Continuous Data Flow Research Articles

Facial Recognition Using Video Streaming

Facial recognition has become an essential technology in numerous applications, including security, surveillance, and human-computer interaction. With the advent of real-time video streaming, facial recognition systems are now tasked with handling the continuous flow of visual data. This paper explores the integration of facial recognition algorithms into video streaming systems, investigating the challenges, methodologies, and applications of real-time face detection and identification. We present a framework for implementing facial recognition in live video streams, assess the performance of different algorithms, and highlight areas for future improvement, including handling large datasets, optimizing computation, and ensuring privacy.

Multi-scale convolutional auto encoder for anomaly detection in 6G environment

With the increasing deployment of 6G networks in all industries, there is a growing risk of vulnerability and complexity due to continuous data flow from edge devices to specialized computers. One potential threat to 6G networks is fuzzing attacks, which involve sending random and invalid inputs to identify vulnerabilities and flaws. A successful fuzzing attack could compromise the network’s protocols, interfaces, and security mechanisms, leading to potential disruption or compromise of critical infrastructure. The proposed framework involves gathering relevant data from different sources within the 6G network and pre-processing it to prepare for analysis. Relevant features that indicate the presence of anomalies are identified using Tuna Swarm Optimization (TSO) Algorithm. The Multi-Scale Convolutional Auto Encoder (MSCAE) is then utilized by the proposed model to extract the feature and classify data. The intrusion Detection System is built to monitor and classify nodes producing anomalies. The proposed model is assessed utilizing various metrics, including recall, precision, accuracy, detection latency, and F1 score. The proposed framework achieved 97.50% accuracy, 94.81% precision, 93.50% F1-score, and 94.50% recall with notable improvements that surmount the deficiencies of previous studies. The results demonstrate that the proposed algorithm is more efficient and safe than current edge security methods, potentially mitigating the risk of fuzzing attacks in 6G networks.

Managing the E-commerce Data Deluge through Text Analytics and Web Management (Overview of Amazon.com)

Today, more than 80% of the big data handled in the e-commerce industry is text and unstructured data. Text analytics is an automated process for analyzing text and extracting useful information from it. It can discover trends and relationships in data. Web analytics is the collection, processing, and analysis of data in order to draw conclusions to optimize usability on a website. Web analytics can be used to improve the usability of a site by analyzing user behavior patterns such as time spent on the site, abandonment rates, most frequently accessed products, click-through rates, etc. It can also help analyze the interests of different user demographics, as it tracks granular details such as user demographics, age and gender, geography, and devices used as data. In order to obtain UpToDate information, the business can utilize business intelligence for real-time data processing, then they can practice stream analysis to analyse continuous flow of data. For instance, the business can collect instant information in Twitter or other social media and analyse it by using social media analysis. For website management, business can practice web analysis to analyse the customer’s behaviours. Tracking the customer’s activity, page view and conversion rate is important for business to analyse how to improve the website performance. Text analytics of comments received on Amazon can be used to group text data and produce results in terms of word frequency distribution and sentiment analysis. Text analytics could be used for decision making, improving service quality, and developing new business models.

Decomposition of non-performing loans dynamics into a debt-servicing capacity and a risk taking indicators

Using Colombian credit vintage data, we decompose non-performing loans into two main components: one capturing the evolution of borrowers’ payment capacity and another reflecting changes in the credit risk assumed by banks when granting loans. We employ intrinsic estimators and penalized regression techniques to address the perfect multicollinearity inherent in the model. Our analysis reveals that these two components have evolved differently over time and that they interact with the real and credit cycles distinctively. In particular, we find that a favorable economic environment and loose financial conditions improve the payment capacity of borrowers to meet their obligations, but coincide with increased risk-taking by financial institutions. Finally, we advocate for the adoption of this decomposition as a policy tool, easily applicable by financial and economic authorities with access to a continuous flow of credit vintage data. This methodology facilitates the identification of credit risk origins, thereby informing economic policies aimed at mitigating systemic financial risks.

