The rapid growth of cities demands sophisticated decision support systems capable of simultaneously tackling social, economic, and environmental issues in urban planning. This research introduces a comprehensive decision-making framework that integrates machine learning to optimize urban infrastructure development. To improve the reliability of the analysis, recursive feature elimination paired with random forest (RF-RFE) is utilized to pinpoint the most critical factors, ensuring that the decision process emphasizes the features that matter most. To handle uncertainty and vagueness in expert evaluations, picture fuzzy sets are implemented, allowing a nuanced expression of hesitation, while the entropy method is employed to objectively determine the weights of social, economic, and environmental criteria, thereby reducing subjective influence. The evaluation based on the ap proach of relative utility and nonlinear standardization (ERUNS) is employed to rank urban planning alternatives, capturing the nonlinear interactions among criteria. This framework provides a comprehensive, data-driven deci sion support tool for urban planners and policymakers to address the intricate and dynamic challenges of modern urbanization.

Biogas-driven combined cooling and power (CCP) systems face the challenge of simultaneously optimizing thermodynamic performance and financial viability under nonlinear design and operational constraints. This study proposes a novel biogas combustion-heat recovery configuration for CCP generation, evaluated through an integrated thermodynamic-financial framework and optimized using machine learning (ML)-driven softcomputing techniques. The system integrates a biogas combustion unit, a gas turbine, a modified supercritical CO2 cycle, and a generator-absorber-exchanger (GAX) cycle. Thermodynamic analyses based on the first and second laws of thermodynamics are employed, while sustainability, financial, and environmental indicators are incorporated into the assessment. A hybrid optimization approach, combining ML with the genetic algorithm optimizer, is implemented to accelerate convergence and explore trade-offs among net present value (NPV), total unit product cost (TUPC), and sustainability index (SI). The optimized configuration achieves an NPV of 13.03 M $, an SI of 1.765, and a TUPC of 26.5 $/GJ. Besides, the system demonstrates an energy efficiency of 62.75%, an exergy efficiency of 43.32%, and a payback period of 3.79 years, confirming technical robustness and economic viability. Overall, ML-driven soft computing enables resilient, investment-ready CCP strategies, offering a scalable plan that aligns biogas utilization with sustainability, efficiency, and competitiveness.

With growing urbanization, there are increasing demands on waste management systems that can be performed in an environmentally friendly way as well as efficiently. Current approaches to managing waste often have issues with efficiency, transparency, and engaging with the public. Blockchain technology has been identified as one potential solution to these problems because it offers several benefits including decentralization, security, and transparency. The selection of the best blockchain-based waste management (BBWM) system is very difficult due to the many different evaluation criteria that may conflict with each other. Therefore this research uses a multi-criteria decision making (MCDM) approach using CIMAS (Criteria Importance Assessment), for determining weights based upon subjective input, and LOPCOW (Logarithmic Percentage Change-Driven Objective Weighing), for determining weights based upon objective data within the MCDM framework. To rank alternatives effectively, an Alternative Ranking Order Method Accounting for Two-Step Normalization (AROMAN) technique is applied, ensuring a precise evaluation process. The use of T-Spherical Fuzzy Sets (T-SFS) captures all three (membership, non-membership, hesitation degree) and is used to address the variability that exists when making an expert judgment. Some of the key factors include; Technological Feasibility, Operational Costs, Scalability, Data Security, Regulatory Compliance, Environmental Impact. Based on the evaluation criteria, it appears that the Blockchain Enabled Waste Tracking System is the most appropriate alternative due to its high potential for Transparency, Regulatory Compliance and Fraud Prevention. In addition, this research will provide Policymakers, Urban Planners and Investors with a methodical way of making Data Driven Decisions on BBWM Investments.

