Search Results (1,087)

This article proposes a prediction model that combines network physical characteristics and deep learning methods to solve the problem of insufficient round-trip time (RTT) prediction accuracy caused by complex dynamic characteristics in civil aviation business data networks. This model is based on a physical information neural network framework, which embeds domain knowledge such as delay load relationships, burst traffic attenuation patterns, and inverse RTT window constraints into a long short-term memory (LSTM) network. In addition, the multi constraint loss function enhances the adaptability of the model to complex network activities. Moreover, in order to overcome the limitations of LSTM networks in modeling long-range dependencies, a multi head attention (MA) mechanism was implemented to capture long-term temporal dependencies in parallel, thereby improving the model’s ability to capture long-range time step correlations. Benchmarking and extension tests were conducted on four civil aviation business data network datasets. The experimental results show that compared with the baseline model, the proposed model significantly improves the root mean square error (RMSE), mean absolute error (MAE), and mean absolute percentage error (MAPE) for bidirectional network communication RTT prediction tasks. The study has verified that combining network physical attribute constraints with attention mechanisms can effectively improve the accuracy of transmission delay prediction, providing an effective method for effective traffic prediction in highly dynamic civil aviation network environments. Full article

Background and Objectives: Pediatric meningitis remains a major diagnostic and therapeutic emergency because clinical presentation may be nonspecific and early etiological distinction directly influences treatment, infection control, and epidemiological surveillance. Although bacterial meningitis has traditionally been emphasized because of its severity, the wider use of rapid molecular methods has increased the detection of viral etiologies. This study aimed to evaluate the etiological distribution of confirmed pediatric meningitis episodes and to assess temporal, age-related, and sex-associated patterns, with particular attention to the observed shift from bacterial toward viral meningitis. Materials and Methods: A retrospective single-center analysis was performed using microbiological data from suspected meningitis episodes investigated between 2023 and 2025 at the “Louis Țurcanu” Emergency Hospital for Children, Timișoara. Confirmed episodes were classified as bacterial, viral, or fungal according to the identified etiological agent. The distribution of suspected and confirmed episodes, annual etiological patterns, age-group differences, sex-associated variation, and identified pathogens were assessed using descriptive statistics and trend analysis. Results: Among 641 suspected episodes, 58 were microbiologically confirmed, corresponding to an overall positivity rate of 9.05%. Viral meningitis was the most frequent etiology, accounting for 30 episodes (51.72%), followed by bacterial meningitis with 22 episodes (37.93%) and fungal meningitis with 6 episodes (10.34%). Etiological distribution changed significantly over the study period: viral episodes increased from 22.22% in 2023 to 77.78% in 2025, while bacterial episodes decreased from 66.67% to 22.22%. Viral meningitis predominated in children aged 1–6 years and was more frequent among male patients, whereas bacterial meningitis remained more frequent in infants. Enterovirus became the dominant identified agent in 2025. Conclusions: The local epidemiology of confirmed pediatric meningitis showed an observed redistribution among confirmed episodes toward viral etiologies, particularly enteroviral meningitis, alongside a proportional decline in bacterial episodes. These findings support continued etiological surveillance and the integration of rapid molecular testing with conventional microbiology in the diagnostic management of suspected pediatric meningitis. Full article

Predictive maintenance of electric vehicle (EV) fleets requires accurate estimation of Remaining Useful Life (RUL), Time-to-Failure (TTF), and State-of-Health (SOH) from heterogeneous Internet of Things (IoT) telemetry. However, real-world degradation patterns are nonlinear, nonstationary, and highly imbalanced near failure. This study proposes Q-TRACNet, a temporal attention framework that combines causal maintenance-aware preprocessing, adaptive temporal condensation, residual refinement, learnable phase modulation, and hybrid Particle Swarm Optimization–Quantum-Guided Descent parameter tuning. The framework is evaluated on the EV-HLM-RUL dataset and three established prognostics benchmarks: NASA CMAPSS, PHM 2012, and XJTU-SY. Chronological training, validation, and testing partitions are used to preserve temporal causality. On EV-HLM-RUL, Q-TRACNet achieves an MAE of 9.8, an RMSE of 14.7, an $R^2$ of 0.979, and a Critical Degradation Awareness Index (CDAI) of 0.91. It reduces RMSE by 20.11% relative to the strongest competing baseline and achieves an NRMSE of 0.102 and a Kendall correlation of 0.89 ($p<{10}^{-4}$). Cross-dataset experiments demonstrate stable performance for RUL, TTF, and short- and long-horizon SOH prediction. Ablation and sensitivity analyses further confirm the contributions of the temporal and attention components and the stability of degradation-aware evaluation. Q-TRACNet also provides lower training cost and inference latency than competing architectures, supporting practical maintenance planning, inspection prioritization, and resource allocation in connected EV fleets. Full article