РАДІОВИМІРЮВАЛЬНИЙ ПЕРЕТВОРЮВАЧ ВОЛОГОСТІ ҐРУНТУ НА БАЗІ ПЕРВИННОГО СЕНСОРА YL-69

Soil moisture monitoring in agriculture is becoming an integral part of ensuring optimal conditions for plant growth and achieving high yields, as well as an important environmental aspect for preserving natural resources and ecosystems. The increasing popularity of wireless sensor networks for measuring soil moisture underscores the need for the development of modern technologies in this field. This enables a continuous flow of real-time data on soil conditions, which is crucial for effective agricultural management and ecological balance. The paper presents the development of a radio-based soil moisture transducer with a moisture-sensitive resistive element based on a transistor structure with negative resistance. The YL-69 resistive sensor was utilized as the primary soil moisture sensor. An analytical expression for the conversion function of the radio-based soil moisture transducer is determined. Computer simulation of the transducer's operation was conducted in the OrCAD environment, resulting in a family of current-voltage characteristics for the investigated radio-based transducer. A region with negative differential resistance is observed on the current-voltage characteristics. An experimental conversion function of the radio-based soil moisture transducer based on the YL-69 primary sensor was obtained. From the conversion function, it is evident that as soil moisture increases in the range from 0% to 50%, the frequency of the output signal of the radio-based soil moisture transducer decreases from 680 kHz to 452 kHz. The highest sensitivity value of the radio-based soil moisture transducer is achieved in the soil moisture measurement range from 20% to 45% and is 8.7 kHz/%. The operation of radio-based soil moisture transducers is based on the functional dependence of semiconductor device impedance on soil moisture, which opens up wide prospects for improving agricultural technologies and preserving the environment. This transducer can be used for measuring soil moisture in agriculture, which can be utilized to create automatic systems for monitoring soil moisture and watering plants.

Effectiveness of a hybrid approach in integrating GLP-1 agonists and lifestyle guidance for obesity and pre-diabetes management: RWE retrospective study

AimEmerging anti-obesity pharmacotherapy provides an option to correct maladaptive physiological and hormonal changes associated with obesity. One of the widely used medications in this context is glucagon-like peptide 1 (GLP-1) agonists. However, the misuse of these medications without any guidance and monitoring of lifestyle modifications can lead to unfavorable outcomes. The study aims to evaluate the effectiveness of a hybrid care model, incorporating GLP-1 and GLP-1/GIP agonist therapies, in managing obese patients with/without pre-diabetes. This study showcases the midway results of a 6-month program, which includes a multidisciplinary care team and digital technology for continuous engagement and monitoring of patients, both in-clinic and remotely. MethodsIn a retrospective observational study, 115 participants were treated with GLP-1s (semaglutide, tirzepatide, and liraglutide). Physicians, dietitians, and coaches worked together to support behavioral changes using a dedicated app provided to patients. At the care team end, an integrated portal enabled continuous data flow allowing for the care team to provide personalized care via chat at regular intervals. Data collected included food logs, continuous glucose monitoring (CGM), and digital biomarkers such as sleep and activity. ResultsAt the midpoint of the program, participants exhibited statistically significant improvements in various metabolic parameters. Mean weight reduction was 8 %, with significant reductions in BMI, fat mass, and cholesterol levels. 24 (20.9 %) of patients lost ≥5 % of body weight, 55 (47.8 %) patients lost ≥10 % weight, and 36 (31.3 %) patients lost ≥15 % weight. Sub-analysis of pre-diabetic patients (n=36) demonstrated substantial improvements, including control of pre-diabetes in 80.6 % of cases and reduced HbA1c levels back to normoglycemia (5.39 ± 0.27). ConclusionThe Zone.Health's program, which combines pharmacotherapy with continuous engagement and monitoring to enable sustainable lifestyle modifications, demonstrated significant improvements in weight, body composition, and metabolic markers. Pre-diabetes was also effectively addressed. It is necessary to conduct further research to assess the long-term sustainability and optimal adoption of such care models into clinical practice.

