Pub Date : 2024-08-09eCollection Date: 2024-01-01DOI: 10.3389/fninf.2024.1461597
Marvin Kaster, Fabian Czappa, Markus Butz-Ostendorf, Felix Wolf
[This corrects the article DOI: 10.3389/fninf.2024.1323203.].
[此处更正了文章 DOI:10.3389/fninf.2024.1323203.]。
{"title":"Corrigendum: Building a realistic, scalable memory model with independent engrams using a homeostatic mechanism.","authors":"Marvin Kaster, Fabian Czappa, Markus Butz-Ostendorf, Felix Wolf","doi":"10.3389/fninf.2024.1461597","DOIUrl":"https://doi.org/10.3389/fninf.2024.1461597","url":null,"abstract":"<p><p>[This corrects the article DOI: 10.3389/fninf.2024.1323203.].</p>","PeriodicalId":12462,"journal":{"name":"Frontiers in Neuroinformatics","volume":null,"pages":null},"PeriodicalIF":2.5,"publicationDate":"2024-08-09","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC11342446/pdf/","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142055340","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":4,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2024-08-09DOI: 10.3389/fninf.2024.1415512
Roy Cox, Frederik D. Weber, E. V. van Someren
While standard polysomnography has revealed the importance of the sleeping brain in health and disease, more specific insight into the relevant brain circuits requires high-density electroencephalography (EEG). However, identifying and handling sleep EEG artifacts becomes increasingly challenging with higher channel counts and/or volume of recordings. Whereas manual cleaning is time-consuming, subjective, and often yields data loss (e.g., complete removal of channels or epochs), automated approaches suitable and practical for overnight sleep EEG remain limited, especially when control over detection and repair behavior is desired. Here, we introduce a flexible approach for automated cleaning of multichannel sleep recordings, as part of the free Matlab-based toolbox SleepTrip. Key functionality includes 1) channel-wise detection of various artifact types encountered in sleep EEG, 2) channel- and time-resolved marking of data segments for repair through interpolation, and 3) visualization options to review and monitor performance. Functionality for Independent Component Analysis is also included. Extensive customization options allow tailoring cleaning behavior to data properties and analysis goals. By enabling computationally efficient and flexible automated data cleaning, this tool helps to facilitate fundamental and clinical sleep EEG research.
{"title":"Customizable automated cleaning of multichannel sleep EEG in SleepTrip","authors":"Roy Cox, Frederik D. Weber, E. V. van Someren","doi":"10.3389/fninf.2024.1415512","DOIUrl":"https://doi.org/10.3389/fninf.2024.1415512","url":null,"abstract":"While standard polysomnography has revealed the importance of the sleeping brain in health and disease, more specific insight into the relevant brain circuits requires high-density electroencephalography (EEG). However, identifying and handling sleep EEG artifacts becomes increasingly challenging with higher channel counts and/or volume of recordings. Whereas manual cleaning is time-consuming, subjective, and often yields data loss (e.g., complete removal of channels or epochs), automated approaches suitable and practical for overnight sleep EEG remain limited, especially when control over detection and repair behavior is desired. Here, we introduce a flexible approach for automated cleaning of multichannel sleep recordings, as part of the free Matlab-based toolbox SleepTrip. Key functionality includes 1) channel-wise detection of various artifact types encountered in sleep EEG, 2) channel- and time-resolved marking of data segments for repair through interpolation, and 3) visualization options to review and monitor performance. Functionality for Independent Component Analysis is also included. Extensive customization options allow tailoring cleaning behavior to data properties and analysis goals. By enabling computationally efficient and flexible automated data cleaning, this tool helps to facilitate fundamental and clinical sleep EEG research.","PeriodicalId":12462,"journal":{"name":"Frontiers in Neuroinformatics","volume":null,"pages":null},"PeriodicalIF":2.5,"publicationDate":"2024-08-09","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"141921600","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":4,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2024-08-05eCollection Date: 2024-01-01DOI: 10.3389/fninf.2024.1454834
Peter A Tass, Hemant Bokil
{"title":"Editorial: Neuromodulation using spatiotemporally complex patterns.","authors":"Peter A Tass, Hemant Bokil","doi":"10.3389/fninf.2024.1454834","DOIUrl":"10.3389/fninf.2024.1454834","url":null,"abstract":"","PeriodicalId":12462,"journal":{"name":"Frontiers in Neuroinformatics","volume":null,"pages":null},"PeriodicalIF":2.5,"publicationDate":"2024-08-05","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC11334158/pdf/","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142008613","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":4,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2024-07-31DOI: 10.3389/fninf.2024.1354708