Due to the explosive growth in cloud-native infrastructures, the attack surface has dramatically increased in modern enterprise identity systems, where privilege escalation has become a major security risk. Conventional rule-based intrusion detection systems fall short in identifying multi-hop privilege inheritance paths and lateral movements over heterogeneous and dynamic identity graphs. This study introduces PEGraphSec-Net, a graph-theoretical framework for detecting privilege-escalation-relevant identity behavior, modeling cloud identity interactions as dynamic heterogeneous graphs of users, services, roles, tokens, and workloads. The core contribution of this framework is a graph-based detection pipeline—an Identity Relationship Graph Constructor, a Privilege-Escalation Path Encoder, and a Temporal Graph Attention Detection layer—evaluated on privilege-escalation-relevant attack categories using a documented proxy identity-graph construction derived from the UNSW-NB15 network-traffic benchmark, and benchmarked against six non-graph tabular classifiers (CNN, LightGBM, XGBoost, Random Forest, SVM, and MLP) trained under identical preprocessing; this pipeline achieves 98.78% accuracy, a weighted F1-score of 0.98692 (macro F1-score of 0.91828), and an AUC of 1.000 on the held-out test partition. PEGraphSec-Net is further benchmarked against three graph neural network baselines (GCN, GAT, and GraphSAGE) trained on the identical identity-graph topology and node attributes; all three substantially underperform PEGraphSec-Net (best case, GraphSAGE: 63.66% accuracy, 0.239 macro F1-score), indicating that a large share of PEGraphSec-Net’s performance derives from its explicit privilege-path encoding and temporal attention mechanisms rather than from the graph topology alone. An Adaptive Containment and Isolation Engine and a Mitigation Policy Reinforcement Optimizer are further proposed as risk-scoring and reward-driven policy-learning components, whose contribution is validated through module-wise ablation on classification performance; live containment action and reinforcement-learning-specific evaluation are left for future validation. The term “privilege escalation” is used throughout to denote the evaluated proxy attack categories (Exploits, Backdoor/Backdoors, and Reconnaissance) under a documented, decade-old (2015) network-intrusion benchmark, rather than production cloud-native IAM behavior, for which native-dataset validation remains an open direction. SHAP-based interpretability analysis links the model’s top-ranked traffic-level features back to the identity-graph risk, role, and trust-transition attributes they populate, evidencing that the learned representation captures semantically meaningful identity-behavior patterns within this proxy setting. Full article

Elevator door systems are critical and failure-prone subsystems in elevators, with vibration signals characterized by strong nonlinearity, non-stationarity, and complex temporal dependencies under varying operating conditions. Existing fault diagnosis methods often struggle to effectively model these characteristics, particularly in capturing cross-scale dependencies between local transient disturbances and global temporal evolution patterns. To address this issue, a cross-scale spatiotemporal feature learning framework integrating adaptive time-frequency decomposition and deep feature fusion is proposed. Variational mode decomposition (VMD) is applied to decompose nonlinear vibration signals into intrinsic mode functions, while fast Fourier transform (FFT) is used to extract global spectral information, forming a multi-channel time-frequency representation. The convolutional neural network (CNN) module is designed to extract local transient features associated with mechanical impacts, while the bidirectional long short-term memory (BiLSTM) module models long-term temporal dependencies. Furthermore, a cross-attention mechanism is introduced to dynamically fuse local and global representations, enhancing discriminative feature learning under complex operating conditions. Experimental results on a representative elevator door fault dataset show that the proposed approach attained an average accuracy of 99.18% and an average F1-score of 99.13%. Compared with conventional machine learning and neural network-based methods, the proposed framework demonstrates superior diagnostic accuracy and robust performance under the evaluated experimental conditions. The findings indicate that the proposed approach serves as an efficient strategy for cross-scale spatiotemporal modeling of nonlinear and non-stationary vibration signals, with considerable promise for intelligent health monitoring and prognostic maintenance of elevator door systems. Full article

