Pub Date : 2024-05-16DOI: 10.3389/fncom.2024.1240348
Kyle Daruwalla, Mikko Lipasti
Deep neural feedforward networks are effective models for a wide array of problems, but training and deploying such networks presents a significant energy cost. Spiking neural networks (SNNs), which are modeled after biologically realistic neurons, offer a potential solution when deployed correctly on neuromorphic computing hardware. Still, many applications train SNNs offline, and running network training directly on neuromorphic hardware is an ongoing research problem. The primary hurdle is that back-propagation, which makes training such artificial deep networks possible, is biologically implausible. Neuroscientists are uncertain about how the brain would propagate a precise error signal backward through a network of neurons. Recent progress addresses part of this question, e.g., the weight transport problem, but a complete solution remains intangible. In contrast, novel learning rules based on the information bottleneck (IB) train each layer of a network independently, circumventing the need to propagate errors across layers. Instead, propagation is implicit due the layers' feedforward connectivity. These rules take the form of a three-factor Hebbian update a global error signal modulates local synaptic updates within each layer. Unfortunately, the global signal for a given layer requires processing multiple samples concurrently, and the brain only sees a single sample at a time. We propose a new three-factor update rule where the global signal correctly captures information across samples via an auxiliary memory network. The auxiliary network can be trained a priori independently of the dataset being used with the primary network. We demonstrate comparable performance to baselines on image classification tasks. Interestingly, unlike back-propagation-like schemes where there is no link between learning and memory, our rule presents a direct connection between working memory and synaptic updates. To the best of our knowledge, this is the first rule to make this link explicit. We explore these implications in initial experiments examining the effect of memory capacity on learning performance. Moving forward, this work suggests an alternate view of learning where each layer balances memory-informed compression against task performance. This view naturally encompasses several key aspects of neural computation, including memory, efficiency, and locality.
{"title":"Information bottleneck-based Hebbian learning rule naturally ties working memory and synaptic updates","authors":"Kyle Daruwalla, Mikko Lipasti","doi":"10.3389/fncom.2024.1240348","DOIUrl":"https://doi.org/10.3389/fncom.2024.1240348","url":null,"abstract":"Deep neural feedforward networks are effective models for a wide array of problems, but training and deploying such networks presents a significant energy cost. Spiking neural networks (SNNs), which are modeled after biologically realistic neurons, offer a potential solution when deployed correctly on neuromorphic computing hardware. Still, many applications train SNNs <jats:italic>offline</jats:italic>, and running network training directly on neuromorphic hardware is an ongoing research problem. The primary hurdle is that back-propagation, which makes training such artificial deep networks possible, is biologically implausible. Neuroscientists are uncertain about how the brain would propagate a precise error signal backward through a network of neurons. Recent progress addresses part of this question, e.g., the weight transport problem, but a complete solution remains intangible. In contrast, novel learning rules based on the information bottleneck (IB) train each layer of a network independently, circumventing the need to propagate errors across layers. Instead, propagation is implicit due the layers' feedforward connectivity. These rules take the form of a three-factor Hebbian update a global error signal modulates local synaptic updates within each layer. Unfortunately, the global signal for a given layer requires processing multiple samples concurrently, and the brain only sees a single sample at a time. We propose a new three-factor update rule where the global signal correctly captures information across samples via an auxiliary memory network. The auxiliary network can be trained <jats:italic>a priori</jats:italic> independently of the dataset being used with the primary network. We demonstrate comparable performance to baselines on image classification tasks. Interestingly, unlike back-propagation-like schemes where there is no link between learning and memory, our rule presents a direct connection between working memory and synaptic updates. To the best of our knowledge, this is the first rule to make this link explicit. We explore these implications in initial experiments examining the effect of memory capacity on learning performance. Moving forward, this work suggests an alternate view of learning where each layer balances memory-informed compression against task performance. This view naturally encompasses several key aspects of neural computation, including memory, efficiency, and locality.","PeriodicalId":12363,"journal":{"name":"Frontiers in Computational Neuroscience","volume":null,"pages":null},"PeriodicalIF":3.2,"publicationDate":"2024-05-16","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"141058716","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}
IntroductionNovel technologies based on virtual reality (VR) are creating attractive virtual environments with high ecological value, used both in basic/clinical neuroscience and modern medical practice. The study aimed to evaluate the effects of VR-based training in an elderly population.Materials and methodsThe study included 36 women over the age of 60, who were randomly divided into two groups subjected to balance-strength and balance-cognitive training. The research applied both conventional clinical tests, such as (a) the Timed Up and Go test, (b) the five-times sit-to-stand test, and (c) the posturographic exam with the Romberg test with eyes open and closed. Training in both groups was conducted for 10 sessions and embraced exercises on a bicycle ergometer and exercises using non-immersive VR created by the ActivLife platform. Machine learning methods with a k-nearest neighbors classifier, which are very effective and popular, were proposed to statistically evaluate the differences in training effects in the two groups.Results and conclusionThe study showed that training using VR brought beneficial improvement in clinical tests and changes in the pattern of posturographic trajectories were observed. An important finding of the research was a statistically significant reduction in the risk of falls in the study population. The use of virtual environments in exercise/training has great potential in promoting healthy aging and preventing balance loss and falls among seniors.
