P. Bellec
Abstract There is a growing interest in using machine learning (ML) models to perform automatic diagnosis of psychiatric conditions; however, generalising the prediction of ML models to completely independent data can lead to sharp decrease in performance. Patients with different psychiatric diagnoses have traditionally been studied independently, yet there is a growing recognition of neuroimaging signatures shared across them as well as rare genetic copy number variants (CNVs). In this work, we assess the potential of multi-task learning (MTL) to improve accuracy by characterising multiple related conditions with a single model, making use of information shared across diagnostic categories and exposing the model to a larger and more diverse dataset. As a proof of concept, we first established the efficacy of MTL in a context where there is clearly information shared across tasks: the same target (age or sex) is predicted at different sites of data collection in a large functional magnetic resonance imaging (fMRI) dataset compiled from multiple studies. MTL generally led to substantial gains relative to independent prediction at each site. Performing scaling experiments on the UK Biobank, we observed that performance was highly dependent on sample size: for large sample sizes (N > 6000) sex prediction was better using MTL across three sites (N = K per site) than prediction at a single site (N = 3K), but for small samples (N < 500) MTL was actually detrimental for age prediction. We then used established machine-learning methods to benchmark the diagnostic accuracy of each of the 7 CNVs (N = 19–103) and 4 psychiatric conditions (N = 44–472) independently, replicating the accuracy previously reported in the literature on psychiatric conditions. We observed that MTL hurt performance when applied across the full set of diagnoses, and complementary analyses failed to identify pairs of conditions which would benefit from MTL. Taken together, our results show that if a successful multi-task diagnostic model of psychiatric conditions were to be developed with resting-state fMRI, it would likely require datasets with thousands of patients across different diagnoses.
{"title":"Challenges in multi-task learning for fMRI-based diagnosis: Benefits for psychiatric conditions and CNVs would likely require thousands of patients","authors":"A. Harvey, Clara A Moreau, K. Kumar, Guillaume Huguet, S. Urchs, H. Sharmarke, K. Jizi, Charles-Olivier Martin, N. Younis, P. Tamer, J.-L. Martineau, P. Orban, Ana Isabel Silva, Jeremy Hall, M. B. van den Bree, Michael J. Owen, David E J Linden, Sarah Lippé, C. Bearden, Guillaume Dumas, Sébastien Jacquemont, P. Bellec","doi":"10.1162/imag_a_00222","DOIUrl":"https://doi.org/10.1162/imag_a_00222","url":null,"abstract":"Abstract There is a growing interest in using machine learning (ML) models to perform automatic diagnosis of psychiatric conditions; however, generalising the prediction of ML models to completely independent data can lead to sharp decrease in performance. Patients with different psychiatric diagnoses have traditionally been studied independently, yet there is a growing recognition of neuroimaging signatures shared across them as well as rare genetic copy number variants (CNVs). In this work, we assess the potential of multi-task learning (MTL) to improve accuracy by characterising multiple related conditions with a single model, making use of information shared across diagnostic categories and exposing the model to a larger and more diverse dataset. As a proof of concept, we first established the efficacy of MTL in a context where there is clearly information shared across tasks: the same target (age or sex) is predicted at different sites of data collection in a large functional magnetic resonance imaging (fMRI) dataset compiled from multiple studies. MTL generally led to substantial gains relative to independent prediction at each site. Performing scaling experiments on the UK Biobank, we observed that performance was highly dependent on sample size: for large sample sizes (N > 6000) sex prediction was better using MTL across three sites (N = K per site) than prediction at a single site (N = 3K), but for small samples (N < 500) MTL was actually detrimental for age prediction. We then used established machine-learning methods to benchmark the diagnostic accuracy of each of the 7 CNVs (N = 19–103) and 4 psychiatric conditions (N = 44–472) independently, replicating the accuracy previously reported in the literature on psychiatric conditions. We observed that MTL hurt performance when applied across the full set of diagnoses, and complementary analyses failed to identify pairs of conditions which would benefit from MTL. Taken together, our results show that if a successful multi-task diagnostic model of psychiatric conditions were to be developed with resting-state fMRI, it would likely require datasets with thousands of patients across different diagnoses.","PeriodicalId":507939,"journal":{"name":"Imaging Neuroscience","volume":"12 6","pages":"1-20"},"PeriodicalIF":0.0,"publicationDate":"2024-07-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"141838421","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Sir-Lord Wiafe, A. Faghiri, Z. Fu, Robyn L. Miller, Adrian Preda, V. Calhoun
