Mean Execution Time Research Articles

Analysis of methods and algorithms for image filtering and quality enhancement

This paper addresses the prevalent issue of image noise and presents methods for its mitigation. The paper describes, analyses and tests a variety of image filtering techniques, with specific reference to their use in different contexts. The filtering methods can be classified into two principal categories: linear filters, which include the Gaussian and mean filters, and non-linear filters, which comprise the median filter, the Fast Fourier Transform (FFT), the Non-Local Means (NLM) filter, and the anisotropic diffusion filter. The efficacy of each filter is mathematically described and evaluated on RGB images using the Python programming language. The study delineates the evaluation metrics and their respective advantages and disadvantages. The Root Mean Square Error (RMSE) and Peak Signal-to-Noise Ratio (PSNR) are employed as criteria for the analysis of algorithm efficiency. Furthermore, the mean execution time for each algorithm is also monitored. The experimental data suggests that linear filters are relatively fast but produce inferior results and are best employed as preparatory measures. Non-linear filters have been demonstrated to be more robust and applicable to a variety of noise types, although it has been established that they require parameter fine-tuning. The study demonstrates that anisotropic diffusion is suitable for both manual image processing and real-time applications, offering an optimal balance between processing speed and denoised image quality. NLM is optimal for high-quality single image processing due to its superior results, despite a slower processing speed. FFT is noted for its efficiency in eliminating periodic noise. Further research will be conducted on advancing filtering techniques for different real-world scenarios and autonomous systems.

Safety and efficacy of surgically performed continuous superficial serratus anterior plane block in uniportal video-assisted thoracic surgery.

The 'surgically performed' continuous superficial serratus anterior plane block (continuous s-SAPB) was never described before in uniportal video-assisted thoracic surgery (uniportal VATS) surgery. The aim of the study was to evaluate the safety and efficacy of the technique. Between March 2022 and April 2023, 50 patients, undergone uniportal VATS surgery at our thoracic surgery department, were scheduled for a surgically performed continuous s-SAPB as post-operative analgesia protocol. The mean execution time for the block was 3.92 ± 2.56 min. Ten patients (20%) required morphine for a visual analogue scale (VAS) score >4 immediately after surgery. The recorded VAS score at chest tube removal was 1.87 ± 1.41, whereas 2 h after the manoeuvre was 0.42 ± 0.72. No complication related to block insertion was recorded. The onset of chronic pain was observed in a total of 2 patients (4%). The surgically performed continuous s-SAPB in uniportal VATS seems to be safe and easy to perform, and it provides a satisfactory analgesic effect.

Individual Updating Strategies-based Elephant Herding Optimization Algorithm for Effective Load Balancing in Cloud Environments

In this manuscript, an Individual Updating Strategies-based Elephant Herding Optimization Algorithm are proposed to facilitate the effective load balancing (LB) process in cloud computing. Primary goal of proposed Individual Updating Strategies-based Elephant Herding Optimization Algorithm focus on issuing the workloads pertaining to network links by the purpose of preventing over-utilization and under-utilization of the resources. Here, NIUS-EHOA-LB-CE is proposed to exploit the merits of traditional Elephant Herd Optimization algorithm to achieve superior results in all dimensions of cloud computing. In this NIUS-EHOA-LB-CE achieves the allocation of Virtual Machines for the incoming tasks of cloud, when the number of currently processing tasks of a specific VM is less than the cumulative number of tasks. Also, it attains potential load balancing process differences with the help of each individual virtual machine’s processing time and the mean processing time (MPT) incurred by complete virtual machine. Efficacy of the proposed technique activates the Cloudsim platform. Experimental results of the proposed method shows lower Mean Response time 11.6%, 18.4%, 20.34%and 28.1%, lower Mean Execution Time 78.2%, 65.4%, 40.32% and 52.6% compared with existing methods, like Improved Artificial Bee Colony utilizing Monarchy Butterfly Optimization approach for Load Balancing in Cloud Environments (IABC-MBOA-LB-CE), An improved Hybrid Fuzzy-Ant Colony Algorithm Applied to Load Balancing in Cloud Computing Environment (FACOA-LB-CE), Hybrid firefly and Improved Multi-Objective Particle Swarm Optimization for energy efficient LB in Cloud environments (FF-IMOPSO-LB-CE) and A hybrid gray wolf optimization and Particle Swarm Optimization algorithm for load balancing in cloud computing environment (GWO-PSO-LB-CE).