Background/Objectives: The earlier, more accurate, and more consistent prediction of the brain tumor recognition process requires automated systems to minimize diagnostic delays and human error. The automated system provides a platform for handling large medical images, speeding up clinical decision-making. However, the existing system is facing difficulties due to the high variability in tumor location, size, and shape, which leads to segmentation complexity. In addition, glioma-related tumors infiltrate the brain tissues, making it challenging to identify the exact tumor region. Method: The above-identified research difficulties are overcome by applying the Swin-UNet with cuttlefish-optimized attention-based Graph Neural Networks (SCAG-Net), thereby improving overall brain tumor recognition accuracy. This integrated approach is utilized to address infiltrative gliomas, tumor variability, and feature redundancy issues by improving diagnostic efficiency. Initially, the collected MRI images are processed using the Swin-UNet approach to identify the region, minimizing prediction error robustly. The region’s features are explored utilizing the cuttlefish algorithm, which minimizes redundant features and speeds up classification by improving accuracy. The selected features are further processed using the attention graph network, which handles structural and heterogeneous information across multiple layers, improving classification accuracy compared to existing methods. Results: The efficiency of the system, implemented with the help of public datasets such as BRATS 2018, BRATS 2019, BRATS 2020, and Figshare is ensured by the proposed SCAG-Net approach, which achieves maximum recognition accuracy. The proposed system achieved a Dice coefficient of 0.989, an Intersection over Union of 0.969, and a classification accuracy of 0.992. This performance surpassed the most recent benchmark models by margins of 1.0% to 1.8% and with statistically significant differences (p < 0.05). These findings present a statistically validated, computationally efficient, clinically deployable framework. Conclusions: The effective analysis of MRI complex structures is used in medical applications and clinical analysis. The proposed SCAG-Net framework significantly improves brain tumor recognition by addressing tumor heterogeneity and infiltrative gliomas using MRI images. The proposed approach provides a robust, efficient, and clinically deployable solution for brain tumor recognition from MRI images, supporting accurate and rapid diagnosis while maintaining expert-level performance.

Objective The study's overarching goal is to improve E-health monitoring systems' precision and performance by developing and implementing an Rendezvous Data Processing Model (RDPM) that is compatible with IoT-cloud architecture. The approach solves a problem with current E-health systems; these systems frequently make incorrect or redundant suggestions because they depend too much on static analytic methods and isolated data augmentation.Methods The RDPM system recommended improves real-time decision-making by digesting historical suggestions and present analytical flaws. Divided features and data streams allow it to validate new hypotheses by comparing them to earlier observations. The state learning process has been improved by earlier efforts to avoid errors and data duplication, the model must distinguish intervening and non-intervening data. Internet-connected sensors collect massive volumes of patient and environment data. Cloud analytics evaluates the system's precision using these data.Results Experimental results show that RDPM reduces data interruptions, analytical errors, and recommendation ratios while improving decision correctness. The model shows that it can quickly interpret many input streams without compromising accuracy. Compared to IoT-based healthcare analytics, the RDPM improves suggestion accuracy and reduces computing redundancy.Conclusion IoT-cloud technologies with the RDPM system establish an adaptive and scalable platform for sophisticated E-health monitoring. State learning and dynamic data validation allow RDPM to make more accurate and convenient health recommendations. This approach allows a healthcare system to self-improve, understand context, and manage massive, real-time datasets.

Environmental problems are intensifying due to the rapid growth of the population, industry, and urban infrastructure. This expansion has resulted in increased air and water pollution, intensified urban heat island effects, and greater runoff from parks and other green spaces. Addressing these challenges requires prioritizing green infrastructure and other sustainable urban development strategies. This study introduces a novel Integrated Decision Support System that combines Pythagorean Fuzzy Sets with the Advanced Alternative Ranking Order Method allowing for Two-Step Normalization (AAROM-TN), enhanced by a dual weighting strategy. The weighting approach integrates the Criteria Importance Through Intercriteria Correlation (CRITIC) method with the Criteria Importance through Means and Standard Deviation (CIMAS) technique. The originality of the proposed framework lies in its ability to objectively quantify criteria importance using CRITIC, incorporate decision-makers' preferences through CIMAS, and capture the uncertainty and hesitation inherent in human judgment via Pythagorean Fuzzy Sets. A case study evaluating green infrastructure alternatives in metropolitan regions demonstrates the applicability and effectiveness of the framework. A sensitivity analysis is conducted to examine how variations in criteria weights affect the rankings and to evaluate the robustness of the results. Furthermore, a comparative analysis highlights the practical and financial implications of each alternative by assessing their respective strengths and weaknesses.