The rapid expansion of road construction has significantly contributed to economic development and regional connectivity. However, linear infrastructure such as roads, railways, and utility corridors has also introduced considerable ecological disruptions. Although increasing global attention is being paid to mitigating these effects, most existing research primarily focuses on quantifying and monitoring external environmental changes (e.g., landscape structure or vegetation coverage) while often neglecting internal ecological dynamics such as ecosystem resilience. Previous road-ecology studies have extensively examined road-buffer effects, land-use/land-cover changes, vegetation-index dynamics, and landscape fragmentation. Therefore, the contribution of this study does not lie in proposing an entirely new class of indicators. Rather, it lies in applying a combined external–internal assessment framework to a recently constructed expressway corridor by jointly examining annual land-cover transitions and leaf area index (LAI)-derived temporal variability/resilience indicators across multiple distance buffers and spatial resolutions. This design allows us to compare whether structural land-cover changes and vegetation time-series responses show similar distance–decay patterns around the expressway corridor. The results show that: (1) Land-cover transformation was mainly concentrated within the first 500–1000 m from the expressway, especially for impervious surface expansion and vegetation loss. Multi-indicator distance-gradient analysis showed that land-cover change intensity and LAI-derived variability indicators gradually approached the distal reference condition at approximately 2000 m, which was therefore used as an empirical corridor-analysis boundary rather than a definitive ecological impact threshold. (2) Within the 2000 m buffer zone, forest area increased from 21.866 km² in 2001 to 45.370 km² in 2023, while impervious surface area increased from 3.016 km² to 6.869 km². During the construction and early operation period from 2018 to 2023, impervious surface area increased from 6.268 km² to 6.869 km², indicating localized artificial surface expansion along the expressway corridor. (3) During 2018–2023, the 30 m LAI product showed a 22.3% increase in coefficient of variation (CV), indicating enhanced relative LAI variability. In contrast, temporal autocorrelation (TAC) did not show the consistent increase expected under classical critical slowing down theory, suggesting that TAC-based evidence for resilience decline was weak or inconclusive during this short period. The observed TAC/CV changes were interpreted as critical slowing down (CSD)-related vegetation variability signals, rather than as a distinct or definitive critical slowing down signature. (4) The multi-resolution comparison showed weak pixel-level correspondence between the 30 m and 250 m LAI products, indicating clear scale dependence rather than robust multi-scale consistency. The 250 m data were useful for characterizing long-term regional background trends, whereas the 30 m data were more suitable for detecting localized corridor-scale vegetation variability. Thus, the multi-resolution analysis should be regarded as a scale-sensitivity assessment rather than as direct cross-scale validation. Full article

Time-series analysis is important for various scientific and industrial fields, such as remote sensing where observations may be disturbed by clouds, have irregular revisits and different sensors. Current transformer-based approaches have enhanced the modeling of distant time relationships, but they are still constrained by specific task designs and poor adaptability to various time-series patterns. To solve these problems, we present TimeHome, a universal sparse transformer basic model for handling heterogeneous time series. TimeHome incorporates a Heterogeneous Mixture-of-Experts (H-MoE) component, where different expert types are chosen dynamically based on a low-rank temperature-controlled gating mechanism to fit various sequence features. Moreover, a hybrid local–global attention mechanism is designed to consider both short-term variations and long-distance correlations, while specific heads are used for unified prediction, missing value estimation and abnormal event detection. TimeHome is pretrained on TS-200B, a huge database of time series including various temporal patterns from different domains. Comprehensive tests on several benchmark datasets and remote sensing extended evaluations show that TimeHome performs well in long-term prediction, missing value replacement and abnormal event detection. The model also exhibits good zero-shot adaptation ability and fast inference speed by adjusting experts dynamically. The source code and pre-training data will be released publicly. Full article