Impact of Columnar Storage Optimizations in Redshift and BigQuery

With the data explosion, companies are utilizing "new age" data warehouses, such as Amazon Redshift and Google BigQuery, to process this information. Space-efficient storage structures (specifically column-oriented storage formats) are the path to improved query performance and lower costs. Columnar storage stores the data sorted by columns and not rows so that only the columns being queried are accessed when a query is made. Fewer disk reads mean faster execution time, and that's good for analytical processing

Enhancing Embedded Object Tracking: A Hardware Acceleration Approach for Real-Time Predictability.

While Siamese object tracking has witnessed significant advancements, its hard real-time behaviour on embedded devices remains inadequately addressed. In many application cases, an embedded implementation should not only have a minimal execution latency, but this latency should ideally also have zero variance, i.e., be predictable. This study aims to address this issue by meticulously analysing real-time predictability across different components of a deep-learning-based video object tracking system. Our detailed experiments not only indicate the superiority of Field-Programmable Gate Array (FPGA) implementations in terms of hard real-time behaviour but also unveil important time predictability bottlenecks. We introduce dedicated hardware accelerators for key processes, focusing on depth-wise cross-correlation and padding operations, utilizing high-level synthesis (HLS). Implemented on a KV260 board, our enhanced tracker exhibits not only a speed up, with a factor of 6.6, in mean execution time but also significant improvements in hard real-time predictability by yielding 11 times less latency variation as compared to our baseline. A subsequent analysis of power consumption reveals our approach's contribution to enhanced power efficiency. These advancements underscore the crucial role of hardware acceleration in realizing time-predictable object tracking on embedded systems, setting new standards for future hardware-software co-design endeavours in this domain.

Optimal Scheme Models with Imperfect Checkpoint

We propose optimal checkpointing models for dual and majority decision structures with imperfect checkpoints. In systems with high reliability, the process returns to the previous checkpoint when errors occur. However, in real-time systems, if the process always returns to the previous checkpoint, its process might not end in time. In this paper, we propose extended checkpoint models in which when errors occur, the process makes forward and backward recoveries, i.e., the process makes the same one again or returns to the first one with some probability. Using dual and majority decision structures as processing modules, these models are applied to constant and random tasks for the processing of objective works. Furthermore, it is shown that processing models are given by the more general forms using a K-out-of-n system. We formulate simultaneous renewal equations for imperfect models and solve them skillfully, using mathematical methods. The mean execution times until the process succeeds are obtained, and optimal policies to minimize them are derived analytically. These results include the former ones and would be useful for some practical checkpoint models.

Using process features to investigate scientific problem-solving in large-scale assessments.

This study investigates the process data from scientific inquiry tasks of fair tests [requiring test-takers to manipulate a target variable while keeping other(s) constant] and exhaustive tests (requiring test-takers to construct all combinations of given variables) in the National Assessment of Educational Progress program. We identify significant associations between item scores and temporal features of preparation time, execution time, and mean execution time. Reflecting, respectively, durations of action planning and execution, and execution efficiency, these process features quantitatively differentiate the high- and low-performing students: in the fair tests, high-performing students tended to exhibit shorter execution time than low-performing ones, but in the exhaustive tests, they showed longer execution time; and in both types of tests, high-performing students had shorter mean execution time than low-performing ones. This study enriches process features reflecting scientific problem-solving process and competence and sheds important light on how to improve performance in large-scale, online delivered scientific inquiry tasks.

Fault Identification and Classification of Asynchronous Motor Drive Using Optimization Approach with Improved Reliability

