Decoding Method Research Articles

A Two-Individual-Based Evolutionary Algorithm for Flexible Assembly Job Shop Scheduling Problem with Uncertain Interval Processing Times

The assembly job shop is a prevalent production organization mode in manufacturing enterprises. During the processing and assembly of products, operation processing times are influenced by numerous factors, leading to significant uncertainty. This paper investigates the flexible assembly job shop scheduling problem (FAJSP) with uncertain processing times, where processing times are represented as variable interval numbers. We develop a robust optimization model for the FAJSP, utilizing confidence level estimation to determine the ranges of processing times and reformulating the model based on the chance-constrained method. A two-individual-based master–apprentice evolutionary (MAE) algorithm is proposed. Two effective encoding schemes are designed to prevent the generation of infeasible solutions under assembly sequence constraints. Additionally, a decoding method based on interval scheduling theory is devised to accurately represent interval processing times. Case studies are conducted to validate the effectiveness of the proposed robust optimization model and demonstrate the superiority of the MAE algorithm.

SoftBoMI: a non-invasive wearable body-machine interface for mapping movement of shoulder to commands

Objective.Customized human-machine interfaces for controlling assistive devices are vital in improving the self-help ability of upper limb amputees and tetraplegic patients. Given that most of them possess residual shoulder mobility, using it to generate commands to operate assistive devices can serve as a complementary approach to brain-computer interfaces.Approach.We propose a hybrid body-machine interface prototype that integrates soft sensors and an inertial measurement unit. This study introduces both a rule-based data decoding method and a user intent inference-based decoding method to map human shoulder movements into continuous commands. Additionally, by incorporating prior knowledge of the user's operational performance into a shared autonomy framework, we implement an adaptive switching command mapping approach. This approach enables seamless transitions between the two decoding methods, enhancing their adaptability across different tasks.Main results.The proposed method has been validated on individuals with cervical spinal cord injury, bilateral arm amputation, and healthy subjects through a series of center-out target reaching tasks and a virtual powered wheelchair driving task. The experimental results show that using both the soft sensors and the gyroscope exhibits the most well-rounded performance in intent inference. Additionally, the rule-based method demonstrates better dynamic performance for wheelchair operation, while the intent inference method is more accurate but has higher latency. Adaptive switching decoding methods offer the best adaptability by seamlessly transitioning between decoding methods for different tasks. Furthermore, we discussed the differences and characteristics among the various types of participants in the experiment.Significance.The proposed method has the potential to be integrated into clothing, enabling non-invasive interaction with assistive devices in daily life, and could serve as a tool for rehabilitation assessment in the future.

Registered Report Stage II: Decoding the category information from evoked potentials to visible and invisible visual objects

Previous studies that use decoding methods and EEG to investigate the neural representation of the category information of visual objects focused mainly on consciously processed visual objects. It remains unclear whether the category information of unconsciously processed visual objects can be decoded and whether the decoding performance is different for consciously and unconsciously processed visual objects. The present study compared the neural decoding of the animacy category of visible and invisible visual objects via EEG and decoding methods. The results revealed that the animacy of visible visual objects could be decoded above the chance level by the P200, N300, and N400, but not by the early N/P100. However, the animacy of invisible visual objects could not be decoded above the chance level by neither early nor late ERP components. The decoding accuracy was greater for visible visual objects than that for invisible visual objects for the P200, N300 and N400. These results suggested that access to animacy category information for visual objects requires conscious processing.

Gesture2Text: A Generalizable Decoder for Word-Gesture Keyboards in XR Through Trajectory Coarse Discretization and Pre-Training.