Abstract In neuroimaging research, understanding the intricate dynamics of brain networks over time is paramount for unraveling the complexities of brain function. One approach commonly used to explore the dynamic nature of brain networks is functional connectivity analysis. However, while functional connectivity offers valuable insights, it fails to consider the diverse timescales of coupling between different brain regions. This gap in understanding leaves a significant aspect of brain dynamics unexplored in neuroimaging research. We propose an innovative approach that delves into the dynamic coupling/connectivity timescales of brain regions relative to one another, focusing on how brain region couplings stretch or shrink over time, rather than relying solely on functional connectivity measures. Our method introduces a novel metric called “warping elasticity,” which utilizes dynamic time warping (DTW) to capture the temporal nuances of connectivity. Unlike traditional methods, our approach allows for (potentially nonlinear) dynamic compression and expansion of the time series, offering a more intricate understanding of how coupling between brain regions evolves. Through the adaptive windows employed by the DTW method, we can effectively capture transient couplings within varying connectivity timescales of brain network pairs. In extensive evaluations, our method exhibits high replicability across subjects and diverse datasets, showcasing robustness against noise. More importantly, it uncovers statistically significant distinctions between healthy control (HC) and schizophrenia (SZ) groups through the identification of warp elasticity states. These states are cluster centroids, representing the warp elasticity across subjects and time, offering a novel perspective on the dynamic nature of brain connectivity, distinct from conventional metrics focused solely on functional connectivity. For instance, controls spend more time in a warp elasticity state characterized by timescale stretching of the visual domain relative to other domains, suggesting disruptions in the visual cortex. Conversely, patients show increased time spent in a warp elasticity state with stretching timescales in higher cognitive areas relative to sensory regions, indicative of prolonged cognitive processing of sensory input. Overall, our approach presents a promising avenue for investigating the temporal dynamics of brain network interactions in functional magnetic resonance imaging (fMRI) data. By focusing on the elasticity of connectivity timescales, rather than adhering to functional connectivity metrics, we pave the way for a deeper understanding of neuropsychiatric disorders in neuroscience research.
{"title":"The dynamics of dynamic time warping in fMRI data: A method to capture inter-network stretching and shrinking via warp elasticity","authors":"Sir-Lord Wiafe, A. Faghiri, Z. Fu, Robyn L. Miller, Adrian Preda, V. Calhoun","doi":"10.1162/imag_a_00187","DOIUrl":"https://doi.org/10.1162/imag_a_00187","url":null,"abstract":"Abstract In neuroimaging research, understanding the intricate dynamics of brain networks over time is paramount for unraveling the complexities of brain function. One approach commonly used to explore the dynamic nature of brain networks is functional connectivity analysis. However, while functional connectivity offers valuable insights, it fails to consider the diverse timescales of coupling between different brain regions. This gap in understanding leaves a significant aspect of brain dynamics unexplored in neuroimaging research. We propose an innovative approach that delves into the dynamic coupling/connectivity timescales of brain regions relative to one another, focusing on how brain region couplings stretch or shrink over time, rather than relying solely on functional connectivity measures. Our method introduces a novel metric called “warping elasticity,” which utilizes dynamic time warping (DTW) to capture the temporal nuances of connectivity. Unlike traditional methods, our approach allows for (potentially nonlinear) dynamic compression and expansion of the time series, offering a more intricate understanding of how coupling between brain regions evolves. Through the adaptive windows employed by the DTW method, we can effectively capture transient couplings within varying connectivity timescales of brain network pairs. In extensive evaluations, our method exhibits high replicability across subjects and diverse datasets, showcasing robustness against noise. More importantly, it uncovers statistically significant distinctions between healthy control (HC) and schizophrenia (SZ) groups through the identification of warp elasticity states. These states are cluster centroids, representing the warp elasticity across subjects and time, offering a novel perspective on the dynamic nature of brain connectivity, distinct from conventional metrics focused solely on functional connectivity. For instance, controls spend more time in a warp elasticity state characterized by timescale stretching of the visual domain relative to other domains, suggesting disruptions in the visual cortex. Conversely, patients show increased time spent in a warp elasticity state with stretching timescales in higher cognitive areas relative to sensory regions, indicative of prolonged cognitive processing of sensory input. Overall, our approach presents a promising avenue for investigating the temporal dynamics of brain network interactions in functional magnetic resonance imaging (fMRI) data. By focusing on the elasticity of connectivity timescales, rather than adhering to functional connectivity metrics, we pave the way for a deeper understanding of neuropsychiatric disorders in neuroscience research.","PeriodicalId":507939,"journal":{"name":"Imaging Neuroscience","volume":"38 6","pages":"1-23"},"PeriodicalIF":0.0,"publicationDate":"2024-06-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"141406886","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Onno P. Kampman, Joe Ziminski, S. Afyouni, Mark van der Wilk, Zoe Kourtzi