Unleashing the potential of digital twins: a new era with aeronautics 4.0

Abstract* Introduction The aerospace value chain consists of several processes, and digitizing it first requires an assessment of these processes and their ability to be transferred to the fully integrated digital thread perspective of smart factories. A digital thread refers to the continuous flow of data and information related to a product throughout its life cycle, integrating and connecting all aspects of a product’s journey. Within this framework, digital twin technology, an essential element of Industry 4.0, comprises the implementation of a virtual presentation of a physical object, system, or process. It brings together the digital replica not only of the physical attributes but also of the behavioral performance of the physical twin. Methods To achieve a digital thread perspective for aeronautics, a research agenda is proposed, including all breakpoints in the current value chain through a PESTLE Analysis, which defines the value-added areas targeted by this transition to digital technology. An aeronautics 4.0 model Digital Thread Twin Smart Aeronautics D2TSAero is proposed, which provides a new perspective in the aeronautics field by integrating its main ecosystems into an interconnected model made possible by the integration of digital twin agent instances. Proof of Concept A use case that deals with the dependability management of an aircraft fuel distribution system is presented. Based on these results, we can see that the proposed twin model can help in reducing real system parts down time, and it can also improve the management of maintenance across the system life cycle by offering a single source of trust for all stakeholders involved in the digital thread cycle of the real twin. Conclusion A forward-looking perspective on the future of aeronautics with this integrated approach is presented, summarizing all the discussed points and the importance of digital twins in supporting the digitalization of the field.

Machine learning technology for predicting intraocular lens power. Diagnostics data generalization

Backgraund. Implantation of recent IOLs allows ophthalmologists to effectively solve the problems of surgical rehabilitation of patients with cataracts. The degree of improving the patient's visual functions is directly dependent on the accuracy of the preoperative calculation of the optical power of IOLs. The most famous formulas to calculate this indicator are SRK II, SRK/T, Hoffer-Q, Holladay II, Haigis, Barrett. All of them work well for an "average patient", however, they are not adequate enough at the boundaries of input variables ranges.
 Aims. To study the possibility of using mathematical models obtained as a result of deep learning of artificial neural networks (ANN models) to generalize data and predict the optical power of modern intraocular lenses.
 Materials and methods. ANN models were trained on large-scale samples, including depersonalized data for patients of ophthalmology clinic. Data provided in 2021 by ophthalmologist K.K. Syrykh, and reflect the results of both preoperative and postoperative observation of patients. The source file used to build the ANN model includes 455 records - 26 columns of input factors and one column for the output factor - calculating IOL (diopters). To conveniently build ANN models, we used a simulator program, previously developed by the authors.
 Results. The resulting models, in contrast to the traditionally used formulas, reflect the regional specificity of patients to a much greater extent. They also make it possible to retrain and optimize the structure based on newly received data, which allows taking into account the non-stationarity of the object. A distinctive feature of such ANN models in comparison with the well-known formulas SRK II, SRK/T, Hoffer-Q, Holladay II, Haigis, Barrett, widely used in the surgical cataract treatment in ophthalmology, is their ability to take into account a significant number of recorded input quantities, which makes it possible to reduce the mean relative error in calculating the optical power of IOL from 10–12% to 3.5%.
 Conclusions. This work shows the fundamental possibility of generalizing a significant amount of empirical data on calculating the optical power of the IOL using training ANN models that have a significantly larger number of input variables than when using traditional formulas and methods. The results obtained allow, in principle, to build an intelligent expert system with a continuous flow of new data from a source and a step-by-step retraining of the ANN model.