Alzheimer’s Disease (AD) is a type of dementia that occurs in patients with increasing severity of deficits in cognitive abilities, memory, and nutritional status, and shifts in behavior. There is virtue in early diagnosis, and this should be compounded with accurate diagnosis. Clinical evaluation and imaging studies represent common diagnostic approaches and they are used and might be inaccurate and require a lot of time. More specific innovations in modern machine learning (ML) and biosensor technology suggest suitable approaches to build less subjective and more sensitive diagnostics. This study examines how biomarker-derived variables can be implemented into ML algorithms that are trained from datasets collected from the integrated biosensor platforms to detect AD. Concerning such biomarkers, we explore how well various ML models can classify AD patients from healthy controls. Further, the potential of optical fiber sensors to improve the sensitivity and specificity of biosensing platforms is also discussed in the study. The use of these technologies has been envisioned in complementing existing approaches in AD diagnosis to enhance its efficiency, thus early intervention, and better results for the patients. Alzheimer’s Detection using Machine Learning and Biosensor Integration.

The present work investigates the thermal and exergy performance of a novel bidirectional wavy absorber solar air heater (BWA-SAH) with sinusoidal corrugations in both the transverse and longitudinal directions to enhance heat transfer through increased surface area and turbulence. As compared to conventional studies on unidirectional wavy geometries, the present study extensively examines the impact of significant geometric parameters longitudinal and transversal wavelengths ((W-ll/d(h)), (W-lt/d(h)), wave amplitudes (A(wl)/d(h)), (A(wt)/d(h)), and Reynolds number on performance parameters such as Nu number, friction factor, thermohydraulic performance parameter (THPP), and exergy efficiency. A complete numerical investigation is conducted with the assistance of CFD simulations confirmed and accelerated by a trained deep neural network (DNN) model of 243 simulations. The DNN delivers R-2 > 0.98 for all performance measures and enables fast prediction with notable speedup compared to CFD simulations. Key findings reveal that smaller longitudinal wavelengths (W-ll/d(h) = 1.35) enhance heat transfer (Nu increases by similar to 35 %) but significantly enhance pressure drop (f increases by similar to 84 %). Similarly, higher longitudinal amplitudes (A(wl)/d(h) = 0.270) raise Nu by similar to 20 % but (f by similar to 100 %, highlighting a significant trade-off. A higher Reynolds number enhances heat transfer (Nu increases by 108 % as Re rises from 5000 to 15,000), but it reduces the THPP by approximately 11 % and exergy efficiency by about 53 % due to increased irreversibilities. The study confirms that intermediate longitudinal wavelengths (W-ll/d(h) = 2.25) and amplitudes (A(wl)/d(h) = 0.16) optimize THPP and exergy efficiency, and low Re (<10,000) minimizes pressure drop penalties. Coupling CFD simulations and DNN modeling provides a robust platform for rapid design optimization, significantly advancing the design of energy-efficient and sustainable solar air heaters. This research paves the way for more affordable and eco-friendly solar thermal systems, further driving global efforts in renewable energy deployment.