Abstract We investigate the utility of Wishart processes (WPs) for estimating time-varying functional connectivity (TVFC), which is a measure of changes in functional coupling as the correlation between brain region activity in functional magnetic resonance imaging (fMRI). The WP is a stochastic process on covariance matrices that can model dynamic covariances between time series, which makes it a natural fit to this task. Recent advances in scalable approximate inference techniques and the availability of robust open-source libraries have rendered the WP practically viable for fMRI applications. We introduce a comprehensive benchmarking framework to assess WP performance compared with a selection of established TVFC estimation methods. The framework comprises simulations with specified ground-truth covariance structures, a subject phenotype prediction task, a test-retest study, a brain state analysis, an external stimulus prediction task, and a novel data-driven imputation benchmark. The WP performed competitively across all the benchmarks. It outperformed a sliding window (SW) approach with adaptive cross-validated window lengths and a dynamic conditional correlation (DCC)-multivariate generalized autoregressive conditional heteroskedasticity (MGARCH) baseline on the external stimulus prediction task, while being less prone to false positives in the TVFC null models.
{"title":"Time-varying functional connectivity as Wishart processes","authors":"Onno P. Kampman, Joe Ziminski, S. Afyouni, Mark van der Wilk, Zoe Kourtzi","doi":"10.1162/imag_a_00184","DOIUrl":"https://doi.org/10.1162/imag_a_00184","url":null,"abstract":"Abstract We investigate the utility of Wishart processes (WPs) for estimating time-varying functional connectivity (TVFC), which is a measure of changes in functional coupling as the correlation between brain region activity in functional magnetic resonance imaging (fMRI). The WP is a stochastic process on covariance matrices that can model dynamic covariances between time series, which makes it a natural fit to this task. Recent advances in scalable approximate inference techniques and the availability of robust open-source libraries have rendered the WP practically viable for fMRI applications. We introduce a comprehensive benchmarking framework to assess WP performance compared with a selection of established TVFC estimation methods. The framework comprises simulations with specified ground-truth covariance structures, a subject phenotype prediction task, a test-retest study, a brain state analysis, an external stimulus prediction task, and a novel data-driven imputation benchmark. The WP performed competitively across all the benchmarks. It outperformed a sliding window (SW) approach with adaptive cross-validated window lengths and a dynamic conditional correlation (DCC)-multivariate generalized autoregressive conditional heteroskedasticity (MGARCH) baseline on the external stimulus prediction task, while being less prone to false positives in the TVFC null models.","PeriodicalId":507939,"journal":{"name":"Imaging Neuroscience","volume":"31 10","pages":"1-28"},"PeriodicalIF":0.0,"publicationDate":"2024-06-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"141274312","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Silvia Federica Cotroneo, Heidi Ala-Salomäki, L. Parkkonen, Mia Liljeström, R. Salmelin
Abstract Reliable individual-level measures of neural activity are essential for capturing interindividual variability in brain activity recorded by magnetoencephalography (MEG). While conventional group-level analyses highlight shared features in the data, individual-level specificity is often lost. Current methods for assessing reproducibility of brain responses focus on group-level statistics and neglect subject-specific temporal and spatial characteristics. This study proposes a combined ICA algorithm (comICA), aimed at extracting within-individual consistent MEG evoked responses. The proposed hypotheses behind comICA are based on the temporal profiles of the evoked responses, the corresponding spatial information, as well as independence and linearity. ComICA is presented and tested against simulated data and test–retest recordings of a high-level cognitive task (picture naming). The results show high reliability in extracting the shared activations in the simulations (success rate >93%) and the ability to successfully reproduce group-level results on reproducibility for the test–retest MEG recordings. Our model offers means for noise reduction, targeted extraction of specific activation components in experimental designs, and potential integration across different recordings.