Text entry with word-gesture keyboards (WGK) is emerging as a popular method and becoming a key interaction for Extended Reality (XR). However, the diversity of interaction modes, keyboard sizes, and visual feedback in these environments introduces divergent word-gesture trajectory data patterns, thus leading to complexity in decoding trajectories into text. Template-matching decoding methods, such as SHARK2 [32], are commonly used for these WGK systems because they are easy to implement and configure. However, these methods are susceptible to decoding inaccuracies for noisy trajectories. While conventional neural-network-based decoders (neural decoders) trained on word-gesture trajectory data have been proposed to improve accuracy, they have their own limitations: they require extensive data for training and deep-learning expertise for implementation. To address these challenges, we propose a novel solution that combines ease of implementation with high decoding accuracy: a generalizable neural decoder enabled by pre-training on large-scale coarsely discretized word-gesture trajectories. This approach produces a ready-to-use WGK decoder that is generalizable across mid-air and on-surface WGK systems in augmented reality (AR) and virtual reality (VR), which is evident by a robust average Top-4 accuracy of 90.4% on four diverse datasets. It significantly outperforms SHARK2 with a 37.2% enhancement and surpasses the conventional neural decoder by 7.4%. Moreover, the Pre-trained Neural Decoder's size is only 4 MB after quantization, without sacrificing accuracy, and it can operate in real-time, executing in just 97 milliseconds on Quest 3.

A temporal–spectral fusion transformer with subject-specific adapter for enhancing RSVP-BCI decoding

The Rapid Serial Visual Presentation (RSVP)-based Brain–Computer Interface (BCI) is an efficient technology for target retrieval using electroencephalography (EEG) signals. The performance improvement of traditional decoding methods relies on a substantial amount of training data from new test subjects, which increases preparation time for BCI systems. Several studies introduce data from existing subjects to reduce the dependence of performance improvement on data from new subjects, but their optimization strategy based on adversarial learning with extensive data increases training time during the preparation procedure. Moreover, most previous methods only focus on the single-view information of EEG signals, but ignore the information from other views which may further improve performance. To enhance decoding performance while reducing preparation time, we propose a Temporal-Spectral fusion transformer with Subject-specific Adapter (TSformer-SA). Specifically, a cross-view interaction module is proposed to facilitate information transfer and extract common representations across two-view features extracted from EEG temporal signals and spectrogram images. Then, an attention-based fusion module fuses the features of two views to obtain comprehensive discriminative features for classification. Furthermore, a multi-view consistency loss is proposed to maximize the feature similarity between two views of the same EEG signal. Finally, we propose a subject-specific adapter to rapidly transfer the knowledge of the model trained on data from existing subjects to decode data from new subjects. Experimental results show that TSformer-SA significantly outperforms comparison methods and achieves outstanding performance with limited training data from new subjects. This facilitates efficient decoding and rapid deployment of BCI systems in practical use.

Neural decoding and feature selection methods for closed-loop control of avoidance behavior

Objective.Many psychiatric disorders involve excessive avoidant or defensive behavior, such as avoidance in anxiety and trauma disorders or defensive rituals in obsessive-compulsive disorders. Developing algorithms to predict these behaviors from local field potentials (LFPs) could serve as the foundational technology for closed-loop control of such disorders. A significant challenge is identifying the LFP features that encode these defensive behaviors.Approach.We analyzed LFP signals from the infralimbic cortex and basolateral amygdala of rats undergoing tone-shock conditioning and extinction, standard for investigating defensive behaviors. We utilized a comprehensive set of neuro-markers across spectral, temporal, and connectivity domains, employing SHapley Additive exPlanations for feature importance evaluation within Light Gradient-Boosting Machine models. Our goal was to decode three commonly studied avoidance/defensive behaviors: freezing, bar-press suppression, and motion (accelerometry), examining the impact of different features on decoding performance.Main results.Band power and band power ratio between channels emerged as optimal features across sessions. High-gamma (80-150 Hz) power, power ratios, and inter-regional correlations were more informative than other bands that are more classically linked to defensive behaviors. Focusing on highly informative features enhanced performance. Across 4 recording sessions with 16 subjects, we achieved an average coefficient of determination of 0.5357 and 0.3476, and Pearson correlation coefficients of 0.7579 and 0.6092 for accelerometry jerk and bar press rate, respectively. Utilizing only the most informative features revealed differential encoding between accelerometry and bar press rate, with the former primarily through local spectral power and the latter via inter-regional connectivity. Our methodology demonstrated remarkably low training/inference time and memory usage, requiring<310 ms for training,<0.051 ms for inference, and 16.6 kB of memory, using a single core of AMD Ryzen Threadripper PRO 5995WX CPU.Significance.Our results demonstrate the feasibility of accurately decoding defensive behaviors with minimal latency, using LFP features from neural circuits strongly linked to these behaviors. This methodology holds promise for real-time decoding to identify physiological targets in closed-loop psychiatric neuromodulation.

