Soft computing and eddy currents to estimate and classify delaminations in biomedical device CFRP plates

Journal of Electrical Engineering, Journal Year: 2025, Volume and Issue: 76(1), P. 72 - 79

Published: Feb. 1, 2025

Abstract This paper presents an approach based on eddy currents induced by suitable magnetic induction fields to test, estimate, and classify subsurface delaminations in Carbon Fibre Reinforced Polymer (CFRP) plates for biomedical devices. The two-dimensional maps obtained, characterised high fuzziness, required the software development of a procedure highly efficient fuzzy classifier that exploits similarity computations with reduced computational load collecting similar (deriving from equally defects) specific defects. hardware implementation what is designed (plate-probe system) detects evaluates entity defects due classification percentage comparable performances obtained more sophisticated classifiers, providing possible tool evaluating potentially useful assess aircraft compliance applicable safety standards.

Language: Английский

---

Smart Electronic Device-Based Monitoring of SAR and Temperature Variations in Indoor Human Tissue Interaction

Applied Sciences, Journal Year: 2025, Volume and Issue: 15(5), P. 2439 - 2439

Published: Feb. 25, 2025

The daily use of devices generating electric and magnetic fields has led to potential human overexposure in home work environments. This paper assesses the possible effects on health at low high frequencies. It presents an electronic monitoring device that captures incidence specific absorption rate (SAR) temperature variation (∆T) body. system transmits data a cloud platform, where feedforward neural network (FFNN) processes received information. SAR surface values are detected indoor environment, stationary moving subjects. results effectively assess distribution due electromagnetic fields. prototype peaks when subjects remained motionless. Predictive analysis confirms need for workplaces with materials shielding external signals attenuating internal sources. Moderate mobile phone could lower values.

Language: Английский

---

Comparison of LSTM- and GRU-Type RNN Networks for Attention and Meditation Prediction on Raw EEG Data from Low-Cost Headsets

Electronics, Journal Year: 2025, Volume and Issue: 14(4), P. 707 - 707

Published: Feb. 12, 2025

This study bridges neuroscience and artificial intelligence by developing advanced models to predict cognitive states—specifically attention meditation—using raw EEG data collected from low-cost commercial devices such as NeuroSky Brainlink. Leveraging the temporal capabilities of recurrent neural networks (RNNs), particularly long short-term memory (LSTM) gated units (GRUs), evaluates their effectiveness in predicting future states. These predictions have applications real-time brain–computer interface (BCI) systems, enhancing responsiveness adaptability dynamic environments like robotic control. The proposed LSTM model demonstrated superior predictive accuracy for meditation states, achieving a Root Mean Squared Error (RMSE) 10.90, while GRU excelled with an RMSE 11.79. Both outperformed results provided proprietary eSense algorithm, reinforcing potential cognitive-state analysis. Notably, inference times were optimized under 50 milliseconds, making suitable applications. findings underline feasibility using signals affordable robust prediction, offering significant step forward applied neuroscience. research lays groundwork further exploration RNN architectures BCI applications, enabling safer, more intuitive, personalized interactions assistive technologies beyond.

Language: Английский

---

Data Reconstruction Methods in Multi-Feature Fusion CNN Model for Enhanced Human Activity Recognition

Sensors, Journal Year: 2025, Volume and Issue: 25(4), P. 1184 - 1184

Published: Feb. 14, 2025

Human activity recognition (HAR) plays a pivotal role in digital healthcare, enabling applications such as exercise monitoring and elderly care. However, traditional HAR methods relying on accelerometer data often require complex preprocessing steps, including noise reduction manual feature extraction. Deep learning-based human using one-dimensional suffers from limited Transforming time-series signals into two-dimensional representations has shown potential for enhancing extraction reducing noise. existing single-feature inputs or extensive face limitations robustness accuracy. This study proposes multi-input, CNN architecture three distinct reconstruction methods. By fusing features reconstructed images, the model enhances capabilities. method was validated custom dataset without requiring steps. The proposed outperformed models single-reconstruction raw data. Compared to baseline, it achieved 16.64%, 13.53%, 16.3% improvements accuracy, precision, recall, respectively. We tested across various levels of noise, consistently demonstrated greater than time-series-based approach. Fusing effectively captured latent patterns variations demonstrates that can be improved multi-input approach with offers practical efficient solution, streamlining performance, making suitable real-world applications.

Language: Английский

---

Efficient and Effective Detection of Repeated Pattern from Fronto-Parallel Images with Unknown Visual Contents

Signals, Journal Year: 2025, Volume and Issue: 6(1), P. 4 - 4

Published: Jan. 24, 2025

The effective detection of repeated patterns from inputs unknown fronto-parallel images is an important computer vision task that supports many real-world applications, such as image retrieval, synthesis, and texture analysis. A pattern defined the smallest unit capable tiling entire image, representing its primary structural visual information. In this paper, a hybrid method proposed, overcoming drawbacks both traditional existing deep learning-based approaches. new leverages features pre-trained Convolutional Neural Network (CNN) to estimate initial sizes refines them using dynamic autocorrelation algorithm. Comprehensive experiments are conducted on dataset textile well another set non-textile demonstrate superiority proposed method. accuracy 67.3%, which represents 20% higher than baseline method, time cost only 11% baseline. has been applied contributed design, it can be adapted other applications.