摘要 可靠的神经活动个体水平测量对于捕捉脑磁图（MEG）记录的大脑活动的个体间变异性至关重要。虽然传统的群体水平分析突出了数据中的共同特征，但个体水平的特异性往往会丢失。目前评估大脑反应可重复性的方法侧重于群体层面的统计，而忽略了受试者特定的时间和空间特征。本研究提出了一种组合 ICA 算法（comICA），旨在提取个体内部一致的 MEG 诱发反应。comICA 背后的假设基于诱发反应的时间轮廓、相应的空间信息以及独立性和线性。ComICA 是通过模拟数据和高级认知任务（图片命名）的重复测试记录来呈现和测试的。结果表明，在模拟数据中提取共享激活的可靠性很高（成功率大于 93%），并能成功再现测试-重复测试 MEG 记录的组级再现结果。我们的模型为降低噪音、有针对性地提取实验设计中的特定激活成分以及跨不同记录的潜在整合提供了方法。
{"title":"Extracting reproducible subject-specific MEG evoked responses with independent component analysis","authors":"Silvia Federica Cotroneo, Heidi Ala-Salomäki, L. Parkkonen, Mia Liljeström, R. Salmelin","doi":"10.1162/imag_a_00182","DOIUrl":"https://doi.org/10.1162/imag_a_00182","url":null,"abstract":"Abstract Reliable individual-level measures of neural activity are essential for capturing interindividual variability in brain activity recorded by magnetoencephalography (MEG). While conventional group-level analyses highlight shared features in the data, individual-level specificity is often lost. Current methods for assessing reproducibility of brain responses focus on group-level statistics and neglect subject-specific temporal and spatial characteristics. This study proposes a combined ICA algorithm (comICA), aimed at extracting within-individual consistent MEG evoked responses. The proposed hypotheses behind comICA are based on the temporal profiles of the evoked responses, the corresponding spatial information, as well as independence and linearity. ComICA is presented and tested against simulated data and test–retest recordings of a high-level cognitive task (picture naming). The results show high reliability in extracting the shared activations in the simulations (success rate >93%) and the ability to successfully reproduce group-level results on reproducibility for the test–retest MEG recordings. Our model offers means for noise reduction, targeted extraction of specific activation components in experimental designs, and potential integration across different recordings.","PeriodicalId":507939,"journal":{"name":"Imaging Neuroscience","volume":"76 9","pages":"1-13"},"PeriodicalIF":0.0,"publicationDate":"2024-06-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"141280651","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Colleen Pappas, Christopher E. Bauer, Valentinos Zachariou, Pauline Maillard, Arvind Caprihan, Xingfeng Shao, Danny J.J. Wang, Brian T. Gold
Abstract Vascular risk factors contribute to cognitive aging, with one such risk factor being dysfunction of the blood brain barrier (BBB). Studies using non-invasive magnetic resonance imaging (MRI) techniques, such as diffusion prepared arterial spin labeling (DP-ASL), can estimate BBB function by measuring water exchange rate (kw). DP-ASL kw has been associated with cognition, but the directionality and strength of the relationship is still under investigation. An additional variable that measures water in extracellular space and impacts cognition, MRI free water (FW), may help explain prior findings. A total of 94 older adults without dementia (Mean age = 74.17 years, 59.6% female) underwent MRI (DP-ASL, diffusion weighted imaging (DWI)) and cognitive assessment. Mean kw was computed across the whole brain (WB), and mean white matter FW was computed across all white matter. The relationship between kw and three cognitive domains (executive function, processing speed, memory) was tested using multiple linear regression. FW was tested as a mediator of the kw-cognitive relationship using the PROCESS macro. A positive association was found between WB kw and executive function [F(4,85) = 7.81, p < .001, R2= 0.269; β = .245, p = .014]. Further, this effect was qualified by subsequent results showing that FW was a mediator of the WB kw-executive function relationship (indirect effect results: standardized effect = .060, bootstrap confidence interval = .0006 to .1411). Results suggest that lower water exchange rate (kw) may contribute to greater total white matter (WM) FW which, in turn, may disrupt executive function. Taken together, proper fluid clearance at the BBB contributes to higher-order cognitive abilities.