Vibration monitoring of structures with indirect load identification and Kalman update

The German Research Foundation (DFG) has launched in September 2022 the priority program SPP 2388 100+ to develop new methods for digital representation, SHM and lifetime management of complex structures, due to the continuously increasing amount of old infrastructure buildings. The present contribution is prepared within the LEMOTRA project as a part of SPP 2388 100+. Among various SHM methods, the approach based on the Kalman update for data assimilation between model and measurement is applied and further developed to create a kind of functional digital twin for SHM. For this purpose, a sound numerical model and a measurement system with a continuous data flow are introduced to provide online predictions of the state and response parameters of the structure. System changes resulting from damage or aging processes can be detected and localized, provided the measurement and the model prediction deal with the same cause. Thus, the load identification is a necessary prerequisite for reliable data assimilation techniques. A two-step update procedure is proposed and applied in this context. At first, one part of the measurement system is used for the load identification. A cluster structure of convolutional neural networks (CNNs) was developed, trained and calibrated to extract load characteristics such as load magnitudes, load velocities or the number of vehicles on the bridge from multiple acceleration signals. This information is then used to reconstruct the actual load. In the second step, a different set of sensors is used for the data assimilation. In contrast to the first set of measurement locations, the measurement data from these sensors should be sensitive to potential system changes or damage. Here, the identified load is used as input for the model predictions which are then compared to the measurement data. A combination of different ensemble based Kalman filters (KF) provides a sequential update of the state parameters (e.g. displacement, velocity, acceleration) and the model parameters (e.g. stiffness, mass, damping). The cluster CNN approach is tested numerically, and the data assimilation technique is tested on a laboratory structure.

Workshop Synthesis: Leveraging Household Survey and Fare Collection Data in Public transport

Public transport is key to the sustainability of the mobility of billions of people around the world. Many sources of data are available to help its planning, namely surveys, fare collection system, and crowdsourced applications. This paper is the synthesis of a workshop that was aimed at characterising these sources of data and find how can we better use them to complement each other. Ultimately, the goal is to obtain a continuous flow of rich and meaningful data that will feed tools and models, targeting better public transport systems. The paper presents the data sources with their strengths and weaknesses, and then discuss about the challenges related to data fusion and the evolving potentialities brought by the increasing availability of this data.

Simulation of Vehicular Bots-Based DDoS Attacks in Connected Vehicles Networks

Connected vehicles are more vulnerable to attacks than wired networks since they involve rapid mobility, continuous data flow across connected nodes, and dynamic network design in a distributed network environment. Distributed Denial of Service (DDOS) is one of the most common and dangerous security attacks on connected vehicle networks. Attackers can remotely control malicious nodes that are programmed to attack other nodes known. The compromised nodes are known as botnets, which will constantly flood the target nodes with User Datagram Protocol (UDP) packets, disrupting the target nodes data flow and operation. Hence, the goal of this research is to create and simulate a vehicular bot-based Distributed Denial of Service (DDoS) assault in connected vehicle networks. A simulation-based methodology is implemented to observe the impact of the number of bots, DDoS rate, and maximum bulk packet size on network performance. Using the NS-3 network simulator, 73 random mobile vehicle nodes with up to 100 vehicle bots were simulated, and the results are discussed. Regardless of the computational constraints, the findings from this study adds to understanding the risks and problems associated with data transmission by analyzing the impact of vehicular bot-based DDoS attacks on connected vehicle performance. Doi: 10.28991/HIJ-2023-04-04-014 Full Text: PDF