Sidsel Winther, Oscar Peulicke, Mariam Andersson, Hans M. Kjer, Jakob A. Bærentzen, Tim B. Dyrby
Brain white matter is a dynamic environment that continuously adapts and reorganizes in response to stimuli and pathological changes. Glial cells, especially, play a key role in tissue repair, inflammation modulation, and neural recovery. The movements of glial cells and changes in their concentrations can influence the surrounding axon morphology. We introduce the White Matter Generator (WMG) tool to enable the study of how axon morphology is influenced through such dynamical processes, and how this, in turn, influences the diffusion-weighted MRI signal. This is made possible by allowing interactive changes to the configuration of the phantom generation throughout the optimization process. The phantoms can consist of myelinated axons, unmyelinated axons, and cell clusters, separated by extra-cellular space. Due to morphological flexibility and computational advantages during the optimization, the tool uses ellipsoids as building blocks for all structures; chains of ellipsoids for axons, and individual ellipsoids for cell clusters. After optimization, the ellipsoid representation can be converted to a mesh representation which can be employed in Monte-Carlo diffusion simulations. This offers an effective method for evaluating tissue microstructure models for diffusion-weighted MRI in controlled bio-mimicking white matter environments. Hence, the WMG offers valuable insights into white matter's adaptive nature and implications for diffusion-weighted MRI microstructure models, and thereby holds the potential to advance clinical diagnosis, treatment, and rehabilitation strategies for various neurological disorders and injuries.
{"title":"Exploring white matter dynamics and morphology through interactive numerical phantoms: the White Matter Generator","authors":"Sidsel Winther, Oscar Peulicke, Mariam Andersson, Hans M. Kjer, Jakob A. Bærentzen, Tim B. Dyrby","doi":"10.3389/fninf.2024.1354708","DOIUrl":"https://doi.org/10.3389/fninf.2024.1354708","url":null,"abstract":"Brain white matter is a dynamic environment that continuously adapts and reorganizes in response to stimuli and pathological changes. Glial cells, especially, play a key role in tissue repair, inflammation modulation, and neural recovery. The movements of glial cells and changes in their concentrations can influence the surrounding axon morphology. We introduce the White Matter Generator (WMG) tool to enable the study of how axon morphology is influenced through such dynamical processes, and how this, in turn, influences the diffusion-weighted MRI signal. This is made possible by allowing interactive changes to the configuration of the phantom generation throughout the optimization process. The phantoms can consist of myelinated axons, unmyelinated axons, and cell clusters, separated by extra-cellular space. Due to morphological flexibility and computational advantages during the optimization, the tool uses ellipsoids as building blocks for all structures; chains of ellipsoids for axons, and individual ellipsoids for cell clusters. After optimization, the ellipsoid representation can be converted to a mesh representation which can be employed in Monte-Carlo diffusion simulations. This offers an effective method for evaluating tissue microstructure models for diffusion-weighted MRI in controlled bio-mimicking white matter environments. Hence, the WMG offers valuable insights into white matter's adaptive nature and implications for diffusion-weighted MRI microstructure models, and thereby holds the potential to advance clinical diagnosis, treatment, and rehabilitation strategies for various neurological disorders and injuries.","PeriodicalId":12462,"journal":{"name":"Frontiers in Neuroinformatics","volume":null,"pages":null},"PeriodicalIF":3.5,"publicationDate":"2024-07-31","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"141863049","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":4,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2024-07-29DOI: 10.3389/fninf.2024.1429670
Kai-Yi Hsu, Chi-Tin Shih, Nan-Yow Chen, Chung-Chuan Lo