{"title":"A machine learning approach to evaluate the impact of virtual balance/cognitive training on fall risk in older women","authors":"Beata Sokołowska, Wiktor Świderski, Edyta Smolis-Bąk, Ewa Sokołowska, Teresa Sadura-Sieklucka","doi":"10.3389/fncom.2024.1390208","DOIUrl":"https://doi.org/10.3389/fncom.2024.1390208","url":null,"abstract":"IntroductionNovel technologies based on virtual reality (VR) are creating attractive virtual environments with high ecological value, used both in basic/clinical neuroscience and modern medical practice. The study aimed to evaluate the effects of VR-based training in an elderly population.Materials and methodsThe study included 36 women over the age of 60, who were randomly divided into two groups subjected to balance-strength and balance-cognitive training. The research applied both conventional clinical tests, such as (a) the Timed Up and Go test, (b) the five-times sit-to-stand test, and (c) the posturographic exam with the Romberg test with eyes open and closed. Training in both groups was conducted for 10 sessions and embraced exercises on a bicycle ergometer and exercises using non-immersive VR created by the ActivLife platform. Machine learning methods with a <jats:italic>k</jats:italic>-nearest neighbors classifier, which are very effective and popular, were proposed to statistically evaluate the differences in training effects in the two groups.Results and conclusionThe study showed that training using VR brought beneficial improvement in clinical tests and changes in the pattern of posturographic trajectories were observed. An important finding of the research was a statistically significant reduction in the risk of falls in the study population. The use of virtual environments in exercise/training has great potential in promoting healthy aging and preventing balance loss and falls among seniors.","PeriodicalId":12363,"journal":{"name":"Frontiers in Computational Neuroscience","volume":null,"pages":null},"PeriodicalIF":3.2,"publicationDate":"2024-05-14","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"140929638","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 : 2024-05-09DOI: 10.3389/fncom.2024.1327986
Peter Kan, Yong Fang Zhu, Junling Ma, Gurmit Singh
ObjectiveNav1.8 expression is restricted to sensory neurons; it was hypothesized that aberrant expression and function of this channel at the site of injury contributed to pathological pain. However, the specific contributions of Nav1.8 to neuropathic pain are not as clear as its role in inflammatory pain. The aim of this study is to understand how Nav1.8 present in peripheral sensory neurons regulate neuronal excitability and induce various electrophysiological features on neuropathic pain.MethodsTo study the effect of changes in sodium channel Nav1.8 kinetics, Hodgkin–Huxley type conductance-based models of spiking neurons were constructed using the NEURON v8.2 simulation software. We constructed a single-compartment model of neuronal soma that contained Nav1.8 channels with the ionic mechanisms adapted from some existing small DRG neuron models. We then validated and compared the model with our experimental data from in vivo recordings on soma of small dorsal root ganglion (DRG) sensory neurons in animal models of neuropathic pain (NEP).ResultsWe show that Nav1.8 is an important parameter for the generation and maintenance of abnormal neuronal electrogenesis and hyperexcitability. The typical increased excitability seen is dominated by a left shift in the steady state of activation of this channel and is further modulated by this channel’s maximum conductance and steady state of inactivation. Therefore, modified action potential shape, decreased threshold, and increased repetitive firing of sensory neurons in our neuropathic animal models may be orchestrated by these modulations on Nav1.8.ConclusionComputational modeling is a novel strategy to understand the generation of chronic pain. In this study, we highlight that changes to the channel functions of Nav1.8 within the small DRG neuron may contribute to neuropathic pain.
