Vol. 60 No. 4 (2026): In Progress

Open Access

Article

Article ID: 1706

Prediction of heart sound using the xLSTM method for hypertrophic cardiomyopathy

by Shatiswaran Vigian, R Kanesaraj Ramasamy, Junaidi Abdullah

Sound & Vibration, Vol.60, No.4, 2026;

Heart disease has emerged as a major global public health concern, driven by poor dietary habits, unhealthy lifestyle choices, and limited health awareness. Accurate diagnosis of cardiac conditions remains challenging for hospitals and clinical institutions. Hypertrophic cardiomyopathy is an autosomal dominant disorder caused by mutations in sarcomere protein genes that affect the contractile function of cardiac muscle. With the growing adoption of digital health systems, large volumes of patient data are now collected and stored, providing opportunities for computational approaches to support clinical decision making. Machine learning methods have become increasingly important for analysing complex and nonlinear patterns in medical data. This study presents an ensemble-based approach for heart sound classification and introduces an Extended Long Short-Term Memory (xLSTM) model for the detection of cardiac abnormalities. The method was evaluated using acoustic features extracted from phonocardiogram recordings. The proposed model achieved 96.93% accuracy, 93.50% sensitivity, and 99.63% specificity, indicating strong performance in distinguishing normal and abnormal heart sounds. In comparison with previously reported techniques, the ensemble strategy and the xLSTM architecture provided improved accuracy. Model performance was assessed using accuracy, precision, recall, and F1 score, confirming the effectiveness of the proposed approach for automated heart sound analysis.

Open Access

Article

Article ID: 4028

A framework of structural health monitoring integrated with deep learning schemes to estimate the displacement of 3D LiDAR scanning on bridges

by Wael A. Altabey

Sound & Vibration, Vol.60, No.4, 2026;

In recent decades, the rapid development of transportation infrastructure safety, such as highways, bridges, and tunnels has greatly promoted the development of the regional economy. The structures with safety hazards and emergencies need continuous monitoring over time. The integrated artificial intelligence algorithms with sensor responses can provide real-time information for further analysis and decision-making for the transportation system, improving the circulation efficiency of the transportation network, ensuring the stability of road structures, and avoiding irreparable damage. This paper aims to develop an efficient and low-cost method to help detect early-stage transportation infrastructure damage through permanent or periodic monitoring. In this research, we used LiDAR scanning units (terrestrial LiDAR fixed on holders and movable units fixed on UAVs) integrated with a novel deep neural network (DNN) for structural monitoring of bridges based on the 3D mapping of bridge displacement compiled from LiDAR scanning over time. The monitoring model is based on a recurrent neural network with long short-term memory blocks (RNN-LSTM) since the LiDAR scanning datasets have a time-dependent and memory-dependent behavior. The response of the proposed DNN achieved a high accuracy rate, regression rate, and F-score equal to 96.43%, 93.77%, and 91.65%, respectively. A deep analysis of the confusion matrix and a side-by-side look at predicted and actual conditions highlight how well the model can tell apart different traditional methods to estimate the bridge displacement in literature. So, the data from LiDAR and DNN models can be combined to analyze the monitoring of transportation infrastructure.

Open Access

Article

Article ID: 4013

Complaint-guided robust optimization of highway noise barriers with hourly traffic variability

by Yogeesh Nijalingappa, Markala Karthik, Asokan Vasudevan, Mohammed Almakki, Zetty Pakir Mastan, Mayibongwe Tafara Mudzengi

Sound & Vibration, Vol.60, No.4, 2026;

Road-traffic noise is a daily problem in many fast-growing tier-2 cities. On busy corridors, mixed traffic and stop-go movement raise both exposure and public annoyance. This study presents an uncertainty-aware framework for optimizing roadside noise barriers by combining hourly traffic variability with community complaint signals. The NH-48 urban approach corridor in Tumakuru, Karnataka, was examined using a 7-day dataset at hourly resolution. A calibrated baseline model related hourly A-weighted equivalent sound levels to log-scaled traffic flow, mean speed, and heavy-vehicle fraction, with good agreement with measurements (overall MAE 2.3 dB(A), RMSE 3.1 dB(A)). Input uncertainty was represented through nested α-cut interval bands, and the measured hourly levels were increasingly captured as the bands widened (coverage from 0.66 at α = 0.8 to 0.92 at α = 0.2). Barrier design was posed as a multi-objective robust optimization problem that minimized conservative exceedance, complaint-weighted nuisance, and a normalized cost index. The evolutionary search produced a Pareto set with a clear cost-performance trade-off. The preferred solutions lowered robust exceedance and complaint-weighted objective values by up to 35% and 42%, respectively, relative to baseline candidates. Receptor-level exceedance hours fell by about 39–45%, and mean upper-bound levels dropped by as much as 3.9 dB(A) at near-road receptors. Overall, the results show that complaint signals can help identify perceptual hotspots, while the uncertainty-aware model maintains robust exposure reduction under day-to-day traffic variation.

Open Access

Research Project

Article ID: 3976

Co-design of flow fields and vibration control for vanadium redox batteries

by Jacer  Hamrouni, Leila  Abdelgader, Abdennaceur  Kachouri, Mounir  Baccar

Sound & Vibration, Vol.60, No.4, 2026;

Vanadium redox flow batteries (VRFBs) are promising candidates for grid-scale energy storage, yet their performance and operational reliability remain constrained by conventional flow field designs, such as serpentine and interdigitated architectures, which inherently trade-off between uniform reactant distribution, hydraulic efficiency, and mechanical stability under dynamic fluid loads. While previous optimization efforts have focused separately on electrochemical performance or pressure drop reduction, no integrated framework has addressed the coupled interaction between flow field topology, species transport, and flow-induced vibration, leaving a critical gap in achieving simultaneously high efficiency and long-term structural reliability. This study introduces a bio-inspired, co-design framework that integrates topology optimization, computational fluid dynamics, electrochemical reaction modeling, and structural dynamics analysis to concurrently optimize flow field architecture and mitigate pressure-induced vibration in VRFBs. The methodology employs a density-based optimization approach guided by Murray's Law and leaf venation principles, constrained by pressure drop limits and manufacturability, and validated through both high-fidelity numerical simulations and experimental prototype testing. The optimized biomimetic flow field achieves a 28% increase in volume-averaged reaction rate, a 27.6% reduction in pressure drop at 40 mL min⁻¹, and a 41% reduction in root-mean-square vibration acceleration compared to a conventional interdigitated design. Voltage efficiency improves by 5.2 percentage points, reaching 89.5% at 120 mA cm⁻², while active area utilization increases from 68% to 91%. These results demonstrate that the proposed co-design framework successfully decoupled the traditional trade-off between electrochemical performance and hydraulic-mechanical stability, providing a validated, nature-inspired pathway toward high-performance, reliable energy storage systems that address practical engineering challenges in noise, vibration, and durability.