Personalized Federated Continual Learning for Task-Incremental Biometrics

In the age of Internet of Things where information is explosively growing, people pay more attention on personal privacy. In the real-world task-incremental scenario for biometrics, every edge device faces continuous task flows of private data without communication with others. Security and Performance are the primary concerns in identity authentication, and Federated Continual Learning (FCL) is a promising solution. In this paper, we design a personalized federated continual learning framework to solve the problem of sequentially identification in every distributed device. For each client, we create an adaptive continual meta-learning model called cTD-αMAML, aiming to align the gradients of previous and new tasks and to make the learning-rate model learnable. For central aggregation, the server gathers the meta-initialization from every local update and allocates the updated global meta-initialization to clients. We propose an extension of FedAvg to locally reserve the learnable learning-rate network to realize the personalization of clients. Results prove that in Continual Learning, our cTD-αMAML can learn to learn the seen tasks and avoid catastrophic forgetting. And in Federated Continual Learning, our personalized method realizes the knowledge transferring across clients, meanwhile improving the local performance and reducing the communication cost. In this way, the proposed personalized federated continual learning framework can obtain a biometric template that is able to learn the expression space for new tasks with rapid adaption.

Enabling coordination in energy communities: A Digital Twin model

Starting from the EU vision for Energy Communities (EC), our purpose is to support them by proposing a Digital Twin (DT) that includes a bi-level optimization model to deliver coordination, economic, social, and environmental benefits to its members that can be quantified as Key Performance Indicators (KPI). The diversity of EC members from the size and interest perspectives leads us to consider a bi-level optimization model. It offers support to individual consumers/prosumers (first level) and coordination for EC (second level). This model is embedded into a DT that replicates the EC and the operation of individual entities such as consumers/prosumers and public assets. The DT is created as an automatic assistant with two components: iEMS – as a member's assistant and eEMS – as an EC assistant. These components optimize the schedule, generate bids for the Local Electricity Market (LEM) and control the flexible appliances of the participants to deal with deficits and surpluses. The DT receives input from EC members, LEM, metering system and improves the operation of the EC in a two-way continuous exchange data flow. Furthermore, it is a reliable framework to test and improve models, regulations and policies in emergent EC as DT provides alternatives regarding its functionalities: optimization, market operation, setting the clearing price, settlement, value sharing for distributing benefits, etc. It can be extended to support grid operators to design tariffs, testing regulation and offer additional energy services. The proposed DT model is tested within an EC case study, both on a seasonal and an annual basis. The average trading index on LEM is 0.6 during the summer and 0.3 during the winter months, while the Degree of Local Sufficiency (DLS) for the EC is 0.45 in summer and 0.28 in winter. Moreover, the proposed LFM model reduces the exchanges with the main grid by an average of 45 kW in summer, that represents almost 15% of the average exchange at peak hours.

A reliable adaptive prototype-based learning for evolving data streams with limited labels

Data stream mining presents notable challenges in the form of concept drift and evolution. Existing learning algorithms, typically designed within a supervised learning framework, require class labels for all data points. However, this is an impractical requirement given the rapid pace of data streams, which often results in label scarcity. Recognizing the realistic necessity of learning from data streams with limited labels, we propose an adaptive, data-driven, prototype-based semi-supervised learning framework specifically tailored to handle evolving data streams. Our method employs a prototype-based data representation, summarizing the continuous flow of streaming data using dynamic prototypes at varying levels of granularity. This technique enables improved data abstraction, capturing the underlying local data distributions more accurately. The model also incorporates reliability modeling and efficient emerging class discovery, dynamically updating the significance of prototypes over time and swiftly adapting to local concept drift. We further leverage these adaptive prototypes to intuitively detect concept evolution, i.e., identifying novel classes from a local density perspective. To minimize the need for manual labeling while optimizing performance, we incorporate active learning into our method. This method employs a dual-criteria approach for data point selection, considering both uncertainty and local density. These manually labeled data points, together with unlabeled data, serve to update the model efficiently and robustly. Empirical validation using several benchmark datasets demonstrates promising performance in comparison to existing state-of-the-art techniques.