Integrated optimization of production and maintenance scheduling with third-party worker resource constraints in distributed parallel machines environment

Outsourcing machine maintenance to third parties has been a trend due to the increasing complex of machines and the benefits of maintenance outsourcing. However, this new phenomenon is ignored in previous studies pertaining to the integrated optimization of production and maintenance scheduling (IOPMS) problems, resulting the lack of theoretical guidance for production managers to formulate optimal scheduling schemes under this new trend. In this study, we investigate an IOPMS problem considering maintenance outsourcing in a distributed parallel machine environment, referred as IOPMSTW, which requires the use of third-party worker resources to perform preventive maintenance. Makespan and total cost are two optimization objectives. We first formulate the problem by developing a mixed integer linear programming. Then a memetic algorithm incorporating iterated greedy method (IG) is proposed to solve the IOPMSTW, in which an improved decoding method and a problem-dependent local search operator based on IG are designed to respectively ensure the feasibility of new generated individuals and improve the searching efficiency. The validity of proposed mathematical model is verified by CPLEX based on eight small instances. Based on 240 constructed instances, a set of comprehensive experiments are conducted. The results demonstrate that the local search operator improved the searching performance of the proposed algorithm by 100%. Comparison results with other well-known algorithms show that the proposed algorithm achieved the best results on more than 85% of the tested instances.

A transient high-dimensional geometry affords stable conjunctive subspaces for efficient action selection

Flexible action selection requires cognitive control mechanisms capable of mapping the same inputs to different output actions depending on the context. From a neural state-space perspective, this requires a control representation that separates similar input neural states by context. Additionally, for action selection to be robust and time-invariant, information must be stable in time, enabling efficient readout. Here, using EEG decoding methods, we investigate how the geometry and dynamics of control representations constrain flexible action selection in the human brain. Participants performed a context-dependent action selection task. A forced response procedure probed action selection different states in neural trajectories. The result shows that before successful responses, there is a transient expansion of representational dimensionality that separated conjunctive subspaces. Further, the dynamics stabilizes in the same time window, with entry into this stable, high-dimensional state predictive of individual trial performance. These results establish the neural geometry and dynamics the human brain needs for flexible control over behavior.

Efficient Neural Decoding Based on Multimodal Training.

Neural decoding methods are often limited by the performance of brain encoders, which map complex brain signals into a latent representation space of perception information. These brain encoders are constrained by the limited amount of paired brain and stimuli data available for training, making it challenging to learn rich neural representations. To address this limitation, we present a novel multimodal training approach using paired image and functional magnetic resonance imaging (fMRI) data to establish a brain masked autoencoder that learns the interactions between images and brain activities. Subsequently, we employ a diffusion model conditioned on brain data to decode realistic images. Our method achieves high-quality decoding results in semantic contents and low-level visual attributes, outperforming previous methods both qualitatively and quantitatively, while maintaining computational efficiency. Additionally, our method is applied to decode artificial patterns across region of interests (ROIs) to explore their functional properties. We not only validate existing knowledge concerning ROIs but also unveil new insights, such as the synergy between early visual cortex and higher-level scene ROIs, as well as the competition within the higher-level scene ROIs. These findings provide valuable insights for future directions in the field of neural decoding.