This article aims to provide a technique for identifying and categorizing interturn insulation problems in variable-speed motor drives by combining Salp Swarm Optimization (SSO) with Recurrent Neural Network (RNN). The goal of the proposed technique is to detect and classify Asynchronous Motor faults at their early stages, under both normal and abnormal operating conditions. The proposed technique uses a recurrent neural network in two phases to identify and label interturn insulation concerns, with the first phase being utilised to establish whether or not the motors are healthy. In the second step, it discovers and categorises potentially dangerous interturn errors. The SSO approach is used in the second phase of the recurrent neural network learning procedure, with the goal function of minimizing error in mind. The proposed CSSRN technique simplifies the system for detecting and categorizing the interturn insulation issue, resulting in increased system precision. In addition, the proposed model is implemented in the MATLAB/Simulink, where metrics such as accuracy, precision, recall, and specificity may be analysed. Similarly, existing methods such as Adaptive Neuro-Fuzzy Inference System (ANFIS), Recurrent Neural Network (RNN), and Salp Swarm Algorithm Artificial Neural Network (SSAANN) are utilised to evaluate metrics such as Root mean squared error (RMSE), Mean bias error (MBE), Mean absolute percentage error (MAPE), consumption, and execution time for comparative analysis.

Luminosity and Contrast Adjustment of Fundus Images with Reflectance

This paper presents an automatic correction method for luminosity and contrast variation in fundus images. Sixty retina or fundus images with different levels of reflectance are selected from online databases and used to assess the effectiveness of the proposed method. There are five stages in the approach, and they are image input, filtering, luminosity correction, histogram stretching and post-processing. First, a color fundus image is read as input, and its three color components, red (R), green (G) and blue (B), are separated into different channels or arrays. Next, the eye region, or the region of interest (ROI), is identified along with its border via thresholding. After that, the original ratios of red-to-green and blue-to-green for every pixel in the ROI are computed and kept together with copies of the three channels. Then, the ROI for the three channels is subjected to lowpass filtering, row-wisely in the horizontal direction and column-wisely in the vertical direction, to create a smooth background luminosity surface. This surface does not contain foreground objects such as blood vessels, optic discs, lesions, microaneurysms and others. Three lowpass filters are tested for this purpose, and their efficacy is compared. The outcome is a smooth luminosity surface that estimates the background illumination of the entire ROI. Once the background illumination is established, the luminosity is equalized for all pixels in the ROI, such that every pixel will have the same background brightness. Afterward, the histogram of the ROI is stretched or equalized to enhance the contrast between the foreground objects and the background. Next, the green channel is further improved by adding details from the blue and red channels. Finally, in the post-filtering stage, the intensities of the blue and red channels are adjusted according to their original ratios to the green channel. When all three channels are recombined, the resulting color image looks similar to the original image but shows improved luminosity and contrast. The method is tested on 60 test images. It reduces luminosity variation and increases the contrast of all images. On average, this method achieves a 30% reduction in luminosity variation and a 90% increment in contrast. The proposed method was executed on AMD 5900HS CPU using MATLAB R2021b, and the mean execution time was nearly 2 s on average.

A deep learning-based mobile application for tree species mapping in RGB images

Tree species mapping is an important type of information demanded in different study fields. However, this task can be expensive and time-consuming, making it difficult to monitor extensive areas. Hence, automatic methods are required to optimize tree species mapping. Here, we propose a deep learning-based mobile application tool for tree species classification in high-spatial-resolution RGB images. Several deep learning architectures were evaluated, including mobile networks and traditional models. A total of 2,349 images were used, of which 1,174 images consisted of the Dipteryx alata species and 1,175 images of other local species. These images were manually annotated and randomly divided into training (70%), validation (20%), and testing (10%) subsets, considering the five-fold cross-validation. We evaluated the accuracy and speed (GPU and CPU) of all the implemented deep learning architectures. We found out that the traditional networks have the best performance in terms of F1 score; however, mobile networks are faster. Inception V3 model achieved the best accuracy (F1 score of 97.4%), and MobileNet the worst (F1 score of 83.84%). The MobileNet obtained the best classification speed for CPU (with a mean execution time of 102.8 ms) and GPU (72.4 ms) units. For comparison, Inception V3 achieved a mean execution time of 1058.3 ms for CPU and 634.5 ms for GPU. We conclude that the mobile application proposed can be successfully used to run mobile networks and traditional networks for image classification, but the balance between accuracy and execution time needs to be carefully assessed. This mobile app is a tool for researchers, policymakers, non-governmental organizations, and the general public who intends to assess the tree species, providing a GUI-based platform for non-programmers to access the capabilities of deep learning models in complex classification tasks. • Deep learning models executed in mobile applications can classify tree species in RGB. • Traditional networks are better (F1-score of 97.4%) than mobile (83.8%) networks. • Mobile networks are faster than traditional models in the mean execution time on task.