The brain atlas, which provides information about the distribution of genes, proteins, neurons, or anatomical regions, plays a crucial role in contemporary neuroscience research. To analyze the spatial distribution of those substances based on images from different brain samples, we often need to warp and register individual brain images to a standard brain template. However, the process of warping and registration may lead to spatial errors, thereby severely reducing the accuracy of the analysis. To address this issue, we develop an automated method for segmenting neuropils in the Drosophila brain for fluorescence images from the FlyCircuit database. This technique allows future brain atlas studies to be conducted accurately at the individual level without warping and aligning to a standard brain template. Our method, LYNSU (Locating by YOLO and Segmenting by U-Net), consists of two stages. In the first stage, we use the YOLOv7 model to quickly locate neuropils and rapidly extract small-scale 3D images as input for the second stage model. This stage achieves a 99.4% accuracy rate in neuropil localization. In the second stage, we employ the 3D U-Net model to segment neuropils. LYNSU can achieve high accuracy in segmentation using a small training set consisting of images from merely 16 brains. We demonstrate LYNSU on six distinct neuropils or structures, achieving a high segmentation accuracy comparable to professional manual annotations with a 3D Intersection-over-Union (IoU) reaching up to 0.869. Our method takes only about 7 s to segment a neuropil while achieving a similar level of performance as the human annotators. To demonstrate a use case of LYNSU, we applied it to all female Drosophila brains from the FlyCircuit database to investigate the asymmetry of the mushroom bodies (MBs), the learning center of fruit flies. We used LYNSU to segment bilateral MBs and compare the volumes between left and right for each individual. Notably, of 8,703 valid brain samples, 10.14% showed bilateral volume differences that exceeded 10%. The study demonstrated the potential of the proposed method in high-throughput anatomical analysis and connectomics construction of the Drosophila brain.
{"title":"LYNSU: automated 3D neuropil segmentation of fluorescent images for Drosophila brains","authors":"Kai-Yi Hsu, Chi-Tin Shih, Nan-Yow Chen, Chung-Chuan Lo","doi":"10.3389/fninf.2024.1429670","DOIUrl":"https://doi.org/10.3389/fninf.2024.1429670","url":null,"abstract":"The brain atlas, which provides information about the distribution of genes, proteins, neurons, or anatomical regions, plays a crucial role in contemporary neuroscience research. To analyze the spatial distribution of those substances based on images from different brain samples, we often need to warp and register individual brain images to a standard brain template. However, the process of warping and registration may lead to spatial errors, thereby severely reducing the accuracy of the analysis. To address this issue, we develop an automated method for segmenting neuropils in the <jats:italic>Drosophila</jats:italic> brain for fluorescence images from the <jats:italic>FlyCircuit</jats:italic> database. This technique allows future brain atlas studies to be conducted accurately at the individual level without warping and aligning to a standard brain template. Our method, LYNSU (Locating by YOLO and Segmenting by U-Net), consists of two stages. In the first stage, we use the YOLOv7 model to quickly locate neuropils and rapidly extract small-scale 3D images as input for the second stage model. This stage achieves a 99.4% accuracy rate in neuropil localization. In the second stage, we employ the 3D U-Net model to segment neuropils. LYNSU can achieve high accuracy in segmentation using a small training set consisting of images from merely 16 brains. We demonstrate LYNSU on six distinct neuropils or structures, achieving a high segmentation accuracy comparable to professional manual annotations with a 3D Intersection-over-Union (IoU) reaching up to 0.869. Our method takes only about 7 s to segment a neuropil while achieving a similar level of performance as the human annotators. To demonstrate a use case of LYNSU, we applied it to all female <jats:italic>Drosophila</jats:italic> brains from the <jats:italic>FlyCircuit</jats:italic> database to investigate the asymmetry of the mushroom bodies (MBs), the learning center of fruit flies. We used LYNSU to segment bilateral MBs and compare the volumes between left and right for each individual. Notably, of 8,703 valid brain samples, 10.14% showed bilateral volume differences that exceeded 10%. The study demonstrated the potential of the proposed method in high-throughput anatomical analysis and connectomics construction of the <jats:italic>Drosophila</jats:italic> brain.","PeriodicalId":12462,"journal":{"name":"Frontiers in Neuroinformatics","volume":null,"pages":null},"PeriodicalIF":3.5,"publicationDate":"2024-07-29","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"141863050","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":4,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