Joint Power Allocation and Hybrid Beamforming for Cell-Free mmWave Multiple-Input Multiple-Output with Statistical Channel State Information.

Cell-free millimeter wave (mmWave) multiple-input multiple-output (MIMO) can effectively overcome the shadow fading effect and provide macro gain to boost the throughput of communication networks. Nevertheless, the majority of the existing studies have overlooked the user-centric characteristics and practical fronthaul capacity limitations. To solve these practical problems, we introduce a resource allocation scheme using statistical channel state information (CSI) for uplink user-centric cell-free mmWave MIMO system. The hybrid beamforming (HBF) architecture is deployed at each access point (AP), while the central processing unit (CPU) only combines the received signals by the large-scale fading decoding (LSFD) method. We further frame the issue of maximizing sum-rate subject to the fronthaul capacity constraint and minimum rate constraint. Based on the alternating optimization (AO) and fractional programming method, we present an algorithm aimed at optimizing the users' transmit power for the power allocation (PA) subproblem. Then, an algorithm relying on the majorization-minimization (MM) method is given for the HBF subproblem, which jointly optimizes the HBF and the LSFD coefficients.

An energy-saving distributed flexible job shop scheduling with machine breakdowns

Due to the change of consumption pattern, distributed flexible production is now becoming a major manufacturing mode. The machine breakdown seriously affects the production in workshops. How to efficiently adjust the scheduling scheme after machine breakdowns in distributed flexible job shop is an important concern for enterprises. Meanwhile, energy consumption is attracting more and more attention as environmental issues become increasingly serious. Thus, the energy-saving scheduling on distributed flexible job shop considering machine breakdowns is studied, and a mixed-integer programming model is established to optimize makespan, total energy consumption and machine load difference, simultaneously. Based on the problem characteristics, an improved memetic algorithm is designed. To improve the quality of initial solutions, an initialization strategy incorporating multiple rules is designed. An active decoding method with an adjustment strategy is adopted to fully utilize machine idleness. To expand the search range, the idea of dual-population collaborative optimization is used. Multi-level intelligent mutation operation strategies are introduced to further enhance the quality of solutions. In addition, an adaptive parameter rule is designed to adjust to appropriate search capabilities at different evolutionary stages. To verify the effectiveness of the proposed strategies and algorithm, the memetic algorithm, the artificial bee colony algorithm, the differential evolution algorithm and the estimation of distribution algorithm are chosen as comparison algorithms. The same iteration number and same running time are used as the termination criteria of all algorithms, and 60 instances of different scales are solved. The experimental results demonstrate that the effectiveness of the strategies and algorithm designed in this paper. Notably, our algorithm consistently achieved good results even when the same running time is used as the termination criterion.

A multi-objective genetic algorithm based on two-stage reinforcement learning for green flexible shop scheduling problem considering machine speed

The consumption of energy and resources in the manufacturing industry has garnered significant attention due to the increasingly severe environmental issues. Green shop scheduling research is focused on optimizing economic and environmental indicators within the current manufacturing model. This paper specifically addresses the flexibility of job-shop scheduling problem considering machine speed (CMS-FJSP), as different machine speeds during the production process can impact energy and resource consumption. The objectives of this study include minimizing the makespan, total energy consumption, and tool wear. To tackle this problem, a multi-objective genetic algorithm that incorporates a two-stage reinforcement learning approach is proposed. In light of the characteristics of the problem, a three-layer encoding approach is suggested, which encompasses machine allocation, operation sequencing, and machine speed selection. Additionally, a decoding method that integrates energy-saving strategies is proposed to enhance the optimization process. To improve the quality of the population, three distinct initialization methods have been developed. Furthermore, a parameter adjustment strategy informed by two-stage reinforcement learning is introduced. This strategy incorporates a state set and action set tailored to the unique characteristics of two-stage reinforcement learning, alongside corresponding reward mechanisms. In 30 test cases, the proposed algorithm demonstrates superior uniformity and convergence compared to five classical algorithms. In a practical case within a hydraulic component production workshop conducted at a hydraulic component company, the proposed algorithm generates 26 scheduling schemes with different focuses, achieving a 14.39% reduction in makespan, a 2.13% decrease in energy consumption, and a 10.65% reduction in tool wear.