A GPU-accelerated compute framework for pathogen genomic variant identification to aid genomic epidemiology of infectious disease: a malaria case study.

As recently demonstrated by the COVID-19 pandemic, large-scale pathogen genomic data are crucial to characterize transmission patterns of human infectious diseases. Yet, current methods to process raw sequence data into analysis-ready variants remain slow to scale, hampering rapid surveillance efforts and epidemiological investigations for disease control. Here, we introduce an accelerated, scalable, reproducible, and cost-effective framework for pathogen genomic variant identification and present an evaluation of its performance and accuracy across benchmark datasets of Plasmodium falciparum malaria genomes. We demonstrate superior performance of the GPU framework relative to standard pipelines with mean execution time and computational costs reduced by 27× and 4.6×, respectively, while delivering 99.9% accuracy at enhanced reproducibility.

Model‐based state of X estimation of lithium‐ion battery for electric vehicle applications

In developing an efficient battery management system (BMS), an accurate and computationally efficient battery states estimation algorithm is always required. In this work, the highly accurate and computationally efficient model-based state of X (SOX) estimation method is proposed to concurrently estimate the different battery states such as state of charge (SOC), state of energy (SOE), state of power (SOP), and state of health (SOH). First, the SOC and SOE estimation is performed using a new joint SOC and SOE estimation method, developed using a multi-time scale dual extended Kalman filter (DEKF). Then, the SOP estimation using T-method and 2RC battery model is performed to evaluate the non-instantaneous peak power during charge/discharge. Finally, the battery current capacity estimation is performed using a simple coulomb counting method (CCM)-based capacity estimation with a sliding window. The performance of the proposed SOX estimation method is compared and analyzed. The experimental results show that the estimated SOC and SOE error is less than 1% under considered dynamic load profile at three different temperatures. After the final convergence, the estimated capacity maximum value absolute error is ±0.08 Ah. In addition, the low value of evaluated mean execution time (MET) justifies the high computational efficiency of the proposed method.

Analysis of midwives' situation and the need to measure their workloads.

to analyze midwives' employment situation of midwives and detect their workload measurement needs. a mixed methodology (quantitative and qualitative), observational, descriptive and cross-sectional study. Two phases were carried out. The first methodological phase consisted of conducting semi-structured individual interviews. The second methodological phase analyzed the Nursing Intervention Classification (NIC) interventions that midwives perform in the labor room during daily practice. 90.3% of midwives have work overload, since for 80.6% the midwife-pregnant mother ratio is not well established, since the mean execution time of a Nursing Intervention Classification (NIC) intervention is 20 minutes. For this reason, 100% of the sample considers the development of workload measurement tools to be highly useful. the need to generate an instrument that considers the diversity of Nursing Intervention Classification (NIC) interventions that develop within the labor room is ratified.

Novel co‐estimation strategy based on forgetting factor dual particle filter algorithm for the state of charge and state of health of the lithium‐ion battery

For the battery management system, accurate estimation of the state of charge and state of health is of great significance. Herein, the ternary Li-ion battery is taken as the research object; the second-order resistor-capacitor (RC) equivalent circuit is taken advantage of to characterize the battery performance. A method for calculating the state of health of Li-ion batteries based on capacity fading was established. A novel forgetting factor dual particle filter algorithm is proposed for co-estimation of the state of charge and state of health by combining the forgetting factor and the particle filter algorithm. The state of charge and state of health of Li-ion batteries under Beijing Bus Dynamic Stress Test conditions are evaluated. In the state of charge estimation, the maximum error, mean absolute error, and root mean square error is 1.1395%, 0.4916%, and 0.5145% in Beijing Bus Dynamic Stress Test condition, 1.8125%, 0.6329%, and 0.7955% in Dynamic Stress Test condition, compared with the extended Kalman filter, unscented Kalman filter, and particle filter algorithms, all reduced obviously. In the state of health estimation, compared with the Random Forest and adaptive dual extended Kalman filter-based fuzzy inference system, the mean execution time and convergence time are 11.14 seconds and 0.44 second in Dynamic Stress Test condition and 15.17 seconds and 0.63 second in Beijing Bus Dynamic Stress Test condition; the results show lower computation complexity and faster convergence speed, which play an important role in promoting the further application of lithium-ion batteries. Novelty Statement A new forgetting factor dual particle filter algorithm is proposed to realize the co-estimation the battery state of charge and state of health. The capacity and the state of health were estimated for Dynamic Stress Test and Beijing Bus Dynamic Stress Test conditions, respectively, and the estimation results of the two common conditions were analyzed. SOC estimator is impressive for its high accuracy and quick convergence. The estimated SOH information has better convergence performance and shorter running time.