IntroductionBrain diseases, particularly the classification of gliomas and brain metastases and the prediction of HT in strokes, pose significant challenges in healthcare. Existing methods, relying predominantly on clinical data or imaging-based techniques such as radiomics, often fall short in achieving satisfactory classification accuracy. These methods fail to adequately capture the nuanced features crucial for accurate diagnosis, often hindered by noise and the inability to integrate information across various scales.MethodsWe propose a novel approach that mask attention mechanisms with multi-scale feature fusion for Multimodal brain disease classification tasks, termed M3, which aims to extract features highly relevant to the disease. The extracted features are then dimensionally reduced using Principal Component Analysis (PCA), followed by classification with a Support Vector Machine (SVM) to obtain the predictive results.ResultsOur methodology underwent rigorous testing on multi-parametric MRI datasets for both brain tumors and strokes. The results demonstrate a significant improvement in addressing critical clinical challenges, including the classification of gliomas, brain metastases, and the prediction of hemorrhagic stroke transformations. Ablation studies further validate the effectiveness of our attention mechanism and feature fusion modules.DiscussionThese findings underscore the potential of our approach to meet and exceed current clinical diagnostic demands, offering promising prospects for enhancing healthcare outcomes in the diagnosis and treatment of brain diseases.
{"title":"M3: using mask-attention and multi-scale for multi-modal brain MRI classification","authors":"Guanqing Kong, Chuanfu Wu, Zongqiu Zhang, Chuansheng Yin, Dawei Qin","doi":"10.3389/fninf.2024.1403732","DOIUrl":"https://doi.org/10.3389/fninf.2024.1403732","url":null,"abstract":"IntroductionBrain diseases, particularly the classification of gliomas and brain metastases and the prediction of HT in strokes, pose significant challenges in healthcare. Existing methods, relying predominantly on clinical data or imaging-based techniques such as radiomics, often fall short in achieving satisfactory classification accuracy. These methods fail to adequately capture the nuanced features crucial for accurate diagnosis, often hindered by noise and the inability to integrate information across various scales.MethodsWe propose a novel approach that mask attention mechanisms with multi-scale feature fusion for Multimodal brain disease classification tasks, termed <jats:italic>M</jats:italic><jats:sup>3</jats:sup>, which aims to extract features highly relevant to the disease. The extracted features are then dimensionally reduced using Principal Component Analysis (PCA), followed by classification with a Support Vector Machine (SVM) to obtain the predictive results.ResultsOur methodology underwent rigorous testing on multi-parametric MRI datasets for both brain tumors and strokes. The results demonstrate a significant improvement in addressing critical clinical challenges, including the classification of gliomas, brain metastases, and the prediction of hemorrhagic stroke transformations. Ablation studies further validate the effectiveness of our attention mechanism and feature fusion modules.DiscussionThese findings underscore the potential of our approach to meet and exceed current clinical diagnostic demands, offering promising prospects for enhancing healthcare outcomes in the diagnosis and treatment of brain diseases.","PeriodicalId":12462,"journal":{"name":"Frontiers in Neuroinformatics","volume":null,"pages":null},"PeriodicalIF":3.5,"publicationDate":"2024-07-29","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"141872820","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":4,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2024-07-29DOI: 10.3389/fninf.2024.1399391
Michelle F. Miranda
Task-evoked functional magnetic resonance imaging studies, such as the Human Connectome Project (HCP), are a powerful tool for exploring how brain activity is influenced by cognitive tasks like memory retention, decision-making, and language processing. A fast Bayesian function-on-scalar model is proposed for estimating population-level activation maps linked to the working memory task. The model is based on the canonical polyadic (CP) tensor decomposition of coefficient maps obtained for each subject. This decomposition effectively yields a tensor basis capable of extracting both common features and subject-specific features from the coefficient maps. These subject-specific features, in turn, are modeled as a function of covariates of interest using a Bayesian model that accounts for the correlation of the CP-extracted features. The dimensionality reduction achieved with the tensor basis allows for a fast MCMC estimation of population-level activation maps. This model is applied to one hundred unrelated subjects from the HCP dataset, yielding significant insights into brain signatures associated with working memory.