The decarbonization of fossil fuel-powered systems remains a critical challenge due to the energy penalties and economic burdens associated with conventional COQ capture technologies. This issue is particularly pronounced in tri- and multi-generation systems, where integrating carbon capture without compromising efficiency and profitability requires advanced system-level solutions. Addressing this challenge is essential to enable reliable low-carbon energy supply while meeting increasing demands for electricity, heating, cooling, and freshwater. This study proposes and systematically examines advanced carbon-neutral tri- and multi-generation energy systems featuring deep thermal integration and alternative heat supply strategies for COQ capture. Two integration scenarios are investigated: a methane-fueled tri-generation system incorporating a MEA-based COQ capture unit (System I), and an extended multi-generation configuration supported by geothermal energy to drive COQ capture and desalination (System II). Comprehensive thermo-enviro-economic assessments, gate-togate life cycle assessment, and data-driven multi-objective optimization using XGBoost surrogate modeling and advanced metaheuristic algorithms are employed. System I achieves 54.00% exergy efficiency and 98.78% COQ removal, reducing emissions to 21.47 kg/h, while System II further enhances performance, attaining 54.51% exergy efficiency, reducing specific GWP to 0.0179 kg/kWh, and increasing NPV to 14.07 M$. Optimized operation yields exergy efficiency up to 55.28% with minimal economic and environmental impact. The proposed systems provide a scalable and economically viable pathway for deploying carbon-neutral multi-generation plants in industrial and urban energy hubs, supporting long-term decarbonization and sustainable energy transition strategies.

IntroductionDiabetic retinopathy (DR) is an inflammatory condition affecting the retina caused by elevated and unregulated blood glucose levels. On a global scale, it is a contributing factor to vision impairment. Several deep learning (DL) methods use retinal images to identify DR severity. However, a significant improvement is required to assist medical professionals in recognizing DR in its early phases.MethodsThus, the author introduced a method based on the DL technique to determine the DR severity grades using retinal images. A ShuffleNet V2 model with vision transformers’ (ViT) attention mechanism was used to extract the features. An improved Whale optimization method (IWO) was used to fine-tune the feature extraction model. We employed a convolutional Kolmogorov-Arnold Network to categorize the DR severity using the extracted features. The EyePACS dataset was utilized to train the proposed DR severity grading model using a five-fold cross-validation strategy. We generalized the model on the Messidor-2 dataset.ResultsThe findings revealed an average accuracy of 93.84% on the MESSIDOR-2 dataset, demonstrating a substantial improvement in detecting DR using the fundus images.DiscussionFurthermore, the model demands minimal processing resources to generate the outcomes, leading to the deployment of the proposed DR severity detection model in healthcare facilities with limited computational resources.

Purpose The earlier prediction of autism spectrum disorder (ASD) placed a serious attention on ensuring the appropriate intervention to improve the child's behavioral, cognitive, and social development. The previous detection process is commonly time-intensive, subjective, and highly dependent on the clinical professions, which leads to limited accessibility in rural areas. The difficulties are addressed by introducing effective ASD detection systems, which provide a scalable, objective, and fast solution, reducing the challenges in the healthcare environment.Method This work integrates the Glowworm Optimization with Extreme Learning Machine Networks (GO-ELMN) model to enhance the efficiency of ASD prediction. During the analysis, ASD screening data for children are collected and processed frequently to obtain behavioral, demographic, and medical features. The extracted features are processed by an extraction learning technique, in which the network hyperparameters are optimized using the glowworm optimization algorithm. The optimized classifier recognizes children's behavior by addressing the issues of limited and imbalanced data.Findings The efficiency of the system is evaluated using experimental results, in which the system ensures high accuracy and convergence speed.Conclusion The ASD detection model provides an interpretable, fast, and reliable solution that is effectively utilized in the pediatric healthcare domain.