{"title":"MRI free water mediates the association between water exchange rate across the blood brain barrier and executive function among older adults","authors":"Colleen Pappas, Christopher E. Bauer, Valentinos Zachariou, Pauline Maillard, Arvind Caprihan, Xingfeng Shao, Danny J.J. Wang, Brian T. Gold","doi":"10.1162/imag_a_00183","DOIUrl":"https://doi.org/10.1162/imag_a_00183","url":null,"abstract":"Abstract Vascular risk factors contribute to cognitive aging, with one such risk factor being dysfunction of the blood brain barrier (BBB). Studies using non-invasive magnetic resonance imaging (MRI) techniques, such as diffusion prepared arterial spin labeling (DP-ASL), can estimate BBB function by measuring water exchange rate (kw). DP-ASL kw has been associated with cognition, but the directionality and strength of the relationship is still under investigation. An additional variable that measures water in extracellular space and impacts cognition, MRI free water (FW), may help explain prior findings. A total of 94 older adults without dementia (Mean age = 74.17 years, 59.6% female) underwent MRI (DP-ASL, diffusion weighted imaging (DWI)) and cognitive assessment. Mean kw was computed across the whole brain (WB), and mean white matter FW was computed across all white matter. The relationship between kw and three cognitive domains (executive function, processing speed, memory) was tested using multiple linear regression. FW was tested as a mediator of the kw-cognitive relationship using the PROCESS macro. A positive association was found between WB kw and executive function [F(4,85) = 7.81, p < .001, R2= 0.269; β = .245, p = .014]. Further, this effect was qualified by subsequent results showing that FW was a mediator of the WB kw-executive function relationship (indirect effect results: standardized effect = .060, bootstrap confidence interval = .0006 to .1411). Results suggest that lower water exchange rate (kw) may contribute to greater total white matter (WM) FW which, in turn, may disrupt executive function. Taken together, proper fluid clearance at the BBB contributes to higher-order cognitive abilities.","PeriodicalId":507939,"journal":{"name":"Imaging Neuroscience","volume":"47 16","pages":"1-15"},"PeriodicalIF":0.0,"publicationDate":"2024-06-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"141275610","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
V. V. Nair, Tyler C. Diorio, Qiuting Wen, Vitaliy L. Rayz, Yunjie Tong
Abstract With growing evidence signifying the impact of cerebrospinal fluid (CSF) flow in facilitating waste clearance from the brain and potential pathophysiological links to neurodegenerative disorders, it is of vital importance to develop effective methods to modulate CSF flow in the brain. Here, we attempt this by means of simple commonly used respiratory challenges—paced breathing and breath holding. Functional Magnetic Resonance Imaging scans of the brain and neck respectively were used to record the craniad and caudad CSF movements at the fourth ventricle from eight healthy volunteers during paced breathing and breath holding. Further, we utilized a novel approach for the first time to combine these separately acquired unidirectional CSF movement signals to compare the CSF flow in both directions (in the fourth ventricle) with the respiratory stimuli as a physiological control. Our results demonstrate that these respiratory challenges enhance the magnitude as well as control the direction of CSF movement in the fourth ventricle. They also reveal the capability of blood CO2 concentration changes (induced by respiratory challenges) in the low-frequency range to bring about these CSF movement modulations. Finally, we also successfully report our novel approach where we use these breathing challenges as a unique control condition to detect the small net CSF flows from independently captured unidirectional signals.