{"title":"Computational modeling to study the impact of changes in Nav1.8 sodium channel on neuropathic pain","authors":"Peter Kan, Yong Fang Zhu, Junling Ma, Gurmit Singh","doi":"10.3389/fncom.2024.1327986","DOIUrl":"https://doi.org/10.3389/fncom.2024.1327986","url":null,"abstract":"ObjectiveNav1.8 expression is restricted to sensory neurons; it was hypothesized that aberrant expression and function of this channel at the site of injury contributed to pathological pain. However, the specific contributions of Nav1.8 to neuropathic pain are not as clear as its role in inflammatory pain. The aim of this study is to understand how Nav1.8 present in peripheral sensory neurons regulate neuronal excitability and induce various electrophysiological features on neuropathic pain.MethodsTo study the effect of changes in sodium channel Nav1.8 kinetics, Hodgkin–Huxley type conductance-based models of spiking neurons were constructed using the NEURON v8.2 simulation software. We constructed a single-compartment model of neuronal soma that contained Nav1.8 channels with the ionic mechanisms adapted from some existing small DRG neuron models. We then validated and compared the model with our experimental data from <jats:italic>in vivo</jats:italic> recordings on soma of small dorsal root ganglion (DRG) sensory neurons in animal models of neuropathic pain (NEP).ResultsWe show that Nav1.8 is an important parameter for the generation and maintenance of abnormal neuronal electrogenesis and hyperexcitability. The typical increased excitability seen is dominated by a left shift in the steady state of activation of this channel and is further modulated by this channel’s maximum conductance and steady state of inactivation. Therefore, modified action potential shape, decreased threshold, and increased repetitive firing of sensory neurons in our neuropathic animal models may be orchestrated by these modulations on Nav1.8.ConclusionComputational modeling is a novel strategy to understand the generation of chronic pain. In this study, we highlight that changes to the channel functions of Nav1.8 within the small DRG neuron may contribute to neuropathic pain.","PeriodicalId":12363,"journal":{"name":"Frontiers in Computational Neuroscience","volume":null,"pages":null},"PeriodicalIF":3.2,"publicationDate":"2024-05-09","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"140929786","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 : 2024-05-09DOI: 10.3389/fncom.2024.1365727
Aaron Kujawa, Reuben Dorent, Steve Connor, Suki Thomson, Marina Ivory, Ali Vahedi, Emily Guilhem, Navodini Wijethilake, Robert Bradford, Neil Kitchen, Sotirios Bisdas, Sebastien Ourselin, Tom Vercauteren, Jonathan Shapey
Automatic segmentation of vestibular schwannoma (VS) from routine clinical MRI has potential to improve clinical workflow, facilitate treatment decisions, and assist patient management. Previous work demonstrated reliable automatic segmentation performance on datasets of standardized MRI images acquired for stereotactic surgery planning. However, diagnostic clinical datasets are generally more diverse and pose a larger challenge to automatic segmentation algorithms, especially when post-operative images are included. In this work, we show for the first time that automatic segmentation of VS on routine MRI datasets is also possible with high accuracy. We acquired and publicly release a curated multi-center routine clinical (MC-RC) dataset of 160 patients with a single sporadic VS. For each patient up to three longitudinal MRI exams with contrast-enhanced T1-weighted (ceT1w) (n = 124) and T2-weighted (T2w) (n = 363) images were included and the VS manually annotated. Segmentations were produced and verified in an iterative process: (1) initial segmentations by a specialized company; (2) review by one of three trained radiologists; and (3) validation by an expert team. Inter- and intra-observer reliability experiments were performed on a subset of the dataset. A state-of-the-art deep learning framework was used to train segmentation models for VS. Model performance was evaluated on a MC-RC hold-out testing set, another public VS datasets, and a partially public dataset. The generalizability and robustness of the VS deep learning segmentation models increased significantly when trained on the MC-RC dataset. Dice similarity coefficients (DSC) achieved by our model are comparable to those achieved by trained radiologists in the inter-observer experiment. On the MC-RC testing set, median DSCs were 86.2(9.5) for ceT1w, 89.4(7.0) for T2w, and 86.4(8.6) for combined ceT1w+T2w input images. On another public dataset acquired for Gamma Knife stereotactic radiosurgery our model achieved median DSCs of 95.3(2.9), 92.8(3.8), and 95.5(3.3), respectively. In contrast, models trained on the Gamma Knife dataset did not generalize well as illustrated by significant underperformance on the MC-RC routine MRI dataset, highlighting the importance of data variability in the development of robust VS segmentation models. The MC-RC dataset and all trained deep learning models were made available online.