{"title":"Frontiers | A canonical polyadic tensor basis for fast Bayesian estimation of multi-subject brain activation patterns","authors":"Michelle F. Miranda","doi":"10.3389/fninf.2024.1399391","DOIUrl":"https://doi.org/10.3389/fninf.2024.1399391","url":null,"abstract":"Task-evoked functional magnetic resonance imaging studies, such as the Human Connectome Project (HCP), are a powerful tool for exploring how brain activity is influenced by cognitive tasks like memory retention, decision-making, and language processing. A fast Bayesian function-on-scalar model is proposed for estimating population-level activation maps linked to the working memory task. The model is based on the canonical polyadic (CP) tensor decomposition of coefficient maps obtained for each subject. This decomposition effectively yields a tensor basis capable of extracting both common features and subject-specific features from the coefficient maps. These subject-specific features, in turn, are modeled as a function of covariates of interest using a Bayesian model that accounts for the correlation of the CP-extracted features. The dimensionality reduction achieved with the tensor basis allows for a fast MCMC estimation of population-level activation maps. This model is applied to one hundred unrelated subjects from the HCP dataset, yielding significant insights into brain signatures associated with working memory.","PeriodicalId":12462,"journal":{"name":"Frontiers in Neuroinformatics","volume":null,"pages":null},"PeriodicalIF":3.5,"publicationDate":"2024-07-29","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"141936367","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":4,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2024-07-12DOI: 10.3389/fninf.2024.1387400
Anass B. El-Yaagoubi, Moo K. Chung, Hernando Ombao
Topological data analysis (TDA) is increasingly recognized as a promising tool in the field of neuroscience, unveiling the underlying topological patterns within brain signals. However, most TDA related methods treat brain signals as if they were static, i.e., they ignore potential non-stationarities and irregularities in the statistical properties of the signals. In this study, we develop a novel fractal dimension-based testing approach that takes into account the dynamic topological properties of brain signals. By representing EEG brain signals as a sequence of Vietoris-Rips filtrations, our approach accommodates the inherent non-stationarities and irregularities of the signals. The application of our novel fractal dimension-based testing approach in analyzing dynamic topological patterns in EEG signals during an epileptic seizure episode exposes noteworthy alterations in total persistence across 0, 1, and 2-dimensional homology. These findings imply a more intricate influence of seizures on brain signals, extending beyond mere amplitude changes.
{"title":"Dynamic topological data analysis: a novel fractal dimension-based testing framework with application to brain signals","authors":"Anass B. El-Yaagoubi, Moo K. Chung, Hernando Ombao","doi":"10.3389/fninf.2024.1387400","DOIUrl":"https://doi.org/10.3389/fninf.2024.1387400","url":null,"abstract":"Topological data analysis (TDA) is increasingly recognized as a promising tool in the field of neuroscience, unveiling the underlying topological patterns within brain signals. However, most TDA related methods treat brain signals as if they were static, i.e., they ignore potential non-stationarities and irregularities in the statistical properties of the signals. In this study, we develop a novel fractal dimension-based testing approach that takes into account the dynamic topological properties of brain signals. By representing EEG brain signals as a sequence of Vietoris-Rips filtrations, our approach accommodates the inherent non-stationarities and irregularities of the signals. The application of our novel fractal dimension-based testing approach in analyzing dynamic