{"title":"Using respiratory challenges to modulate CSF movement across different physiological pathways: An fMRI study","authors":"V. V. Nair, Tyler C. Diorio, Qiuting Wen, Vitaliy L. Rayz, Yunjie Tong","doi":"10.1162/imag_a_00192","DOIUrl":"https://doi.org/10.1162/imag_a_00192","url":null,"abstract":"Abstract With growing evidence signifying the impact of cerebrospinal fluid (CSF) flow in facilitating waste clearance from the brain and potential pathophysiological links to neurodegenerative disorders, it is of vital importance to develop effective methods to modulate CSF flow in the brain. Here, we attempt this by means of simple commonly used respiratory challenges—paced breathing and breath holding. Functional Magnetic Resonance Imaging scans of the brain and neck respectively were used to record the craniad and caudad CSF movements at the fourth ventricle from eight healthy volunteers during paced breathing and breath holding. Further, we utilized a novel approach for the first time to combine these separately acquired unidirectional CSF movement signals to compare the CSF flow in both directions (in the fourth ventricle) with the respiratory stimuli as a physiological control. Our results demonstrate that these respiratory challenges enhance the magnitude as well as control the direction of CSF movement in the fourth ventricle. They also reveal the capability of blood CO2 concentration changes (induced by respiratory challenges) in the low-frequency range to bring about these CSF movement modulations. Finally, we also successfully report our novel approach where we use these breathing challenges as a unique control condition to detect the small net CSF flows from independently captured unidirectional signals.","PeriodicalId":507939,"journal":{"name":"Imaging Neuroscience","volume":"8 8","pages":"1-14"},"PeriodicalIF":0.0,"publicationDate":"2024-06-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"141409401","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Ekaterina Shatalina, E. Onwordi, T. Whitehurst, Alex Whittington, A. Mansur, A. Arumuham, B. Statton, A. Berry, T. R. Marques, Roger N. Gunn, S. Natesan, Matthew M. Nour, E. Rabiner, Matthew B Wall, Oliver D. Howes
Abstract Synaptic terminal density is thought to influence cognitive function and neural activity, yet its role in cognition has not been explored in healthy humans. We examined these relationships using [11C]UCB-J positron emission tomography (PET) and functional magnetic resonance imaging (fMRI) in 25 healthy adults performing cognitive function tasks in the scanner. We found a significant positive association between synaptic terminal density, indicated by [11C]UCB-J PET distribution volume ratio (DVRcs), and neural activity during task switching (PLS-CA, second canonical component, r = 0.63, p = 0.043) with the thalamus-putamen data positively contributing to this relationship (PLS-CA loading 0.679, exploratory Pearson’s correlation r = 0.42, p = 0.044, uncorrected). Furthermore, synaptic terminal density predicted switch cost (PLS-R, R2 = 0.45, RMSE = 0.06, p = 0.022), with DVRcs negatively correlating with switch cost in key brain regions including the dorsolateral prefrontal cortex and posterior frontal cortex. Conversely, no significant relationships were observed between [11C]UCB-J DVRcs and neural activity or performance measures in the N-back working memory task, suggesting interindividual differences in synaptic terminal density may be more closely related to some cognitive functions and not others.
{"title":"The relationship between SV2A levels, neural activity, and cognitive function in healthy humans: A [11C]UCB-J PET and fMRI study","authors":"Ekaterina Shatalina, E. Onwordi, T. Whitehurst, Alex Whittington, A. Mansur, A. Arumuham, B. Statton, A. Berry, T. R. Marques, Roger N. Gunn, S. Natesan, Matthew M. Nour, E. Rabiner, Matthew B Wall, Oliver D. Howes","doi":"10.1162/imag_a_00190","DOIUrl":"https://doi.org/10.1162/imag_a_00190","url":null,"abstract":"Abstract Synaptic terminal density is thought to influence cognitive function and neural activity, yet its role in cognition has not been explored in healthy humans. We examined these relationships using [11C]UCB-J positron emission tomography (PET) and functional magnetic resonance imaging (fMRI) in 25 healthy adults performing cognitive function tasks in the scanner. We found a significant positive association between synaptic terminal density, indicated by [11C]UCB-J PET distribution volume ratio (DVRcs), and neural activity during task switching (PLS-CA, second canonical component, r = 0.63, p = 0.043) with the thalamus-putamen data positively contributing to this relationship (PLS-CA loading 0.679, exploratory Pearson’s correlation r = 0.42, p = 0.044, uncorrected). Furthermore, synaptic terminal density predicted switch cost (PLS-R, R2 = 0.45, RMSE = 0.06, p = 0.022), with DVRcs negatively correlating with switch cost in key brain regions including the dorsolateral prefrontal cortex and posterior frontal cortex. Conversely, no significant relationships were observed between [11C]UCB-J DVRcs and neural activity or performance measures in the N-back working memory task, suggesting interindividual differences in synaptic terminal density may be more closely related to some cognitive functions and not others.","PeriodicalId":507939,"journal":{"name":"Imaging Neuroscience","volume":"137 S239","pages":"1-16"},"PeriodicalIF":0.0,"publicationDate":"2024-06-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"141413762","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Huili Sun, M. Rosenblatt, J. Dadashkarimi, Raimundo Rodriguez, Link Tejavibulya, Dustin Scheinost