A Fully Online Clustering Approach for Enhanced Performance of Health Information System

Health Data clustering is a significant research direction aiming to extract knowledge from a continuous health data flow to support online health decisions. However, processing health data clusters is still a challenging task. Existing clustering approaches are subject to various limitations in terms of considering the neighbor clusters and conducting multiple operations during the maintenance process. In this paper, we model, design and implement a novel framework called IClustMaint for efficiently clustering and maintaining health data clusters incrementally. A two-phase algorithm is embedded in the framework. We first employ the Principal Component Analysis (PCA) method to efficiently reduce the high costs of the initial clustering phase. Next, in the maintenance phase, we propose the incremental Cluster maintenance (ICM) approach for managing the generated cluster during a period of time. Technically, when the data clusters are evolving over time and need to be maintained frequently, the ICM approach improves the performance of cluster maintenance by only tracking the edge points. The experimental results on a real medical dataset verify the efficiency of the proposed approaches.

Acceleration-based condition monitoring of track longitudinal level using multiple regression models

In this paper, a condition monitoring system for railway track geometry is presented. The methodology has been designed for high-speed application, where the train travels at the maximum allowed speed for most of the trip. The system is designed to rely on acceleration data recorded by in-service vehicles to provide estimations of the track longitudinal level, based on pre-built regression models. It exploits synthetic indicators sampled over predefined track sections 100 m long. Different predictors are considered, computed both from acceleration data and from track geometry measured by the diagnostic train. The proposed modelling strategy allows distinguishing between isolated and distributed defects that populate the railway track as well as reproducing the evolution over time of the maximum longitudinal level registered in the considered track section; moreover, also accurate predictions of the defect amplitude are made. The results have been validated against track geometry data recorded by the diagnostic train during a monitoring period of 2 years. It is proven that the proposed system could support current maintenance strategies, providing a continuous flow of data to monitor the track infrastructure.

Applying Neural Networks to Recover Values of Monitoring Parameters for COVID-19 Patients in the ICU

From the moment a patient is admitted to the hospital, monitoring begins, and specific information is collected. The continuous flow of parameters, including clinical and analytical data, serves as a significant source of information. However, there are situations in which not all values from medical tests can be obtained. This paper aims to predict the medical test values of COVID-19 patients in the intensive care unit (ICU). By retrieving the missing medical test values, the model provides healthcare professionals with an additional tool and more information with which to combat COVID-19. The proposed approach utilizes a customizable deep learning model. Three types of neural networks, namely Multilayer Perceptron (MLP), Long/Short-Term Memory (LSTM), and Gated Recurrent Units (GRU), are employed. The parameters of these neural networks are configured to determine the model that delivers the optimal performance. Evaluation of the model’s performance is conducted using metrics such as Root Mean Square Error (RMSE), Mean Absolute Percentage Error (MAPE), and Mean Absolute Error (MAE). The application of the proposed model achieves predictions of the retrieved medical test values, resulting in RMSE = 7.237, MAPE = 5.572, and MAE = 4.791. Moreover, the article explores various scenarios in which the model exhibits higher accuracy. This model can be adapted and utilized in the diagnosis of future infectious diseases that share characteristics with Coronavirus Disease 2019 (COVID-19).

Telemedicine and Its Application in Cystic Fibrosis.

The care of cystic fibrosis (CF) traditionally consists of regular visits to the clinic where a multidisciplinary team can visit the patient, adjust treatments and monitor the disease. During the COVID-19 pandemic when access to hospitals and medical environments was very limited, the role of telemedicine was crucial to keep in touch with patients with chronic diseases such as CF. Increasing evidence demonstrates that electronic health can successfully support healthcare professionals in the management of people with CF. The use of devices connected to digital platforms or smartphones results in a continuous flow of data that can be shared with the clinician and the team in order to improve the knowledge of patients' diseases and the level of care needed. This narrative review aims to describe the application of telemedicine in CF disease with pros and cons. A literature analysis showed that telemedicine has several advantages in the management of patients with CF. With the evolving support of digital technology, telemedicine can promote clinical visits, adherence to daily treatment, including respiratory physiotherapy and physical exercise, early identification of pulmonary exacerbations and management of psychological issues. The main disadvantages are missed physical exam findings, lack of physical contact that can prevent conversation on sensitive topics, lack of access to technology and lack of technological skills. Furthermore, healthcare operators need appropriate training for telemedicine systems and need time to organise and analyse data generated remotely, which may increase the burden of daily work. Hybrid personalised care models that marge telemedicine and traditional care can be an ideal solution.