The rapid growth of Internet of Things (IoT) devices in smart grids and industrial control systems means that the global state has attained a level of technological evolution. Yet this growth has also created an enormous attack surface with millions of vulnerable endpoints, thus revealing inherent weaknesses of traditional security models. These conventional systems are fraught with data integrity challenges, single points of failure, and no proactive defense against new, adaptive cyber-physical threats. To overcome these limitations, this paper presents "Causio-TwinChain," a new security model that synergistically integrates three leading-edge technologies to establish a proactive, self-diagnostic, and tamper-resistant security framework for critical IoT infrastructure. A digital twin is a virtual replica that can monitor physical devices in real time via sandboxing. A permissioned blockchain provides an immutable, tamper-proof ledger for all device data and transactions, ensuring data integrity and auditability. Two kinds of machine-learning engines form the core intelligence: contrastive Learning, which detects subtle anomalies by modeling normal operations; and structural causal Learning, which diagnoses root causes of security incidents and predicts their potential impact. The model’s superior efficacy is demonstrated on an industrial IoT dataset. Causio-TwinChain yielded a 15.3% higher F1-score in novel attack detection, and reduced the mean time for incident diagnosis by 68% compared to benchmark intrusion detection systems. This model reduced the false-positive rate by 22%, demonstrating its robustness in noisy environments. Moving beyond mere attack detection to explainable diagnosis and predictive mitigation, this work establishes a new benchmark for building proactive, resilient, and self-healing security frameworks that safeguard the most critical IoT applications and enhance trust and continuity in operational services.

IntroductionVision impairment caused by retinal diseases stands out as one of the significant sources of irreversible blindness, highlighting the importance of developing robust automated screening tools to detect eye diseases at an early stage. Despite the promising results of recent deep learning approaches to the automatic classification of retinal images, most existing solutions focus on image-based representations and do not consider the importance of vascular topology. To overcome the limitations, this study proposes VG-RETFound – a hybrid framework based on a retinal foundation model and graph-based vascular-topology modeling that can be used to classify multiple retinal categories and detect vision impairment.MethodsThe authors use RETFound—a retinal foundation model—pretrained on a vast number of retinal images, as well as the Graph Attention Network (GAT), to learn vascular topology features by representing retinal vessels as graphs. The presented approach employs a cross-modal multi-head attention fusion method to integrate features from both modalities in real time. To evaluate the efficiency of VG-RETFound, experiments were conducted using the RFMiD and ODIR-5 K datasets for model training and testing across six retinal categories: Normal, Diabetic Retinopathy, Pathological Myopia, Age-related Macular Degeneration, Hypertensive Retinopathy, and Other Retinal Diseases.ResultsThe accuracy of 96.39%, F1-score of 95.84%, MCC of 94.87%, and Kappa of 94.71% were attained on an internal test dataset, while cross-dataset external validation on the BRSET dataset revealed accuracy of 94.81%, F1-score of 94.25%, MCC of 93.09%, and Kappa of 92.94%.DiscussionThese findings demonstrate that integrating retinal foundation models with anatomically informed vessel-topology learning significantly improves classification accuracy, interpretability, and generalization, highlighting the potential of VG-RETFound as a tool for large-scale vision impairment. screening and retinal disease diagnosis.

The rapid growth of cities demands sophisticated decision support systems capable of simultaneously tackling social, economic, and environmental issues in urban planning. This research introduces a comprehensive decision-making framework that integrates machine learning to optimize urban infrastructure development. To improve the reliability of the analysis, recursive feature elimination paired with random forest (RF-RFE) is utilized to pinpoint the most critical factors, ensuring that the decision process emphasizes the features that matter most. To handle uncertainty and vagueness in expert evaluations, picture fuzzy sets are implemented, allowing a nuanced expression of hesitation, while the entropy method is employed to objectively determine the weights of social, economic, and environmental criteria, thereby reducing subjective influence. The evaluation based on the approach of relative utility and nonlinear standardization (ERUNS) is employed to rank urban planning alternatives, capturing the nonlinear interactions among criteria. This framework provides a comprehensive, data-driven decision support tool for urban planners and policymakers to address the intricate and dynamic challenges of modern urbanization.