从常规临床磁共振成像中自动分割前庭分裂瘤(VS)有望改善临床工作流程、促进治疗决策并协助患者管理。之前的工作表明,在为立体定向手术规划而获取的标准化磁共振成像数据集上,自动分割性能可靠。然而,临床诊断数据集通常更加多样化,对自动分割算法提出了更大的挑战,尤其是在包含术后图像的情况下。在这项工作中,我们首次展示了在常规磁共振成像数据集上自动分割 VS 的高准确性。我们获得并公开发布了一个由 160 名单个散发性 VS 患者组成的多中心常规临床(MC-RC)数据集。每位患者最多可接受三次纵向 MRI 检查,包括对比增强 T1 加权(ceT1w)(124 人)和 T2 加权(T2w)(363 人)图像,并对 VS 进行人工标注。分段的制作和验证是一个反复的过程:(1) 由一家专业公司进行初步分段;(2) 由三位训练有素的放射科医生之一进行审查;(3) 由一个专家组进行验证。在数据集的一个子集上进行了观察者之间和观察者内部的可靠性实验。最先进的深度学习框架用于训练 VS 的分割模型。在 MC-RC 暂缓测试集、另一个公开 VS 数据集和一个部分公开数据集上对模型性能进行了评估。在 MC-RC 数据集上训练的 VS 深度学习分割模型的泛化能力和鲁棒性显著提高。在观察者间实验中,我们的模型获得的骰子相似系数(DSC)与经过培训的放射科医生获得的相似系数相当。在 MC-RC 测试集中,ceT1w 的 DSC 中位数为 86.2(9.5),T2w 为 89.4(7.0),ceT1w+T2w 组合输入图像的 DSC 中位数为 86.4(8.6)。在为伽马刀立体定向放射外科手术获取的另一个公共数据集上,我们的模型分别获得了 95.3(2.9)、92.8(3.8) 和 95.5(3.3) 的中位 DSCs。相比之下,在伽马刀数据集上训练的模型并不能很好地泛化,在 MC-RC 常规 MRI 数据集上的表现就说明了这一点,这突出了数据可变性在开发稳健的 VS 分割模型中的重要性。MC-RC 数据集和所有经过训练的深度学习模型均可在线获取。
{"title":"Deep learning for automatic segmentation of vestibular schwannoma: a retrospective study from multi-center routine MRI","authors":"Aaron Kujawa, Reuben Dorent, Steve Connor, Suki Thomson, Marina Ivory, Ali Vahedi, Emily Guilhem, Navodini Wijethilake, Robert Bradford, Neil Kitchen, Sotirios Bisdas, Sebastien Ourselin, Tom Vercauteren, Jonathan Shapey","doi":"10.3389/fncom.2024.1365727","DOIUrl":"https://doi.org/10.3389/fncom.2024.1365727","url":null,"abstract":"Automatic segmentation of vestibular schwannoma (VS) from routine clinical MRI has potential to improve clinical workflow, facilitate treatment decisions, and assist patient management. Previous work demonstrated reliable automatic segmentation performance on datasets of standardized MRI images acquired for stereotactic surgery planning. However, diagnostic clinical datasets are generally more diverse and pose a larger challenge to automatic segmentation algorithms, especially when post-operative images are included. In this work, we show for the first time that automatic segmentation of VS on routine MRI datasets is also possible with high accuracy. We acquired and publicly release a curated multi-center routine clinical (MC-RC) dataset of 160 patients with a single sporadic VS. For each patient up to three longitudinal MRI exams with contrast-enhanced T1-weighted (ceT1w) (<jats:italic>n</jats:italic> = 124) and T2-weighted (T2w) (<jats:italic>n</jats:italic> = 363) images were included and the VS manually annotated. Segmentations were produced and verified in an iterative process: (1) initial segmentations by a specialized company; (2) review by one of three trained radiologists; and (3) validation by an expert team. Inter- and intra-observer reliability experiments were performed on a subset of the dataset. A state-of-the-art deep learning framework was used to train segmentation models for VS. Model performance was evaluated on a MC-RC hold-out testing set, another public VS datasets, and a partially public dataset. The generalizability and robustness of the VS deep learning segmentation models increased significantly when trained on the MC-RC dataset. Dice similarity coefficients (DSC) achieved by our model are comparable to those achieved by trained radiologists in the inter-observer experiment. On the MC-RC testing set, median DSCs were 86.2(9.5) for ceT1w, 89.4(7.0) for T2w, and 86.4(8.6) for combined ceT1w+T2w input images. On another public dataset acquired for Gamma Knife stereotactic radiosurgery our model achieved median DSCs of 95.3(2.9), 92.8(3.8), and 95.5(3.3), respectively. In contrast, models trained on the Gamma Knife dataset did not generalize well as illustrated by significant underperformance on the MC-RC routine MRI dataset, highlighting the importance of data variability in the development of robust VS segmentation models. The MC-RC dataset and all trained deep learning models were made available online.","PeriodicalId":12363,"journal":{"name":"Frontiers in Computational Neuroscience","volume":null,"pages":null},"PeriodicalIF":3.2,"publicationDate":"2024-05-09","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"140929649","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 : 2024-05-02DOI: 10.3389/fncom.2024.1340019
Bozhi Qiu, Sheng Li, Lei Wang
Harnessing the remarkable ability of the human brain to recognize and process complex data is a significant challenge for researchers, particularly in the domain of point cloud classification—a technology that aims to replicate the neural structure of the brain for spatial recognition. The initial 3D point cloud data often suffers from noise, sparsity, and disorder, making accurate classification a formidable task, especially when extracting local information features. Therefore, in this study, we propose a novel attention-based end-to-end point cloud downsampling classification method, termed as PointAS, which is an experimental algorithm designed to be adaptable to various downstream tasks. PointAS consists of two primary modules: the adaptive sampling module and the attention module. Specifically, the attention module aggregates global features with the input point cloud data, while the adaptive module extracts local features. In the point cloud classification task, our method surpasses existing downsampling methods by a significant margin, allowing for more precise extraction of edge data points to capture overall contour features accurately. The classification accuracy of PointAS consistently exceeds 80% across various sampling ratios, with a remarkable accuracy of 75.37% even at ultra-high sampling ratios. Moreover, our method exhibits robustness in experiments, maintaining classification accuracies of 72.50% or higher under different noise disturbances. Both qualitative and quantitative experiments affirm the efficacy of our approach in the sampling classification task, providing researchers with a more accurate method to identify and classify neurons, synapses, and other structures, thereby promoting a deeper understanding of the nervous system.