Abstract Network control theory models how gray matter regions transition between cognitive states through associated white matter connections, where controllability quantifies the contribution of each region to driving these state transitions. Current applications predominantly adopt node-centric views and overlook the potential contribution of brain network connections. To bridge this gap, we use edge-centric network control theory (E-NCT) to assess the role of brain connectivity (i.e., edges) in governing brain dynamic processes. We applied this framework to diffusion MRI data from individuals in the Human Connectome Project. We first validate edge controllability through comparisons against null models, node controllability, and structural and functional connectomes. Notably, edge controllability predicted individual differences in phenotypic information. Using E-NCT, we estimate the brain’s energy consumption for activating specific networks. Our results reveal that the activation of a complex, whole-brain network predicting executive function (EF) is more energy efficient than the corresponding canonical network pairs. Overall, E-NCT provides an edge-centric perspective on the brain’s network control mechanism. It captures control energy patterns and brain-behavior phenotypes with a more comprehensive understanding of brain dynamics.
{"title":"Edge-centric network control on the human brain structural network","authors":"Huili Sun, M. Rosenblatt, J. Dadashkarimi, Raimundo Rodriguez, Link Tejavibulya, Dustin Scheinost","doi":"10.1162/imag_a_00191","DOIUrl":"https://doi.org/10.1162/imag_a_00191","url":null,"abstract":"Abstract Network control theory models how gray matter regions transition between cognitive states through associated white matter connections, where controllability quantifies the contribution of each region to driving these state transitions. Current applications predominantly adopt node-centric views and overlook the potential contribution of brain network connections. To bridge this gap, we use edge-centric network control theory (E-NCT) to assess the role of brain connectivity (i.e., edges) in governing brain dynamic processes. We applied this framework to diffusion MRI data from individuals in the Human Connectome Project. We first validate edge controllability through comparisons against null models, node controllability, and structural and functional connectomes. Notably, edge controllability predicted individual differences in phenotypic information. Using E-NCT, we estimate the brain’s energy consumption for activating specific networks. Our results reveal that the activation of a complex, whole-brain network predicting executive function (EF) is more energy efficient than the corresponding canonical network pairs. Overall, E-NCT provides an edge-centric perspective on the brain’s network control mechanism. It captures control energy patterns and brain-behavior phenotypes with a more comprehensive understanding of brain dynamics.","PeriodicalId":507939,"journal":{"name":"Imaging Neuroscience","volume":"31 1","pages":"1-15"},"PeriodicalIF":0.0,"publicationDate":"2024-06-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"141399838","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Nathan Blanke, Alexander J. Gray, Rhiannon Robinson, A. Novoseltseva, D. Rosene, Irving J. Bigio
Abstract Despite the interest in studying and quantifying the structural integrity of myelin in postmortem brain tissue, current methods for high-resolution imaging of myelin with optical microscopy are not sufficient. While imaging methods must have adequate resolution and sensitivity to detect microstructural alterations to myelin that are relevant in aging and neurodegenerative disease, an equally critical aspect is to minimize myelin damage that is induced during tissue processing steps. Birefringence microscopy (BRM) is a powerful technique that leverages the structural anisotropy of myelin to provide detailed, label-free images of myelin at any diffraction-limited optical resolution, while maintaining a simple and low-cost setup. Building on our previous work, we have developed a new BRM system and image processing pipeline that enable efficient, high-throughput imaging of myelin structure at multiple scales. Here, we utilize this system to systematically assess the damage to myelin that is induced by several common tissue processing steps in brain sections from the rhesus monkey. Images taken of the same myelinated axons, before and after each tissue processing step, provide direct evidence that mishandling of tissue during sample preparation can cause significant structural alterations to myelin. First, we report on key advancements to our BRM system, imaging procedure, and image processing