A digital engineering framework to facilitate automated data exchange between geometric inspection and structural analysis

Digital engineering relies on continuous and interoperable flow of data across the various stages within the product lifecycle, which permits engineers to make more-informed downstream decisions in the lifecycle. A model-based feature information network (MFIN) was developed to enable storing, tracking, and retrieving externally stored metadata in the context of the geometric features of the corresponding computer aided-design (CAD) model. In this paper, we use the MFIN framework to dynamically capture deviations in the part geometry via physical measurements, and to create a new instance of the CAD model and the associated structural analysis. The contribution of this work is the creation of automated workflow software, which enables engineers to improve decision making by assessing (i) as-manufactured part geometries that fall out of specifications during the materials review board or (ii) in-service damages in parts during the maintenance, repair, or overhaul stages of the lifecycle. We demonstrate a use case to assess the structural integrity of a turbofan blade that had experienced foreign object damages during service. The automated flow of information from geometric inspection within structural analysis, enabled by MFIN, presents opportunities for effectively assessing products by utilizing their current geometries and improving decision making during the product lifecycle.

Accelerating GRAPPA reconstruction using SoC design for real-time cardiac MRI

Real-time cardiac MRI is a rapidly developing area of research that has the potential to improve the diagnosis and treatment of cardiovascular diseases. However, the acquisition of high-quality real-time cardiac MR (CMR) images is challenging as it requires a high frame rate and temporal resolution. To overcome this challenge, there have been recent efforts on several approaches including hardware-based improvements and image reconstruction techniques such as compressed sensing and parallel MRI. The use of parallel MRI techniques such as GRAPPA (Generalized Autocalibrating Partial Parallel Acquisition) is a promising approach for improving the temporal resolution of MRI and expanding its applications in clinical practice. However, the GRAPPA algorithm involves a significant amount of computation, particularly for high acceleration factors and large datasets. This can result in long reconstruction times, which can limit the ability to achieve real-time imaging or high frame rates. One solution to this challenge is to use specialized hardware i.e. field-programmable gate arrays (FPGAs). In this work, a novel 32-bit floating-point FPGA-based GRAPPA accelerator is proposed with an aim to reconstruct high-quality cardiac MR images at higher frame rates, making it well suited for real-time clinical applications. The proposed FPGA-based accelerator consists of custom-designed data processing units named as dedicated computational engines (DCEs) that allow for a continuous flow of data between the calibration and synthesis stages of GRAPPA reconstruction process. This greatly increases the throughput and reduces the latency of the overall proposed system. Moreover, a high-speed memory module (DDR4-SDRAM) is integrated with the proposed architecture to store the multi-coil MR data. An on-chip quad-core ARM Cortex-A53 processor is used to manage access control information required for data transfer between the DCEs and DDR4-SDRAM. The proposed accelerator is implemented on Xilinx Zynq UltraScale + MPSoC using high-level synthesis (HLS) and hardware descriptive language (HDL) with an aim to explore the trade-offs between the reconstruction time, resource utilization and design effort. Several experiments have been performed using in-vivo cardiac datasets i.e. 18-receiver coil and 30-receiver coil to evaluate the performance of the proposed accelerator. A comparison is performed with the contemporary CPU and GPU-based GRAPPA reconstruction methods in terms of reconstruction time, frames-per-second and reconstruction accuracy (RMSE and SNR). The results show that the proposed accelerator achieves speed-up factors up to 121× and 9× as compared to the contemporary CPU-based and GPU-based GRAPPA reconstruction methods, respectively. Moreover, it has been demonstrated that the proposed accelerator can achieve reconstruction rates of up to ∼27 frames-per-second while maintaining the visual quality of the reconstructed images.