{"title":"PointAS: an attention based sampling neural network for visual perception","authors":"Bozhi Qiu, Sheng Li, Lei Wang","doi":"10.3389/fncom.2024.1340019","DOIUrl":"https://doi.org/10.3389/fncom.2024.1340019","url":null,"abstract":"Harnessing the remarkable ability of the human brain to recognize and process complex data is a significant challenge for researchers, particularly in the domain of point cloud classification—a technology that aims to replicate the neural structure of the brain for spatial recognition. The initial 3D point cloud data often suffers from noise, sparsity, and disorder, making accurate classification a formidable task, especially when extracting local information features. Therefore, in this study, we propose a novel attention-based end-to-end point cloud downsampling classification method, termed as PointAS, which is an experimental algorithm designed to be adaptable to various downstream tasks. PointAS consists of two primary modules: the adaptive sampling module and the attention module. Specifically, the attention module aggregates global features with the input point cloud data, while the adaptive module extracts local features. In the point cloud classification task, our method surpasses existing downsampling methods by a significant margin, allowing for more precise extraction of edge data points to capture overall contour features accurately. The classification accuracy of PointAS consistently exceeds 80% across various sampling ratios, with a remarkable accuracy of 75.37% even at ultra-high sampling ratios. Moreover, our method exhibits robustness in experiments, maintaining classification accuracies of 72.50% or higher under different noise disturbances. Both qualitative and quantitative experiments affirm the efficacy of our approach in the sampling classification task, providing researchers with a more accurate method to identify and classify neurons, synapses, and other structures, thereby promoting a deeper understanding of the nervous system.","PeriodicalId":12363,"journal":{"name":"Frontiers in Computational Neuroscience","volume":null,"pages":null},"PeriodicalIF":3.2,"publicationDate":"2024-05-02","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"140834094","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 : 2024-05-02DOI: 10.3389/fncom.2024.1385047
Hui Tian, Xin Su, Yanfang Hou
BackgroundAs an important mathematical model, the finite state machine (FSM) has been used in many fields, such as manufacturing system, health care, and so on. This paper analyzes the current development status of FSMs. It is pointed out that the traditional methods are often inconvenient for analysis and design, or encounter high computational complexity problems when studying FSMs.MethodThe deep Q-network (DQN) technique, which is a model-free optimization method, is introduced to solve the stabilization problem of probabilistic finite state machines (PFSMs). In order to better understand the technique, some preliminaries, including Markov decision process, ϵ-greedy strategy, DQN, and so on, are recalled.ResultsFirst, a necessary and sufficient stabilizability condition for PFSMs is derived. Next, the feedback stabilization problem of PFSMs is transformed into an optimization problem. Finally, by using the stabilizability condition and deep Q-network, an algorithm for solving the optimization problem (equivalently, computing a state feedback stabilizer) is provided.DiscussionCompared with the traditional Q learning, DQN avoids the limited capacity problem. So our method can deal with high-dimensional complex systems efficiently. The effectiveness of our method is further demonstrated through an illustrative example.
{"title":"Feedback stabilization of probabilistic finite state machines based on deep Q-network","authors":"Hui Tian, Xin Su, Yanfang Hou","doi":"10.3389/fncom.2024.1385047","DOIUrl":"https://doi.org/10.3389/fncom.2024.1385047","url":null,"abstract":"BackgroundAs an important mathematical model, the finite state machine (FSM) has been used in many fields, such as manufacturing system, health care, and so on. This paper analyzes the current development status of FSMs. It is pointed out that the traditional methods are often inconvenient for analysis and design, or encounter high computational complexity problems when studying FSMs.MethodThe deep Q-network (DQN) technique, which is a model-free optimization method, is introduced to solve the stabilization problem of probabilistic finite state machines (PFSMs). In order to better understand the technique, some preliminaries, including Markov decision process, ϵ-greedy strategy, DQN, and so on, are recalled.ResultsFirst, a necessary and sufficient stabilizability condition for PFSMs is derived. Next, the feedback stabilization problem of PFSMs is transformed into an optimization problem. Finally, by using the stabilizability condition and deep Q-network, an algorithm for solving the optimization problem (equivalently, computing a state feedback stabilizer) is provided.DiscussionCompared with the traditional Q learning, DQN avoids the limited capacity problem. So our method can deal with high-dimensional complex systems efficiently. The effectiveness of our method is further demonstrated through an illustrative example.","PeriodicalId":12363,"journal":{"name":"Frontiers in Computational Neuroscience","volume":null,"pages":null},"PeriodicalIF":3.2,"publicationDate":"2024-05-02","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"140834210","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 : 2024-04-25DOI: 10.3389/fncom.2024.1393849
Nabeela Yaqoob, Muhammad Attique Khan, Saleha Masood, Hussain Mobarak Albarakati, Ameer Hamza, Fatimah Alhayan, Leila Jamel, Anum Masood
Alzheimer's disease (AD) is a neurodegenerative illness that impairs cognition, function, and behavior by causing irreversible damage to multiple brain areas, including the hippocampus. The suffering of the patients and their family members will be lessened with an early diagnosis of AD. The automatic diagnosis technique is widely required due to the shortage of medical experts and eases the burden of medical staff. The automatic artificial intelligence (AI)-based computerized method can help experts achieve better diagnosis accuracy and precision rates. This study proposes a new automated framework for AD stage prediction based on the ResNet-Self architecture and Fuzzy Entropy-controlled Path-Finding Algorithm (FEcPFA). A data augmentation technique has been utilized to resolve the dataset imbalance issue. In the next step, we proposed a new deep-learning model based on the self-attention module. A ResNet-50 architecture is modified and connected with a self-attention block for important information extraction. The hyperparameters were optimized using Bayesian optimization (BO) and then utilized to train the model, which was subsequently employed for feature extraction. The self-attention extracted features were optimized using the proposed FEcPFA. The best features were selected using FEcPFA and passed to the machine learning classifiers for the final classification. The experimental process utilized a publicly available MRI dataset and achieved an improved accuracy of 99.9%. The results were compared with state-of-the-art (SOTA) techniques, demonstrating the improvement of the proposed framework in terms of accuracy and time efficiency.