Fruit preharvest–postharvest quality assessment is essential for precision harvesting, intelligent grading, storage management, and supply-chain loss reduction. However, conventional approaches mainly rely on single-point postharvest inspection and cannot adequately capture the long-term effects of preharvest fruit phenotypes and environmental dynamics on quality formation. To address limited prediction accuracy under few-label conditions, insufficient multimodal fusion, and weak cross-orchard generalization, this study proposes FruitSSL-QNet, a self-supervised multimodal learning framework for jointly modeling preharvest fruit images, environmental sensor time series, and postharvest quality indicators. The framework employs visual masked reconstruction to learn fine-grained phenotype features, including color, texture, lenticel distribution, disease spots, and maturity patterns. Environmental temporal masked modeling is used to capture the cumulative effects of temperature, humidity, light intensity, soil moisture, and rainfall. Bidirectional cross-attention, gated fusion, and contrastive alignment are further integrated to learn complementary and semantically consistent image–environment representations. Experimental results demonstrate that FruitSSL-QNet outperforms SVM, Random Forest, XGBoost, LSTM, GRU, TCN, Transformer, and MM-Transformer across multiple quality assessment tasks. The proposed model achieves a maturity recognition accuracy of $89.6\%$, exceeding MM-Transformer by $4.4$ percentage points. Compared with the corresponding baseline results, the prediction errors for sugar content, firmness, and shelf life are reduced by $21.1\%$, $22.2\%$, and $22.5\%$, respectively. The decay-risk AUC reaches $0.921$, and the cross-site F1-score reaches $0.867$, indicating strong risk discrimination and stable generalization across orchard environments. Ablation experiments further confirm the contributions of visual self-supervision, environmental temporal self-supervision, and cross-modal alignment. Full article

Background: Social cognition research increasingly emphasizes cognitive processes in social contexts, with growing attention to socially shared retrieval-induced forgetting (SS-RIF). Behavioral evidence suggests that the mechanisms of SS-RIF are fundamentally aligned with those of retrieval-induced forgetting (RIF). However, SS-RIF differs from RIF due to the influence of social interaction factors. This study investigated the cognitive neural mechanisms of SS-RIF, focusing on its commonalities and distinctions from RIF. Methods: Experiment 1 examined neural evidence associated with retrieval-related processes. Twenty-eight university students participated in a within-subject design including retrieval practice, social retrieval practice, and baseline conditions. Correct recall rates and P2/FN400 amplitudes were measured. Experiment 2 investigated the temporal dynamics of inhibitory mechanisms by narrowing the design to SS-RIF only, using a within-participant design with first, second, and third repetition blocks occurring within the social retrieval phase, and recall was tested across two days. Results: Classic RIF and SS-RIF effects were observed, accompanied by neural activity indicative of inhibition during both retrieval practice and social retrieval conditions. These findings are consistent with the hypothesis that SS-RIF and RIF may share overlapping underlying cognitive operations. Typical RIF and SS-RIF phenomena were also observed across repetition conditions, while neural data showed no significant differences in P2 and FN400 amplitudes for SS-RIF, highlighting a distinction from patterns observed in RIF. Conclusions: Cognitive suppression accounts offer a compelling framework for explaining SS-RIF during social retrieval practice. However, compared with RIF, the duration of inhibition differs in SS-RIF. Unlike RIF, SS-RIF shows no reduction in specific EEG signal amplitudes over time. These findings provide cognitive neural evidence for inhibitory mechanisms in social retrieval and preliminary insights into the shared and distinct mechanisms of socially shared and individual RIF. Full article

The carotid and vertebral arteries are critical components of the cerebral circulation, supplying the brain and spinal cord. Their embryologic development involves complex transformations of the aortic arches, paired dorsal aortae, and cervical intersegmental arteries. Congenital anomalies and variations involving their origin, course, and branching patterns are relatively common and are increasingly recognized using modern imaging. While many of these variants are incidental, some carry important clinical implications, particularly in contexts of stroke imaging, head and neck surgery, and neurointerventional procedures. This article reviews the embryologic development of the carotid and vertebral arterial systems, highlights normal and variant anatomy, and presents a case-based, illustrated overview of these congenital anomalies. Emphasis is placed on key findings utilizing computed tomography angiography (CTA), magnetic resonance angiography (MRA), and catheter angiography. Particular attention is given to structural and pathway variations, including vessel agenesis or hypoplasia, the aberrant internal carotid artery within the temporal bone, abnormal branching patterns, fenestrations, and persistent fetal anastomoses (such as the primitive trigeminal, hypoglossal, and proatlantal intersegmental arteries). Ultimately, familiarity with these anomalies, their embryological origins, and their distinct imaging characteristics is essential for ensuring accurate interpretation and avoiding diagnostic pitfalls and procedural complications. Full article