Development of a small-size laser triangulation displacement sensor and temperature drift compensation method

In practical applications, some special working scenarios need to take into account both narrow space and extreme temperature, which puts more strict requirements on the size and temperature adaptability of the laser triangulation displacement sensor (LTDS). In this paper, a small-size LTDS was developed and the measurement error caused by temperature drift of the sensor was corrected. Firstly, based on direct projection laser triangulation imaging with a reflector unit, a mathematical model of the compact geometry optical structure was established; then, an optimization process of the optical parameters was formulated, and the sensor was assembled. Secondly, a temperature drift error database was built by an extreme temperature experiment, and a general regression neural network model was constructed for error compensation. In addition, root mean square error and execution time were used to evaluate and compare different regression methods. Finally, the experimental results showed that the repeatability accuracy of the proposed sensor was ±2.3 µm, and the temperature linearity was 0.001% F.S. °C−1.

Robust multimodal registration of fluorescein angiography and optical coherence tomography angiography images using evolutionary algorithms

Optical coherence tomography angiography (OCTA) and fluorescein angiography (FA) are two different vascular imaging modalities widely used in clinical practice to diagnose and grade different relevant retinal pathologies. Although each of them has its advantages and disadvantages, the joint analysis of the images produced by both techniques to analyze a specific area of the retina is of increasing interest, given that they provide common and complementary visual information. However, in order to facilitate this analysis task, a previous registration of the pair of FA and OCTA images is desirable in order to superimpose their common areas and focus the gaze on the regions of interest. Normally, this task is manually carried out by the expert clinician, but it turns out to be tedious and time-consuming. Here, we present a three-stage methodology for robust multimodal registration of FA and superficial plexus OCTA images. The first one is a preprocessing stage devoted to reducing the noise and segmenting the main vessels in both types of images. The second stage uses the vessel information to do an approximate registration based on template matching. Lastly, the third stage uses an evolutionary algorithm based on differential evolution to refine the previous registration and obtain the optimal registration. The method was evaluated in a dataset with 172 pairs of FA and OCTA images, obtaining a success rate of 98.8%. The best mean execution time of the method was less than 5 s per image.

Evaluating Automatic Program Repair Capabilities to Repair API Misuses

API misuses are well-known causes of software crashes and security vulnerabilities. However, their detection and repair is challenging given that the correct usages of (third-party) <sc xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">api</small> s might be obscure to the developers of client programs. This paper presents the first empirical study to assess the ability of existing automated bug repair tools to repair <sc xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">api</small> misuses, which is a class of bugs previously unexplored. Our study examines and compares 14 Java test-suite-based repair tools (11 proposed before 2018, and three afterwards) on a manually curated benchmark ( <sc xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">APIRepBench</small> ) consisting of 101 <sc xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">api</small> misuses. We develop an extensible execution framework ( <sc xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">APIARTy</small> ) to automatically execute multiple repair tools. Our results show that the repair tools are able to generate patches for 28 percent of the <sc xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">api</small> misuses considered. While the 11 less recent tools are generally fast (the median execution time of the repair attempts is 3.87 minutes and the mean execution time is 30.79 minutes), the three most recent are less efficient (i.e., 98 percent slower) than their predecessors. The tools generate patches for <sc xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">api</small> misuses that mostly belong to the categories of missing <monospace xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">null</monospace> check, missing value, missing exception, and missing call. Most of the patches generated by all tools are plausible (65 percent), but only few of these patches are semantically correct to human patches (25 percent). Our findings suggest that the design of future repair tools should support the localisation of complex bugs, including different categories of <sc xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">api</small> misuses, handling of timeout issues, and ability to configure large software projects. Both <sc xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">APIRepBench</small> and <sc xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">APIARTy</small> have been made publicly available for other researchers to evaluate the capabilities of repair tools on detecting and fixing <sc xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">api</small> misuses.