{"title":"Prediction of Alzheimer's disease stages based on ResNet-Self-attention architecture with Bayesian optimization and best features selection","authors":"Nabeela Yaqoob, Muhammad Attique Khan, Saleha Masood, Hussain Mobarak Albarakati, Ameer Hamza, Fatimah Alhayan, Leila Jamel, Anum Masood","doi":"10.3389/fncom.2024.1393849","DOIUrl":"https://doi.org/10.3389/fncom.2024.1393849","url":null,"abstract":"Alzheimer's disease (AD) is a neurodegenerative illness that impairs cognition, function, and behavior by causing irreversible damage to multiple brain areas, including the hippocampus. The suffering of the patients and their family members will be lessened with an early diagnosis of AD. The automatic diagnosis technique is widely required due to the shortage of medical experts and eases the burden of medical staff. The automatic artificial intelligence (AI)-based computerized method can help experts achieve better diagnosis accuracy and precision rates. This study proposes a new automated framework for AD stage prediction based on the ResNet-Self architecture and Fuzzy Entropy-controlled Path-Finding Algorithm (FEcPFA). A data augmentation technique has been utilized to resolve the dataset imbalance issue. In the next step, we proposed a new deep-learning model based on the self-attention module. A ResNet-50 architecture is modified and connected with a self-attention block for important information extraction. The hyperparameters were optimized using Bayesian optimization (BO) and then utilized to train the model, which was subsequently employed for feature extraction. The self-attention extracted features were optimized using the proposed FEcPFA. The best features were selected using FEcPFA and passed to the machine learning classifiers for the final classification. The experimental process utilized a publicly available MRI dataset and achieved an improved accuracy of 99.9%. The results were compared with state-of-the-art (SOTA) techniques, demonstrating the improvement of the proposed framework in terms of accuracy and time efficiency.","PeriodicalId":12363,"journal":{"name":"Frontiers in Computational Neuroscience","volume":null,"pages":null},"PeriodicalIF":3.2,"publicationDate":"2024-04-25","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"140653862","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 : 2024-04-19DOI: 10.3389/fncom.2024.1276292
Robin Dietrich, Nicolai Waniek, Martin Stemmler, Alois Knoll
IntroductionRecent work on bats flying over long distances has revealed that single hippocampal cells have multiple place fields of different sizes. At the network level, a multi-scale, multi-field place cell code outperforms classical single-scale, single-field place codes, yet the performance boundaries of such a code remain an open question. In particular, it is unknown how general multi-field codes compare to a highly regular grid code, in which cells form distinct modules with different scales.MethodsIn this work, we address the coding properties of theoretical spatial coding models with rigorous analyses of comprehensive simulations. Starting from a multi-scale, multi-field network, we performed evolutionary optimization. The resulting multi-field networks sometimes retained the multi-scale property at the single-cell level but most often converged to a single scale, with all place fields in a given cell having the same size. We compared the results against a single-scale single-field code and a one-dimensional grid code, focusing on two main characteristics: the performance of the code itself and the dynamics of the network generating it.ResultsOur simulation experiments revealed that, under normal conditions, a regular grid code outperforms all other codes with respect to decoding accuracy, achieving a given precision with fewer neurons and fields. In contrast, multi-field codes are more robust against noise and lesions, such as random drop-out of neurons, given that the significantly higher number of fields provides redundancy. Contrary to our expectations, the network dynamics of all models, from the original multi-scale models before optimization to the multi-field models that resulted from optimization, did not maintain activity bumps at their original locations when a position-specific external input was removed.DiscussionOptimized multi-field codes appear to strike a compromise between a place code and a grid code that reflects a trade-off between accurate positional encoding and robustness. Surprisingly, the recurrent neural network models we implemented and optimized for either multi- or single-scale, multi-field codes did not intrinsically produce a persistent “memory” of attractor states. These models, therefore, were not continuous attractor networks.