Background/Objectives: Cross-modal emotional priming effects of music on language have been widely demonstrated, but how musical emotion interacts with different types of conceptual representation remains unclear. This study examined whether emotional musical contexts exert differential influences on the processing of concrete and abstract words, whose processing is thought to depend to different degrees on sensorimotor and affective experience. Methods: Two ERP experiments were conducted using a long stimulus-onset asynchrony priming paradigm with two independent groups of participants. In Experiment 1, 24 female participants performed valence judgments on concrete words following positive or negative musical primes. In Experiment 2, 24 female participants completed the same task with abstract words. ERP components associated with early perceptual or attentional processing, semantic processing, and later evaluative or integrative processing were analyzed. Results: Musical emotion modulated the processing of both concrete and abstract words, but the temporal patterns differed. For concrete words, music–word valence-congruency effects were mainly observed following negative music primes, with negative words eliciting larger P2 and late positive component amplitudes than positive words. Musical context also modulated the N400 amplitudes, suggesting effects across multiple processing stages. For abstract words, no reliable interaction was observed in the P2 or N400 time windows. Instead, a late-stage valence effect emerged mainly following positive music primes, with negative words eliciting larger late positive component amplitudes than positive words. Conclusions: These findings suggest that musical emotion interacts with concrete and abstract word processing through distinct temporal patterns. The results offer preliminary support for the contextual compensation hypothesis, suggesting that concrete words may be more sensitive to external musical contexts during emotional processing, whereas abstract words appear to rely more on later-stage emotional integration. These findings therefore highlight context-dependent effects of musical emotion and music–word valence congruency on concrete and abstract word processing. Full article

Accurate solar radiation forecasting is essential for photovoltaic energy generation, smart grid stability, and renewable energy management. This study proposes a lightweight hybrid deep learning framework that combines a transformer encoder and Gated Rrecurrent Uunit (GRU) network for short-term solar radiation forecasting in Makkah and Madinah, Saudi Arabia. Hourly meteorological data from the NASA POWER dataset (2020–2025) were utilized, including solar radiation intensity, temperature, humidity, wind speed, cloud amount, rainfall, surface pressure, and dew point temperature. A preprocessing pipeline consisting of missing value treatment, outlier removal, normalization, timestamp alignment, and data cleaning was applied to improve data quality. Feature engineering techniques were incorporated to capture temporal dependency, meteorological interactions, weather dynamics, and solar variability patterns. The transformer encoder was used to learn long-range temporal dependencies through multi-head self-attention, while the GRU layer modeled sequential temporal dynamics efficiently. Hyperparameter optimization was performed using Bayesian optimization with Optuna. The experimental results demonstrate that the proposed transformer GRU framework achieved a Mean Absolute Error (MAE) of 0.014, Root Mean Square Error (RMSE) of 0.0219, and a coefficient of determination (${\mathrm{R}}^2$) of 0.98. The proposed model outperformed ARIMA, LSTM, GRU, and XGBoost models while maintaining stable performance across varying weather conditions and forecasting horizons. Full article

Background/Objectives: Speech disorders caused by neurological, articulatory, cognitive, and behavioral impairments require accurate multimodal analysis frameworks capable of jointly understanding acoustic abnormalities and neurolinguistic inconsistencies for reliable computer-assisted speech disorder assessment. Recent multimodal deep learning approaches have utilized speech signals, linguistic transcripts, attention mechanisms, and contextual fusion strategies to improve disordered speech analysis and intelligent disorder prediction. However, existing methods frequently suffer from insufficient temporal synchronization, ineffective local–global feature learning, multimodal redundancy, poor interpretability, and reduced robustness under heterogeneous speech disorders. To address these challenges, Methods: The proposes NSX-Net (NeuroSpeech Explainable Network), an intelligent multimodal deep learning framework for speech disorder classification using Acoustic Speech Signal Data and Neurolinguistic Text/Linguistic Data. Initially, the Adaptive Speech Refinement Module (ASRM) performs noise removal, silence elimination, signal normalization, spectrogram generation, MFCC extraction, and transcript preprocessing to improve multimodal speech consistency. Subsequently, the Hierarchical Multimodal Feature Learning Unit (HMFLU) extracts discriminative feature representations through the Cross-Domain Representation Encoder (CDRE), Fine-Grained Speech Pattern Analyzer (FGSPA), and Global Sequential Dependency Learner (GSDL) for capturing both local articulation abnormalities and long-range semantic dependencies. Furthermore, the Dual-Path Attention Enhancement Block (DPAEB) emphasizes clinically important disordered speech regions using adaptive local–global attention mechanisms, while the Temporal Resolution Synchronization Module (TRSM) aligns rhythm-level and beat-level multimodal contextual structures to improve temporal consistency and suppress noise disturbances. Results: Experimental evaluation across six publicly available multimodal speech disorder datasets demonstrated that NSX-Net achieved superior performance, with 99.27% Accuracy, 99.11% Precision, 98.97% Recall, 99.02% F1-Score, and 99.41% AUC, significantly outperforming existing state-of-the-art frameworks. Conclusions: Finally, optimized multimodal contextual representations are forwarded into a Softmax classification layer for intelligent multiclass speech disorder prediction with improved interpretability and potential clinical applicability. Full article