IntroductionDevelopmental dyslexia refers to a common neurodevelopmental disorder, which impairs the accuracy and fluency of reading, and early identification is vital for initiating timely intervention. Nonetheless, the traditional methods of formal assessment are time- and resource-intensive, which limits their scalability. Machine-learning approaches and eye-tracking technologies provide objective, data-driven solutions for dyslexia screening. This research integrates current evidence on eye-movement-based and emerging multimodal computational methods for dyslexia screening, risk identification, and algorithmic classification during reading tasks.MethodsPubMed, Scopus, Web of Science, and CINAHL were searched systematically to identify studies published between January 2015 and March 2026. Eligible studies included analysis of eye-movement obtained via eye tracking or electrooculography (EOG), with or without predictive modeling. Methodological quality was assessed using JBI, PROBAST, ROBINS-I, and COSMIN tools.ResultsTwenty-three articles were included out of 50 full-text articles screened comprising eye-movement biomarker/observational studies (n = 5), machine-learning prediction-model studies (n = 14), intervention response studies (n = 2), and reliability/feasibility studies (n = 2). The sample sizes ranged from small experimental cohorts (<20 participants) to larger datasets (>300 participants). In the literature, dyslexic readers were consistently found to exhibit longer fixation durations, increased regression behavior and reduced saccadic efficiency. Machine-learning algorithms using fixation, saccade, scan path, and signal-based features demonstrated classification accuracies ranging from approximately 80 to 95% with some studies reporting values approaching 99% under specific experimental conditions.DiscussionNevertheless, there was a significant heterogeneity in datasets, feature extraction methods, outcome definitions and validation schemes. Notably, numerous studies used proxy diagnostic labels, small or internally derived datasets, and internal cross-validation, which introduces the risk of overfitting and performance inflation. Explicit multimodal or multi-source modeling was identified in three of 23 studies involving combinations of gaze data with demographic, cognitive, linguistic, VR-bed, text-derived, saliency-map, or CNN-based features. Two additional studies used EOG as an alternative eye-movement signal modality rather than true multi-source fusion. Therefore, the evidence base remains dominated by eye-movement and gaze-derived approaches, while multimodal evidence should be interpreted as emerging and exploratory. Altogether, eye-movement based computational systems are a promising, non-invasive method for scalable dyslexia screening.Systematic review registrationPROSPERO, identifier (RD42061332527).

Objectives Brain tumors have been a major factor in the development of mental disorders like anxiety and depression. The primary target of this research is to develop a smart, economically feasible system that can detect and classify brain tumors by analyzing MRI images, reducing manual labor, and shortening diagnostic time. Methods Although diagnosis with medical imaging is efficient for various health issues, accurately classifying brain tumors remains challenging for medical experts. This research introduces a new energy channel-based hybrid optimized network (ECHO-Net) as an automatic identification tool for brain tumors, aiming to facilitate medical diagnosis remotely. Energy shape prior segmentation (ESPS) is part of the framework, providing accurate segmentation and cropping of tumor regions from MRIs. Channel and spatial attention-based neural network (CSA-Net) differentiates between normal and tumor-affected images. Additionally, the hybrid chimp-based whale optimization (HCWO) algorithm enhances prediction precision and optimally sets sigmoid activation function parameters for better convergence and generalization. Results Performance evaluation of ECHO-Net using publicly available MRI brain datasets, including Figshare, BRATS 2018–2020, and a clinical MRI dataset, shows a peak signal-to-noise ratio of 41.87 dB, a Structural Similarity Index Measure of 0.992, a sensitivity of 99.4%, an accuracy of 99.2%, and an average computational time of 2.68 s, outperforming current state-of-the-art methods. Conclusion The proposed ECHO-Net is an automated, accurate, fast, and robust system for brain tumor detection. With low computational costs, it effectively segments abnormal regions and recognizes tumor types, demonstrating strong potential as a tool in intelligent healthcare systems and real-world clinical applications.