{"title":"Grid codes vs. multi-scale, multi-field place codes for space","authors":"Robin Dietrich, Nicolai Waniek, Martin Stemmler, Alois Knoll","doi":"10.3389/fncom.2024.1276292","DOIUrl":"https://doi.org/10.3389/fncom.2024.1276292","url":null,"abstract":"IntroductionRecent work on bats flying over long distances has revealed that single hippocampal cells have multiple place fields of different sizes. At the network level, a multi-scale, multi-field place cell code outperforms classical single-scale, single-field place codes, yet the performance boundaries of such a code remain an open question. In particular, it is unknown how general multi-field codes compare to a highly regular grid code, in which cells form distinct modules with different scales.MethodsIn this work, we address the coding properties of theoretical spatial coding models with rigorous analyses of comprehensive simulations. Starting from a multi-scale, multi-field network, we performed evolutionary optimization. The resulting multi-field networks sometimes retained the multi-scale property at the single-cell level but most often converged to a single scale, with all place fields in a given cell having the same size. We compared the results against a single-scale single-field code and a one-dimensional grid code, focusing on two main characteristics: the performance of the code itself and the dynamics of the network generating it.ResultsOur simulation experiments revealed that, under normal conditions, a regular grid code outperforms all other codes with respect to decoding accuracy, achieving a given precision with fewer neurons and fields. In contrast, multi-field codes are more robust against noise and lesions, such as random drop-out of neurons, given that the significantly higher number of fields provides redundancy. Contrary to our expectations, the network dynamics of all models, from the original multi-scale models before optimization to the multi-field models that resulted from optimization, did not maintain activity bumps at their original locations when a position-specific external input was removed.DiscussionOptimized multi-field codes appear to strike a compromise between a place code and a grid code that reflects a trade-off between accurate positional encoding and robustness. Surprisingly, the recurrent neural network models we implemented and optimized for either multi- or single-scale, multi-field codes did not intrinsically produce a persistent “memory” of attractor states. These models, therefore, were not continuous attractor networks.","PeriodicalId":12363,"journal":{"name":"Frontiers in Computational Neuroscience","volume":null,"pages":null},"PeriodicalIF":3.2,"publicationDate":"2024-04-19","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"140627809","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 : 2024-04-17DOI: 10.3389/fncom.2024.1350916
Kai Cheng
Existing methods for classifying image emotions often overlook the subjective impact emotions evoke in observers, focusing primarily on emotion categories. However, this approach falls short in meeting practical needs as it neglects the nuanced emotional responses captured within an image. This study proposes a novel approach employing the weighted closest neighbor algorithm to predict the discrete distribution of emotion in abstract paintings. Initially, emotional features are extracted from the images and assigned varying K-values. Subsequently, an encoder-decoder architecture is utilized to derive sentiment features from abstract paintings, augmented by a pre-trained model to enhance classification model generalization and convergence speed. By incorporating a blank attention mechanism into the decoder and integrating it with the encoder's output sequence, the semantics of abstract painting images are learned, facilitating precise and sensible emotional understanding. Experimental results demonstrate that the classification algorithm, utilizing the attention mechanism, achieves a higher accuracy of 80.7% compared to current methods. This innovative approach successfully addresses the intricate challenge of discerning emotions in abstract paintings, underscoring the significance of considering subjective emotional responses in image classification. The integration of advanced techniques such as weighted closest neighbor algorithm and attention mechanisms holds promise for enhancing the comprehension and classification of emotional content in visual art.
现有的图像情绪分类方法往往忽视情绪对观察者的主观影响,而主要关注情绪类别。然而,这种方法忽略了图像中细微的情感反应,无法满足实际需要。本研究提出了一种采用加权近邻算法预测抽象绘画中情感离散分布的新方法。首先,从图像中提取情感特征并赋予不同的 K 值。随后,利用编码器-解码器架构从抽象绘画中提取情感特征,并通过预训练模型增强分类模型的泛化和收敛速度。通过在解码器中加入空白关注机制,并将其与编码器的输出序列相结合,可以学习抽象绘画图像的语义,从而促进精确、合理的情感理解。实验结果表明,与现有方法相比,利用注意力机制的分类算法的准确率高达 80.7%。这一创新方法成功地解决了辨别抽象画中情感这一复杂难题,强调了在图像分类中考虑主观情感反应的重要性。加权近邻算法和注意力机制等先进技术的整合有望提高对视觉艺术中情感内容的理解和分类。
{"title":"Prediction of emotion distribution of images based on weighted K-nearest neighbor-attention mechanism","authors":"Kai Cheng","doi":"10.3389/fncom.2024.1350916","DOIUrl":"https://doi.org/10.3389/fncom.2024.1350916","url":null,"abstract":"Existing methods for classifying image emotions often overlook the subjective impact emotions evoke in observers, focusing primarily on emotion categories. However, this approach falls short in meeting practical needs as it neglects the nuanced emotional responses captured within an image. This study proposes a novel approach employing the weighted closest neighbor algorithm to predict the discrete distribution of emotion in abstract paintings. Initially, emotional features are extracted from the images and assigned varying <jats:italic>K</jats:italic>-values. Subsequently, an encoder-decoder architecture is utilized to derive sentiment features from abstract paintings, augmented by a pre-trained model to enhance classification model generalization and convergence speed. By incorporating a blank attention mechanism into the decoder and integrating it with the encoder's output sequence, the semantics