topological patterns in EEG signals during an epileptic seizure episode exposes noteworthy alterations in total persistence across 0, 1, and 2-dimensional homology. These findings imply a more intricate influence of seizures on brain signals, extending beyond mere amplitude changes.","PeriodicalId":12462,"journal":{"name":"Frontiers in Neuroinformatics","volume":null,"pages":null},"PeriodicalIF":3.5,"publicationDate":"2024-07-12","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"141609793","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":4,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2024-07-08DOI: 10.3389/fninf.2024.1448591
Antonio Fernández-Caballero, Michel Le Van Quyen
{"title":"Editorial: Innovative methods for sleep staging using neuroinformatics","authors":"Antonio Fernández-Caballero, Michel Le Van Quyen","doi":"10.3389/fninf.2024.1448591","DOIUrl":"https://doi.org/10.3389/fninf.2024.1448591","url":null,"abstract":"","PeriodicalId":12462,"journal":{"name":"Frontiers in Neuroinformatics","volume":null,"pages":null},"PeriodicalIF":2.5,"publicationDate":"2024-07-08","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"141669416","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":4,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2024-06-28DOI: 10.3389/fninf.2024.1392661
Km Bhavna, Azman Akhter, Romi Banerjee, Dipanjan Roy
Decoding of cognitive states aims to identify individuals' brain states and brain fingerprints to predict behavior. Deep learning provides an important platform for analyzing brain signals at different developmental stages to understand brain dynamics. Due to their internal architecture and feature extraction techniques, existing machine-learning and deep-learning approaches are suffering from low classification performance and explainability issues that must be improved. In the current study, we hypothesized that even at the early childhood stage (as early as 3-years), connectivity between brain regions could decode brain states and predict behavioral performance in false-belief tasks. To this end, we proposed an explainable deep learning framework to decode brain states (Theory of Mind and Pain states) and predict individual performance on ToM-related false-belief tasks in a developmental dataset. We proposed an explainable spatiotemporal connectivity-based Graph Convolutional Neural Network (Ex-stGCNN) model for decoding brain states. Here, we consider a developmental dataset, N = 155 (122 children; 3–12 yrs and 33 adults; 18–39 yrs), in which participants watched a short, soundless animated movie, shown to activate Theory-of-Mind (ToM) and pain networs. After scanning, the participants underwent a ToM-related false-belief task, leading to categorization into the pass, fail, and inconsistent groups based on performance. We trained our proposed model using Functional Connectivity (FC) and Inter-Subject Functional Correlations (ISFC) matrices separately. We observed that the stimulus-driven feature set (ISFC) could capture ToM and Pain brain states more accurately with an average accuracy of 94%, whereas it achieved 85% accuracy using FC matrices. We also validated our results using five-fold cross-validation and achieved an average accuracy of 92%. Besides this study, we applied the SHapley Additive exPlanations (SHAP) approach to identify brain fingerprints that contributed the most to predictions. We hypothesized that ToM network brain connectivity could predict individual performance on false-belief tasks. We proposed an Explainable Convolutional Variational Auto-Encoder (Ex-Convolutional VAE) model to predict individual performance on false-belief tasks and trained the model using FC and ISFC matrices separately. ISFC matrices again outperformed the FC matrices in prediction of individual performance. We achieved 93.5% accuracy with an F1-score of 0.94 using ISFC matrices and achieved 90% accuracy with an F1-score of 0.91 using FC matrices.