Growing global energy demand, and relying heavily on fossil fuels to meet these energy needs led to serious environmental and economic concerns. Within the gate-to-gate operational boundary, this work introduces a revolutionary polygeneration carbon neutral process that can concurrently produce valuable products (electricity/heating/cooling/syngas) with negligible CO2 emissions. This configuration combines a biomass combustor, with biogas bi-reforming unit, Rankine cycle with orthoxylene organic working-fluid, and absorption chiller, which are simulated using Aspen-HYSYS software. The study applies the evaluation process integrated with life cycle assessment and profitability analysis, while the optimal operating is determined using comparative machine learning-assisted multi-objective optimization approach. With a net electricity of 356 kW, heating and cooling capacities of 282 kW and 325 kW, and a syngas production of 0.7754 kg/s, the system shows exceptional efficiencies of 96.92% and 71.63% for energy and exergy. Furthermore, the system maintains a nearly zero CO2 footprint (0.000858 kg/kWh) along with negative global warming potential (-0.0038 kgCO2-eq/ kWh), which highlight its novelty contribution toward low-carbon system. Moreover, the system demonstrated indicates that the energy production and product generation processes operate at a competitive cost level, with the cost of energy of 0.19 $/kWh and the total unit cost of product (TUCP) of 4.21 $/GJ, which results net present value of 8.26 million $, and a payback-period of 6.21 years. The Firefly optimization algorithm find the optimal configuration with efficiency of 73.60% for exergy and TUCP of 4.03 $/GJ. The Monte-Carlo sensitivity evaluation indicates that the mass flow of biogas as the most influential variables affecting system performance.

Early and precise recognition of developmental disabilities (DDs) is essential for the timely intervention of preschool-age children. Although various predictive models have been developed in the past, they cannot provide accurate detections and face challenges like high false prediction, limited generalization, etc. To address these issues, a novel artificial intelligence (AI)-based algorithm was proposed for accurate DD detection. The proposed system integrates digital biomarkers from speech and behavioral domains for predicting DDs. Consequently, an Enhanced Gaussian-Based Noise Filtering (EGNF) is applied to speech data, and Z-score normalization is applied to behavioral data for preprocessing, which makes the raw data reliable for subsequent analysis. Further, feature extraction is performed using Convolutional-Bidirectional Long Short-Term Memory (Conv-BiLSTM) for speech data and Graph Attention Network (GAT) for behavioral data. Subsequently, a Hybrid Feature Selection (HFS) method was developed by combining the Harmony Search Algorithm (HSA) and ReliefF ranking for selecting informative and relevant attributes for model training. Finally, a hybrid classification model named DeepSTNet has been developed by integrating MnasNet and Independently Recurrent Neural Networks (IndRNN), which allows the capture of spatial and temporal long-term dependencies within the data, resulting in high detection performances. The proposed framework was implemented in Python and validated using comprehensive data formed by combining the Dysarthria Detection dataset and the Autism Dataset for Toddlers. The execution outcomes highlighted that the developed framework achieved an accuracy of 99.46%, a precision of 97.74%, and a recall of 98.05%, illustrating that this AI-centric system can be equipped by end users, including healthcare professionals and caregivers, with tools for early screening and detection of DDs.

In recent days, fifth-generation infrastructure based on the Internet of Things has been instrumental in terahertz communication, leading to molecule absorption, blocking, and deep fades. These issues lead to intermittent behaviour in wireless networks, potentially causing delays. The growing user density and application demands are met by sharing machine type and device-to-device information. However, the limit for user support is indefinite due to varying allocations of applications and service channels. Irrespective of the user and application requirements, the scalability management is powered by a novel Dynamicity-assisted Service Assignment scheme. The proposed scheme balances service assignment and user management with scalable computing by following three steps. First, the reliability of this scheme is enhanced through federated learning, which maintains balanced density and service assignment. Second, the proposed scheme categorizes latency-aware and service-aware requests from the user equipment for granting independent and collaborative service access. The requests are classified into distinguished time and service intervals, regardless of density. Third, this alters the service distribution and request satisfaction from the previous to the current processing level, supporting heterogeneous users. Therefore, this scheme’s service access and response ratio are considerably high. The other metrics, such as latency, backlogs, and computational complexity, are verified to validate the performance of the proposed scheme.