of abstract painting images are learned, facilitating precise and sensible emotional understanding. Experimental results demonstrate that the classification algorithm, utilizing the attention mechanism, achieves a higher accuracy of 80.7% compared to current methods. This innovative approach successfully addresses the intricate challenge of discerning emotions in abstract paintings, underscoring the significance of considering subjective emotional responses in image classification. The integration of advanced techniques such as weighted closest neighbor algorithm and attention mechanisms holds promise for enhancing the comprehension and classification of emotional content in visual art.","PeriodicalId":12363,"journal":{"name":"Frontiers in Computational Neuroscience","volume":null,"pages":null},"PeriodicalIF":3.2,"publicationDate":"2024-04-17","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"140613294","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}
IntroductionThe blood oxygen level-dependent (BOLD) signal derived from functional neuroimaging is commonly used in brain network analysis and dementia diagnosis. Missing the BOLD signal may lead to bad performance and misinterpretation of findings when analyzing neurological disease. Few studies have focused on the restoration of brain functional time-series data.MethodsIn this paper, a novel U-shaped convolutional transformer GAN (UCT-GAN) model is proposed to restore the missing brain functional time-series data. The proposed model leverages the power of generative adversarial networks (GANs) while incorporating a U-shaped architecture to effectively capture hierarchical features in the restoration process. Besides, the multi-level temporal-correlated attention and the convolutional sampling in the transformer-based generator are devised to capture the global and local temporal features for the missing time series and associate their long-range relationship with the other brain regions. Furthermore, by introducing multi-resolution consistency loss, the proposed model can promote the learning of diverse temporal patterns and maintain consistency across different temporal resolutions, thus effectively restoring complex brain functional dynamics.ResultsWe theoretically tested our model on the public Alzheimer's Disease Neuroimaging Initiative (ADNI) dataset, and our experiments demonstrate that the proposed model outperforms existing methods in terms of both quantitative metrics and qualitative assessments. The model's ability to preserve the underlying topological structure of the brain functional networks during restoration is a particularly notable achievement.ConclusionOverall, the proposed model offers a promising solution for restoring brain functional time-series and contributes to the advancement of neuroscience research by providing enhanced tools for disease analysis and interpretation.
导言从功能神经成像中获得的血氧水平依赖性(BOLD)信号常用于脑网络分析和痴呆诊断。在分析神经系统疾病时,BOLD 信号的缺失可能会导致不良表现和结果误读。本文提出了一种新型 U 形卷积变换器 GAN(UCT-GAN)模型,用于恢复缺失的脑功能时间序列数据。该模型充分利用了生成式对抗网络(GAN)的强大功能,同时结合了 U 型结构,从而在还原过程中有效捕捉分层特征。此外,基于变压器的生成器中还设计了多级时间相关注意和卷积采样,以捕捉缺失时间序列的全局和局部时间特征,并将其与其他脑区的长程关系联系起来。此外,通过引入多分辨率一致性损失,所提出的模型可以促进对不同时间模式的学习,并在不同时间分辨率之间保持一致性,从而有效地恢复复杂的大脑功能动态。结果我们在公开的阿尔茨海默病神经影像倡议(ADNI)数据集上对我们的模型进行了理论测试,实验证明所提出的模型在定量指标和定性评估方面都优于现有方法。总之,所提出的模型为恢复大脑功能时间序列提供了一种很有前景的解决方案,并通过为疾病分析和解释提供增强工具,为神经科学研究的进步做出了贡献。
{"title":"U-shaped convolutional transformer GAN with multi-resolution consistency loss for restoring brain functional time-series and dementia diagnosis","authors":"Qiankun Zuo, Ruiheng Li, Binghua Shi, Jin Hong, Yanfei Zhu, Xuhang Chen, Yixian Wu, Jia Guo","doi":"10.3389/fncom.2024.1387004","DOIUrl":"https://doi.org/10.3389/fncom.2024.1387004","url":null,"abstract":"IntroductionThe blood oxygen level-dependent (BOLD) signal derived from functional neuroimaging is commonly used in brain network analysis and dementia diagnosis. Missing the BOLD signal may lead to bad performance and misinterpretation of findings when analyzing neurological disease. Few studies have focused on the restoration of brain functional time-series data.MethodsIn this paper, a novel <jats:italic>U</jats:italic>-shaped convolutional transformer GAN (UCT-GAN) model is proposed to restore the missing brain functional time-series data. The proposed model leverages the power of generative adversarial networks (GANs) while incorporating a <jats:italic>U</jats:italic>-shaped architecture to effectively capture hierarchical features in the restoration process. Besides, the multi-level temporal-correlated attention and the convolutional sampling in the transformer-based generator are devised to capture the global and local temporal features for the missing time series and associate their long-range relationship with the other brain regions. Furthermore, by introducing multi-resolution consistency loss, the proposed model can promote the learning of diverse temporal patterns and maintain consistency across different temporal resolutions, thus effectively restoring complex brain functional dynamics.ResultsWe theoretically tested our model on the public Alzheimer's Disease Neuroimaging Initiative (ADNI) dataset, and our experiments demonstrate that the proposed model outperforms existing methods in terms of both quantitative metrics and qualitative assessments. The model's ability to preserve the underlying topological structure of the brain functional networks during restoration is a particularly notable achievement.ConclusionOverall, the proposed model offers a promising solution for restoring brain functional time-series and contributes to the advancement of neuroscience research by providing enhanced tools for disease analysis and interpretation.","PeriodicalId":12363,"journal":{"name":"Frontiers in Computational Neuroscience","volume":null,"pages":null},"PeriodicalIF":3.2,"publicationDate":"2024-04-17","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"140613436","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}