认知状态解码旨在识别个人的大脑状态和大脑指纹,从而预测行为。深度学习为分析不同发育阶段的大脑信号以了解大脑动态提供了一个重要平台。由于其内部架构和特征提取技术的原因,现有的机器学习和深度学习方法存在分类性能低、可解释性差等问题,必须加以改进。在本研究中,我们假设即使在幼儿阶段(早至 3 岁),大脑区域之间的连接性也能解码大脑状态,并预测虚假信念任务中的行为表现。为此,我们提出了一个可解释的深度学习框架,以解码大脑状态(心智理论和疼痛状态),并预测发育数据集中与心智理论相关的虚假信念任务中的个体表现。我们提出了一种可解释的基于时空连接的图卷积神经网络(Ex-stGCNN)模型,用于解码大脑状态。在这里,我们考虑了一个发育数据集,N = 155(122 名儿童;3-12 岁和 33 名成人;18-39 岁),其中参与者观看了一部无声动画短片,影片显示激活了心智理论(ToM)和疼痛网络。扫描结束后,参与者接受了与 ToM 相关的虚假信念任务,根据表现分为通过组、失败组和不一致组。我们使用功能连接(FC)和受试者间功能相关性(ISFC)矩阵分别训练了我们提出的模型。我们观察到,刺激驱动特征集(ISFC)能更准确地捕捉 ToM 和疼痛的大脑状态,平均准确率为 94%,而使用 FC 矩阵的准确率为 85%。我们还使用五倍交叉验证对结果进行了验证,平均准确率达到 92%。除了这项研究,我们还应用了 SHapley Additive exPlanations(SHAP)方法来识别对预测贡献最大的大脑指纹。我们假设 ToM 网络的大脑连通性可以预测个人在错误信念任务中的表现。我们提出了一个可解释卷积变异自动编码器(Ex-Convolutional VAE)模型来预测个人在虚假信念任务中的表现,并分别使用 FC 和 ISFC 矩阵对该模型进行了训练。在预测个人表现方面,ISFC 矩阵的表现再次优于 FC 矩阵。使用 ISFC 矩阵,我们的准确率达到了 93.5%,F1 分数为 0.94;使用 FC 矩阵,我们的准确率达到了 90%,F1 分数为 0.91。
{"title":"Explainable deep-learning framework: decoding brain states and prediction of individual performance in false-belief task at early childhood stage","authors":"Km Bhavna, Azman Akhter, Romi Banerjee, Dipanjan Roy","doi":"10.3389/fninf.2024.1392661","DOIUrl":"https://doi.org/10.3389/fninf.2024.1392661","url":null,"abstract":"Decoding of cognitive states aims to identify individuals' brain states and brain fingerprints to predict behavior. Deep learning provides an important platform for analyzing brain signals at different developmental stages to understand brain dynamics. Due to their internal architecture and feature extraction techniques, existing machine-learning and deep-learning approaches are suffering from low classification performance and explainability issues that must be improved. In the current study, we hypothesized that even at the early childhood stage (as early as 3-years), connectivity between brain regions could decode brain states and predict behavioral performance in false-belief tasks. To this end, we proposed an explainable deep learning framework to decode brain states (Theory of Mind and Pain states) and predict individual performance on ToM-related false-belief tasks in a developmental dataset. We proposed an explainable spatiotemporal connectivity-based Graph Convolutional Neural Network (Ex-stGCNN) model for decoding brain states. Here, we consider a developmental dataset, <jats:italic>N</jats:italic> = 155 (122 children; 3–12 yrs and 33 adults; 18–39 yrs), in which participants watched a short, soundless animated movie, shown to activate Theory-of-Mind (ToM) and pain networs. After scanning, the participants underwent a ToM-related false-belief task, leading to categorization into the pass, fail, and inconsistent groups based on performance. We trained our proposed model using Functional Connectivity (FC) and Inter-Subject Functional Correlations (ISFC) matrices separately. We observed that the stimulus-driven feature set (ISFC) could capture ToM and Pain brain states more accurately with an average accuracy of 94%, whereas it achieved 85% accuracy using FC matrices. We also validated our results using five-fold cross-validation and achieved an average accuracy of 92%. Besides this study, we applied the SHapley Additive exPlanations (SHAP) approach to identify brain fingerprints that contributed the most to predictions. We hypothesized that ToM network brain connectivity could predict individual performance on false-belief tasks. We proposed an Explainable Convolutional Variational Auto-Encoder (Ex-Convolutional VAE) model to predict individual performance on false-belief tasks and trained the model using FC and ISFC matrices separately. ISFC matrices again outperformed the FC matrices in prediction of individual performance. We achieved 93.5% accuracy with an F1-score of 0.94 using ISFC matrices and achieved 90% accuracy with an F1-score of 0.91 using FC matrices.","PeriodicalId":12462,"journal":{"name":"Frontiers in Neuroinformatics","volume":null,"pages":null},"PeriodicalIF":3.5,"publicationDate":"2024-06-28","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"141502400","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":4,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}