Skeleton-based human action recognition is an important problem in applied vision systems, yet many existing approaches depend on a single skeleton descriptor or a single feature-learning mechanism. This restriction can weaken the representation of local body kinematics, long-range joint relations, and temporal dependencies within an action sequence. To address these limitations, this paper proposes HILF-GT (Hybrid Invariant Latent Feature Graph Transformer), a hybrid Graph Convolutional Network (GCN)-Transformer framework based on multiple spatio-temporal invariant latent features. The representation module constructs complementary structured tensors from skeleton graphs, inter-joint distances, adjacent-frame joint displacements, and inter-limb angles. Instead of transforming these descriptors into image-like maps for separate Convolutional Neural Network (CNN)-based classification, HILF-GT keeps their graph and temporal organization during learning. A local GCN branch models skeleton-aware kinematic patterns, whereas a graph-aware Transformer branch uses biased self-attention and cross-attention to capture dependencies among distant joints, frames, and latent-feature streams. A Perceiver-style latent bottleneck is further introduced to reduce the memory cost of global attention over frame-joint tokens. Experiments were conducted on four standard benchmark datasets, including NTU-RGB+D 60, NTU-RGB+D 120, NW-UCLA, and UTD-MHAD. The proposed method achieved 93.1% and 97.20% accuracy on the NTU-RGB+D 60 Cross-Subject and Cross-View protocols, 88.15% and 90.20% on the NTU-RGB+D 120 Cross-Subject and Cross-Setup protocols, 98.50% on NW-UCLA, and 97.50% on UTD-MHAD. Full article

Functional near-infrared spectroscopy (fNIRS) is a portable, non-invasive tool for studying cortical function in children with neurodevelopmental and neurological disorders. Although fNIRS use is increasing, heterogeneous study paradigms and cortical targets have limited cross-condition comparisons and the identification of shared research priorities. This systematic review maps cortical regions and reporting patterns in five key conditions—autism spectrum disorder (ASD), attention-deficit/hyperactivity disorder (ADHD), cerebral palsy (CP), hypoxic–ischemic encephalopathy (HIE), and epilepsy (Ep)—from January 2015 to December 2025. A systematic search across five databases identified 72 relevant studies meeting PRISMA 2020 criteria, revealing both similarities and differences across disorders. The prefrontal cortex (PFC) was studied in all five conditions (5/5: 100%), making it the most consistently investigated cortical region. The parietal and temporal cortices were studied in 4 of 5 conditions (80%). The frontal cortex had the most regions investigated, while the temporal cortex showed the most consistent coverage across conditions (2.25 conditions per region). Beyond regional preferences, condition-specific patterns were aligned with their disorder phenotypes: social brain networks in ASD, prefrontal executive systems in ADHD, sensorimotor changes in CP, cerebrovascular monitoring in HIE, and state-dependent changes in Ep. Across conditions, researchers found altered prefrontal activity, disrupted connectivity, and compensatory brain responses, supporting broader frameworks. Collectively, these findings identify the PFC as a shared target for transdiagnostic fNIRS investigations while highlighting important gaps in regional coverage and methodological consistency. Despite methodological differences, fNIRS shows promise for identifying both shared and unique brain patterns in pediatric neurodevelopmental and neurological disorders. This review provides a framework for prioritizing cortical targets and guiding future standardized fNIRS research in pediatric populations. Full article