BSP-Net: automatic skin lesion segmentation improved by boundary enhancement and progressive decoding methods

BSP-Net: automatic skin lesion segmentation improved by boundary enhancement and progressive decoding methods

A deep transfer learning approach for cross-subject recognition of mental tasks based on functional near-infrared spectroscopy

In the field of brain-computer interfaces (BCIs) based on functional near-infrared spectroscopy (fNIRS), traditional subject-specific decoding methods suffer from the limitations of long calibration time and low cross-subject generalizability, which restricts the promotion and application of BCI systems in daily life and clinic. To address the above dilemma, this study proposes a novel deep transfer learning approach that combines the revised inception-residual network (rIRN) model and the model-based transfer learning (TL) strategy, referred to as TL-rIRN. This study performed cross-subject recognition experiments on mental arithmetic (MA) and mental singing (MS) tasks to validate the effectiveness and superiority of the TL-rIRN approach. The results show that the TL-rIRN significantly shortens the calibration time, reduces the training time of the target model and the consumption of computational resources, and dramatically enhances the cross-subject decoding performance compared to subject-specific decoding methods and other deep transfer learning methods. To sum up, this study provides a basis for the selection of cross-subject, cross-task, and real-time decoding algorithms for fNIRS-BCI systems, which has potential applications in constructing a convenient and universal BCI system.

Estimating muscle force based on a neuromuscular decoding approach adaptive to fatigue conditions

Electromyography (EMG) has been extensively employed for precise muscle force estimation. Although methods based on motor unit (MU) activities (the smallest functional unit of the neuromuscular control system) have achieved superior performance, their effectiveness in fatiguing conditions remains unknown. This study aims to investigate the performance of existing MU-based muscle force estimation methods in fatiguing conditions, including the representative neuromuscular decoding (NMD) method and firing rate (FR) −based method. Compared to FR-based method, the NMD method specifically accounts for the differences in force contributions among MUs and the force twitch response of muscle fibers. We conducted fatigue induction experiment to collect force and high-density surface EMG data from the abductor pollicis brevis muscles of eight subjects during their performance of thumb abduction. Subsequently, the analysis of myoelectric manifestation of fatigue was performed and four training–testing strategies were designed to examine the effect of fatigue on force estimation methods. Experimental results indicated that the NMD method achieved superior and stable performance (assessed through root mean square difference between estimated and actual force) across four strategies (p >0.05 revealed by ANOVA), with only slight performance decline when fatigue data was involved in training or/and testing phase. Conversely, fatigue caused significant performance degradation when using the FR-based and other five comparative methods (p <0.05). These findings demonstrate that the NMD approach is adaptive to fatiguing MU activities, which can provide precise and robust muscle force estimation against fatigue in motor control and human–machine interfaces.

Pupil-Linked Arousal Modulates Precision of Stimulus Representation in Cortex.