pipeline, which provide significant increases to the speed and efficiency of BRM. These include integrating fast piezoelectric rotational stages, minimizing the number of images required (to three images) for determining birefringence parameter maps, and implementing an analytical solution for directly determining birefringence parameter maps. Second, using this BRM system, we demonstrate that effective myelin imaging requires (1) the avoidance of prolonged drying or dehydration of tissue, (2) the selection of the optimal mounting medium (85% glycerol), (3) the avoidance of tissue permeabilization with detergents (i.e., Triton X-100 and Saponin), and (4) the selection of a suitable tissue-section thickness (15, 30 and 60 µm) based on the region of interest. In addition to serving as a guide for new users interested in imaging myelin, these basic experiments in sample preparation highlight that BRM is very sensitive to changes in the underlying lipid structure of myelin and suggest that optimized BRM can enable new studies of myelin breakdown in disease. In this work, we show that BRM is a leading method for detailed imaging and characterization of myelin, and we provide direct evidence that the structure of myelin is highly sensitive to damage during inadequate preparation of brain tissue for imaging, which has previously not been properly characterized for birefringence imaging of myelin. For the most effective, high-resolution imaging of myelin structure, tissue processing should be kept to a minimum, with sections prevented from dehydration and moun
{"title":"Practical considerations for birefringence microscopy of myelin structure: Microscope design and tissue processing for effective imaging","authors":"Nathan Blanke, Alexander J. Gray, Rhiannon Robinson, A. Novoseltseva, D. Rosene, Irving J. Bigio","doi":"10.1162/imag_a_00186","DOIUrl":"https://doi.org/10.1162/imag_a_00186","url":null,"abstract":"Abstract Despite the interest in studying and quantifying the structural integrity of myelin in postmortem brain tissue, current methods for high-resolution imaging of myelin with optical microscopy are not sufficient. While imaging methods must have adequate resolution and sensitivity to detect microstructural alterations to myelin that are relevant in aging and neurodegenerative disease, an equally critical aspect is to minimize myelin damage that is induced during tissue processing steps. Birefringence microscopy (BRM) is a powerful technique that leverages the structural anisotropy of myelin to provide detailed, label-free images of myelin at any diffraction-limited optical resolution, while maintaining a simple and low-cost setup. Building on our previous work, we have developed a new BRM system and image processing pipeline that enable efficient, high-throughput imaging of myelin structure at multiple scales. Here, we utilize this system to systematically assess the damage to myelin that is induced by several common tissue processing steps in brain sections from the rhesus monkey. Images taken of the same myelinated axons, before and after each tissue processing step, provide direct evidence that mishandling of tissue during sample preparation can cause significant structural alterations to myelin. First, we report on key advancements to our BRM system, imaging procedure, and image processing pipeline, which provide significant increases to the speed and efficiency of BRM. These include integrating fast piezoelectric rotational stages, minimizing the number of images required (to three images) for determining birefringence parameter maps, and implementing an analytical solution for directly determining birefringence parameter maps. Second, using this BRM system, we demonstrate that effective myelin imaging requires (1) the avoidance of prolonged drying or dehydration of tissue, (2) the selection of the optimal mounting medium (85% glycerol), (3) the avoidance of tissue permeabilization with detergents (i.e., Triton X-100 and Saponin), and (4) the selection of a suitable tissue-section thickness (15, 30 and 60 µm) based on the region of interest. In addition to serving as a guide for new users interested in imaging myelin, these basic experiments in sample preparation highlight that BRM is very sensitive to changes in the underlying lipid structure of myelin and suggest that optimized BRM can enable new studies of myelin breakdown in disease. In this work, we show that BRM is a leading method for detailed imaging and characterization of myelin, and we provide direct evidence that the structure of myelin is highly sensitive to damage during inadequate preparation of brain tissue for imaging, which has previously not been properly characterized for birefringence imaging of myelin. For the most effective, high-resolution imaging of myelin structure, tissue processing should be kept to a minimum, with sections prevented from dehydration and moun","PeriodicalId":507939,"journal":{"name":"Imaging Neuroscience","volume":"1 4","pages":"1-22"},"PeriodicalIF":0.0,"publicationDate":"2024-06-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"141281020","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
京公网安备 11010802042870号
京ICP备2023020795号-1
扫码关注我们
求助内容：
应助结果提醒方式：