High-performance parallel implementations of flow accumulation algorithms for multicore architectures

The calculation of flow accumulation is one of the tasks in digital terrain analysis that is not easy to parallelize. The aim of this work was to develop new, faster ways to calculate flow accumulation and achieve shorter execution times than popular software tools for this purpose. We prepared six implementations of algorithms based on both top-down and bottom-up approaches and compared their performance using 118 different data sets (including 59 subcatchments and 59 full frames) of various sizes but the same area and resolution. Our results clearly show that the parallel top-down algorithm (without the use of OpenMP tasks) is the most suitable implementation for flow accumulation calculations of all we have tested. The mean and median execution times of this algorithm are the shortest in all cases studied. The implementation is characterized by high speedups. The execution times of the parallel top-down implementation are two orders of magnitude shorter compared to the Flow Accumulation tool from ArcGIS Desktop. This is important, considering the performance of popular GIS platforms, where it takes hours to perform the same kind of operations with the use of similar equipment.

Detection of epileptic seizures from compressively sensed EEG signals for wireless body area networks

Wireless body area networks (WBANs) are gaining popularity for tele-monitoring of biomedical signals such as the electroencephalogram (EEG), with diagnosing and monitoring of epileptic seizures being one of the most important applications. Most seizure-detection algorithms cannot be applied directly to the compressed data in WBANs as they require full reconstruction of original EEG signals. In this study, we propose a novel feature for real-time automatic single-channel seizure detection which does not require complete reconstruction of original EEGs. The feature is based on iteratively applying the orthogonal matching pursuit (OMP) algorithm on the compressed EEG data and computing the rate by which the energies of partially reconstructed signals are increased. The feature, i.e. partial energy difference (PED), is then used for classifying seizure and non-seizure states. We also extend this method to the case for multichannel EEG. In multichannel case, the simultaneous OMP (SOMP) with a low number of iterations (1 and 15 iteration) is applied to the compressed data and the difference between the Frobenius norms of partially reconstructed signals is used as a multivariate feature for the classification of seizure and non-seizure states. The proposed features are used to detect seizure intervals in the EEG database provided by the University of Bonn and the CHB-MIT database. The results show that the proposed features can classify seizure epochs from non-seizures even for compression ratios as small as 0.05. The results also show that the proposed single-channel method achieves improvement of up to 4% for the area under the curve (AUC) with significantly less execution time compared to the benchmark matrix determinant (MD) classification approach. The results of applying the proposed multivariate feature to seizure and non-seizure segments of length 5 s from the CHB-MIT database show that it achieves mean AUC values of 0.941, 0.941, and 0.939 with mean execution times of 9.5, 10.7, and 13.1 s for compression ratios of 0.05, 0.1, and 0.2, respectively. Applying a threshold to the PED in a leave-one-out cross-validation (LOO-CV) scenario generates a sensitivity of 0.873,specificity of 0.710, and accuracy of 0.791 at the compression ratio of 0.05. The proposed single- and multichannel features have the potential for deployment in WBANs for the real-time tele-monitoring of epilepsy patients in healthcare applications.

Speeding up an Adaptive Filter based ECG Signal Pre-processing on Embedded Architectures

Medical applications increasingly require complex calculations with constraints of accelerated processing time. These applications are therefore oriented towards the integration of high-performance embedded architectures. In this context, the detection of cardiac abnormalities is a task that remains a high priority in emergency medicine. ECG analysis is a complex task that requires significant computing time since a large amount of information must be analyzed in parallel with high frequencies. Real-time processing is the biggest challenge for researchers, when talking about applications that require time constraints like that of cardiac activity monitoring. This work evaluates the Adaptive Dual Threshold Filter (ADTF) algorithm dedicated to ECG signal filtering using various embedded architectures: A Raspberry 3B+ and Odroid XU4. The implementation has been based on C/C++ and OpenMP to exploit the parallelism in the used architectures. The evaluation was validated using several ECG signals proposed in MIT-BIH Arrhythmia database with a sampling frequency of 360 Hz. Based on an algorithmic complexity study and a parallelization of the functional blocks which present significant workloads, the evaluation results show a mean execution time of 7.5 ms on the Raspberry 3B+ and 0.34 ms on the Odroid XU4. With an efficient parallelization on the Odroid XU4 architecture, real-time performance can be achieved.