Neural responses are naturally variable from one moment to the next, even when the stimulus is held constant. What factors might underlie this variability in neural population activity? We hypothesized that spontaneous fluctuations in cortical stimulus representations are created by changes in arousal state. We tested the hypothesis using a combination of fMRI, probabilistic decoding methods, and pupillometry. Human participants (20 female, 12 male) were presented with gratings of random orientation. Shortly after viewing the grating, participants reported its orientation and gave their level of confidence in this judgment. Using a probabilistic fMRI decoding technique, we quantified the precision of the stimulus representation in the visual cortex on a trial-by-trial basis. Pupil size was recorded and analyzed to index the observer's arousal state. We found that the precision of the cortical stimulus representation, reported confidence, and variability in the behavioral orientation judgments varied from trial to trial. Interestingly, these trial-by-trial changes in cortical and behavioral precision and confidence were linked to pupil size and its temporal rate of change. Specifically, when the cortical stimulus representation was more precise, the pupil dilated more strongly prior to stimulus onset and remained larger during stimulus presentation. Similarly, stronger pupil dilation during stimulus presentation was associated with higher levels of subjective confidence, a secondary measure of sensory precision, as well as improved behavioral performance. Taken together, our findings support the hypothesis that spontaneous fluctuations in arousal state modulate the fidelity of the stimulus representation in the human visual cortex, with clear consequences for behavior.

ECMSH: An Energy-efficient and Cost-effective data harvesting protocol for Mobile Sink-based Heterogeneous WSNs using PSO-TVAC

Efficient energy consumption is crucial in Wireless Sensor Networks (WSNs). Uncontrolled energy usage can lead to the hotspot issue, hindering network lifetime and successful packet delivery. Sink mobility has been suggested as a solution, but it comes with challenges such as high data gathering delay and poor packet reception. These problems stem from the short contact time of nodes with the Mobile Sink (MS). To tackle these issues, we present an MS-based heterogeneous WSN with super and normal nodes. Most previous studies only considered the energy heterogeneity of sensors. These methods also suffered from different issues such as fixed MS tours, including inappropriate criteria in cluster construction, proposing greedy schemes, and employing basic metaheuristic algorithms. In our proposed model, super nodes are richer in initial energy and transmission range than normal sensors. In each round, the nodes are organized into clusters, and the MS visits the Cluster Heads (CHs) to gather data packets. Super nodes, owing to their elevated initial energy, are more adept at executing energy-sensitive tasks compared to normal sensors. Additionally, as CH, super nodes extend the contact time with the MS due to their longer transmission range, delivering more packets. The clusters are constructed using a variant of Particle Swarm Optimization (PSO), namely PSO-TVAC. We empower this method with effective initialization and decoding methods. Furthermore, we propose a heuristic intra-cluster multi-hop routing algorithm to enhance network lifetime. Our other contribution is to propose an efficient algorithm to determine the time to reconfigure the network, while the other algorithms mainly reconfigure the WSN periodically. Simulation results demonstrate superior performance compared to state-of-the-art algorithms, showcasing lower energy consumption, higher energy efficiency, higher lifetime, reduced packet delivery delay, and higher number of received packets by 30%, 38.2%, four times, 20.6%, and 22.6%, respectively.

Enhancement of English-Bengali Machine Translation Leveraging Back-Translation

An English-Bengali machine translation (MT) application can convert an English text into a corresponding Bengali translation. To build a better model for this task, we can optimize English-Bengali MT. MT for languages with rich resources, like English-German, started decades ago. However, MT for languages lacking many parallel corpora remains challenging. In our study, we employed back-translation to improve the translation accuracy. With back-translation, we can have a pseudo-parallel corpus, and the generated (pseudo) corpus can be added to the original dataset to obtain an augmented dataset. However, the new data can be regarded as noisy data because they are generated by models that may not be trained very well or not evaluated well, like human translators. Since the original output of a translation model is a probability distribution of candidate words, to make the model more robust, different decoding methods are used, such as beam search, top-k random sampling and random sampling with temperature T, and others. Notably, top-k random sampling and random sampling with temperature T are more commonly used and more optimal decoding methods than the beam search. To this end, our study compares LSTM (Long-Short Term Memory, as a baseline) and Transformer. Our results show that Transformer (BLEU: 27.80 in validation, 1.33 in test) outperforms LSTM (3.62 in validation, 0.00 in test) by a large margin in the English-Bengali translation task. (Evaluating LSTM and Transformer without any augmented data is our baseline study.) We also incorporate two decoding methods, top-k random sampling and random sampling with temperature T, for back-translation that help improve the translation accuracy of the model. The results show that data generated by back-translation without top-k or temperature sampling (“no strategy”) help improve the accuracy (BLEU 38.22, +10.42 on validation, 2.07, +0.74 on test). Specifically, back-translation with top-k sampling is less effective (k=10, BLEU 29.43, +1.83 on validation, 1.36, +0.03 on test), while sampling with a proper value of T, T=0.5 makes the model achieve a higher score (T=0.5, BLEU 35.02, +7.22 on validation, 2.35, +1.02 on test). This implies that in English-Bengali MT, we can augment the training set through back-translation using random sampling with a proper temperature T.