Language: Английский

---

Volumetric Assessment and Graph Theoretical Analysis of Thalamic Nuclei in Essential Tremor

Brain and Behavior, Journal Year: 2025, Volume and Issue: 15(2)

Published: Feb. 1, 2025

ABSTRACT Introduction Essential tremor (ET) is a neurological disorder primarily characterized by upper limb action tremor. It widely recognized that the thalamus implicated in ET pathophysiology, playing central role treatment approaches. This study aimed to explore thalamic morphology, assessing macrostructural changes and intrinsic networks patients. Methods A total of 109 (41 with 68 without resting tremor) 81 healthy controls (HC) were enrolled study. An automatic probabilistic segmentation nuclei was employed on T1‐weighted MRI images using FreeSurfer 7.4. Subsequently, volumetric data extracted, graph theoretical analysis applied cortical–thalamic network, global local network properties. Results No significant differences observed volume between patients HC. exhibited alterations suggesting less efficient brain comparison also showed such as lower eccentricity path length ventral reduced efficiency pulvinar, indicating interconnected network. rest Conclusion Our demonstrates patients, impaired communication interconnection regions. These findings confirm involvement lateral pulvinar key regions pathophysiology supporting targeting these for therapeutic

Language: Английский

---

Development of a Prototype for the Acquisition of Biopotentials Implementing a New Interconnection Method for Shielding

Applied Sciences, Journal Year: 2025, Volume and Issue: 15(5), P. 2442 - 2442

Published: Feb. 25, 2025

A biological system can emit signals, and if these signals are correctly acquired, they provide valuable information about the processes occurring within system, enhancing our knowledge of system. For this reason, we present a prototype for acquiring various biopotentials using main module that integrates amplification, high-pass filtering, band-reject offset adjustment stages. This configuration allows adjustable gain when working with different includes dedicated filtering modules each biopotential type. We also propose new topology shielded controller used in interconnection between electrodes amplification stage to reduce noise introduced by electrical network. Biopotentials acquired proposed show improved reduction signal definition compared those other topologies found literature. The design utilizes basic electronics, making it low-cost solution. Ultimately, is simple, efficient, suitable applications requiring acquisition multiple types biopotentials.

Language: Английский

---

Electromyography Signals in Embedded Systems: A Review of Processing and Classification Techniques

Biomimetics, Journal Year: 2025, Volume and Issue: 10(3), P. 166 - 166

Published: March 10, 2025

This article provides an overview of the implementation electromyography (EMG) signal classification algorithms in various embedded system architectures. They address specifications used for different devices, such as number movements and type method. Architectures analyzed include microcontrollers, DSP, FPGA, SoC, neuromorphic computers/chips terms precision, processing time, energy consumption, cost. analysis highlights capabilities each technology real-time wearable applications smart prosthetics gesture control well importance local inference artificial intelligence models to minimize execution times resource consumption. The results show that choice device depends on required specifications, robustness model, be classified, limits knowledge concerning design budget. work a reference selecting technologies developing biomedical solutions based EMG.

Language: Английский

---

Development and Validation of a Low-Cost External Signal Acquisition Device for Smart Rail Pads: A Comparative Performance Study

Sensors, Journal Year: 2025, Volume and Issue: 25(6), P. 1933 - 1933

Published: March 20, 2025

The development of cost-effective and reliable railway monitoring technologies is crucial for the maintenance modern infrastructure. Embedding sensors into rail pads has emerged as a promising approach wheel–track interactions, but successful implementation these systems requires robust framework signal data acquisition analysis. This study validates custom-designed External Signal Acquisition Device (ESAD) use with smart pads, comparing its performance against high-precision commercial analog module. While module delivers exceptional accuracy, high cost, bulky size, complex installation requirements limit practicality large-scale applications. Laboratory-scale full-scale experiments simulating real-world conditions demonstrated that custom ESAD performs comparably to During simulated train passages, showed reduced dispersion load speed increased, confirming ability provide calibration data. Moreover, device maintained over 95% reliability in analyzing load-to-signal linearity, ensuring consistent dependable both laboratory field settings. However, does have limitations, including slightly lower resolution low frequencies potential sensitivity extreme environmental conditions, which may affect specific scenarios. These findings highlight ESAD’s strike balance between cost functionality, making it viable solution widespread research contributes advancement affordable efficient technologies, fostering adoption preventive practices enhancing overall infrastructure performance.

Language: Английский

---

Convolutional Neural Network for Interface Defect Detection in Adhesively Bonded Dissimilar Structures

Applied Sciences, Journal Year: 2024, Volume and Issue: 14(22), P. 10351 - 10351

Published: Nov. 11, 2024

This study presents an ultrasonic non-destructive method with convolutional neural networks (CNN) used for the detection of interface defects in adhesively bonded dissimilar structures. Adhesive bonding, as weakest part such structures, is prone to defects, making their challenging due various factors, including surface curvature, which causes amplitude variations. Conventional methods and processing algorithms may be insufficient enhance detectability, some influential factors cannot fully eliminated. Even after aligning signals reflected from sample interface, cases, non-parallel interfaces, persistent variations remain, significantly affecting defect detectability. To address this problem, a proposed that integrates NDT CNN, able recognize complex patterns non-linear relationships, developed work. Traditional pulse-echo testing was performed on adhesive structures collect experimental data generate C-scan images, covering time gate first reflection point where reflections were attenuated. Two classes datasets, representing defective defect-free areas, fed into network. One subset dataset model training, while another validation. Additionally, collected different during independent experiment evaluate generalization performance The results demonstrated integration CNN enabled high prediction accuracy automation analysis process, enhancing efficiency reliability detecting defects.

Language: Английский

---