Benchmarking brain–computer interface algorithms: Riemannian approaches vs convolutional neural networks

Objective.To date, a comprehensive comparison of Riemannian decoding methods with deep convolutional neural networks for EEG-based brain-computer interfaces remains absent from published work. We address this research gap by using MOABB, The Mother Of All BCI Benchmarks, to compare novel convolutional neural networks to state-of-the-art Riemannian approaches across a broad range of EEG datasets, including motor imagery, P300, and steady-state visual evoked potentials paradigms.Approach.We systematically evaluated the performance of convolutional neural networks, specifically EEGNet, shallow ConvNet, and deep ConvNet, against well-established Riemannian decoding methods using MOABB processing pipelines. This evaluation included within-session, cross-session, and cross-subject methods, to provide a practical analysis of model effectiveness and to find an overall solution that performs well across different experimental settings.Main results.We find no significant differences in decoding performance between convolutional neural networks and Riemannian methods for within-session, cross-session, and cross-subject analyses.Significance.The results show that, when using traditional Brain-Computer Interface paradigms, the choice between CNNs and Riemannian methods may not heavily impact decoding performances in many experimental settings. These findings provide researchers with flexibility in choosing decoding approaches based on factors such as ease of implementation, computational efficiency or individual preferences.

Macroscopic brain dynamics beyond contralateral primary motor cortex for movement prediction

This study investigates the complex relationship between upper limb movement direction and macroscopic neural signals in the brain, which is critical for understanding brain-computer interfaces (BCI). Conventional BCI research has primarily focused on a local area, such as the contralateral primary motor cortex (M1), relying on the population-based decoding method with microelectrode arrays. In contrast, macroscopic approaches such as electroencephalography (EEG) and magnetoencephalography (MEG) utilize numerous electrodes to cover broader brain regions. This study probes the potential differences in the mechanisms of microscopic and macroscopic methods. It is important to determine which neural activities effectively predict movements. To investigate this, we analyzed MEG data from nine right-handed participants while performing arm-reaching tasks. We employed dynamic statistical parametric mapping (dSPM) to estimate source activity and built a decoding model composed of long short-term memory (LSTM) and a multilayer perceptron to predict movement trajectories. This model achieved a high correlation coefficient of 0.79 between actual and predicted trajectories. Subsequently, we identified brain regions sensitive to predicting movement direction using the integrated gradients (IG) method, which assesses the predictive contribution of each source activity. The resulting salience map demonstrated a distribution without significant differences across motor-related regions, including M1. Predictions based solely on M1 activity yielded a correlation coefficient of 0.42, nearly half as effective as predictions incorporating all source activities. This suggests that upper limb movements are influenced by various factors such as movement coordination, planning, body and target position recognition, and control, beyond simple muscle activity. All of the activities are needed in the decoding model using macroscopic signals. Our findings also revealed that contralateral and ipsilateral hemispheres contribute equally to movement prediction, implying that BCIs could potentially benefit patients with brain damage in the contralateral hemisphere by utilizing brain signals from the ipsilateral hemisphere. In conclusion, this study demonstrates that macroscopic activity from large brain regions significantly contributes to predicting upper limb movement. Non-invasive BCI systems would require a comprehensive collection of neural signals from multiple brain regions.