Description:TECHNICAL FIELD
[0001] The present disclosure relates to the field of electrocardiogram (ECG) signal processing. More particularly, the present disclosure relates to a system and a method for QRS complex detection, beat classification, and arrhythmia detection in electrocardiogram (ECG) signals.

BACKGROUND
[0002] The following description of the related art is intended to provide background information pertaining to the field of the present disclosure. This section may include certain aspects of the art that may be related to various features of the present disclosure. However, it should be appreciated that this section is used only to enhance the understanding of the reader with respect to the present disclosure, and not as admissions of the prior art.
[0003] Electrocardiogram (ECG) signal processing is essential for analyzing the heart’s electrophysiological activity. Various segments of an ECG waveform correspond to different phases of cardiac function, with the QRS complex representing ventricular depolarization. The QRS complex representing ventricular depolarization is a critical feature for evaluating heart health and diagnosing cardiac arrhythmias. Traditional QRS detection methods utilize digital signal processing (DSP) techniques, including noise filtering, low-frequency suppression, peak enhancement, and threshold-based detection. The existing methods aim to isolate and enhance the QRS complex while reducing interference from other ECG components, such as P and T waves. However, despite their widespread use, these techniques face several limitations that impact their reliability in real-world applications.
[0004] One major disadvantage of traditional QRS detection techniques pertains to sensitivity to noise. ECG signals often contain baseline wander, power line interference, and high-frequency artifacts, which can significantly impact detection accuracy. Parameter tuning is another challenge, as threshold values and filtering parameters must be manually adjusted for different patient populations and signal qualities, making the process time-consuming and inconsistent. Additionally, traditional methods struggle with false positives and negatives, particularly in cases of rapid heart rates, arrhythmias, or abnormal waveforms. The variability of QRS morphology across individuals further complicates detection, necessitating frequent manual adjustments and reducing the robustness of these techniques.
[0005] Traditional QRS detection methods also exhibit limited adaptability in dynamic environments, such as ambulatory monitoring or exercise ECG, where signal characteristics can change rapidly. Detecting low-amplitude QRS complexes remains difficult due to reliance on threshold-based methods, increasing the likelihood of missed detections. Multi-lead ECG analysis adds another layer of complexity, as integrating information from different leads requires additional processing steps. Moreover, abnormal rhythms such as ventricular fibrillation can significantly alter QRS morphology, rendering traditional detection methods unreliable. The above limitations have driven the development of more advanced and adaptive techniques, such as machine learning-based approaches, which offer improved accuracy, noise resilience, and real-time adaptability in ECG signal analysis.
[0006] There is, therefore, a need in the art to provide a system and a method for QRS complex detection, beat classification, and arrhythmia detection in electrocardiogram (ECG) signals that can overcome the shortcomings of the existing prior arts.

OBJECTS OF THE PRESENT DISCLOSURE
[0007] Some of the objects of the present disclosure, which at least one embodiment herein satisfies are as listed herein below.
[0008] It is an object of the present disclosure to provide a system and a method for QRS complex detection, beat classification, and arrhythmia detection in electrocardiogram (ECG) signals using a deep learning model.
[0009] It is another object of the present disclosure to provide a system and a method provides a non-invasive, cost-effective, reliable that enhances QRS detection accuracy, ensuring improved precision even in noisy, low-amplitude, or motion affected ECG signals.
[0010] It is another object of the present disclosure to provide a system and a method employs an adaptive thresholding mechanism driven by deep learning models that dynamically adjust based on real-time ECG variations, reducing false positives and false negatives in R-peak detection.
[0011] It is another object of the present disclosure to provide a system and a method provides a robust, real-time solution which supports multi-lead ECG analysis, leveraging spatial features to improve detection accuracy and providing a comprehensive view of cardiac activity.
[0012] It is another object of the present disclosure to provide a system and a method that detects and classifies QRS complexes and heartbeat types into normal and abnormal categories using a deep learning-based approach, facilitating automated real-time cardiac monitoring and diagnostic support.
[0013] It is another object of the present disclosure to provide a system and a method that detects abnormalities in ECG signals such as A-Fib, ST changes, any conduction abnormality such as AV blocks, LBBB, RBBB etc.
[0014] It is another object of the present disclosure to provide a system and a method provides a low-latency processing, enabling real-time ECG monitoring in wearable health devices, clinical settings, and ambulatory ECG systems.

SUMMARY
[0015] This summary is provided to introduce certain objects and aspects of the present disclosure in a simplified form that are further described below in the detailed description. This summary is not intended to identify the key features or the scope of the claimed subject matter.
[0016] The present disclosure relates to the field of electrocardiogram (ECG) signal processing. More particularly, the present disclosure relates to a system and a method for QRS complex detection, beat classification, and arrhythmia detection in electrocardiogram (ECG) signals using a deep learning model.
[0017] An aspect of the present disclosure relates to a system for automated QRS complex detection, beat classification, and arrhythmia detection in electrocardiogram (ECG) signals. The system may include processors, and a memory coupled to the processors. The memory stores instructions executable by the processors, enabling the system to receive ECG signals from a user. The system can filter noise including baseline wander and high-frequency artifacts from the ECG signals to improve signal quality. The system can segment the ECG signals into a fixed-length intervals and extract relevant features. The system can analyse the extracted relevant features using predefined parameters to generate probability masks for QRS complexes detection. The system can identify R-peaks within the QRS complexes and assign corresponding confidence scores. The system can classify the identified QRS complexes into predefined heartbeat categories based on the corresponding confidence scores using a neural network model. The system can display the classified QRS complexes on a computing device associated with the user for clinical interpretation.
[0018] In an aspect, the system may be configured to apply a dynamic thresholding mechanism to detect the R-peaks within the QRS complexes.
[0019] In an aspect, the system can analyse the one or more extracted relevant features based on the one or more predefined parameters to improve the QRS complexes detection and beat classification, where the one or more predefined parameters can include at least one of a duration, an amplitude, and a polarity.
[0020] In an aspect, the relevant features are extracted from the ECG signals by using a deep residual convolutional neural network (ResNet) along with a convolutional block attention module (CBAM).
[0021] In an aspect, the CBAM module can include at least one of a channel attention sub-module and a spatial attention sub-module configured to enhance one or more ECG signals.
[0022] In an aspect, the system can include a Bidirectional Gated Recurrent Unit (Bi-GRU) layer configured to interpret the one or more ECG signals by using at least one of: temporal dependencies and long-term dependencies.
[0023] In an aspect, the system can include the predefined heartbeat categories which includes at least one of: normal beats, premature ventricular contractions (PVCs), and bundle branch blocks (BBBs).
[0024] In an aspect, the extracted features may include at least one of a temporal feature, a morphological feature, and a spatial feature.
[0025] In an aspect, the temporal feature pertains to timing and duration of ECG events for analyzing the one or more ECG signals, wherein the temporal feature may include at least one of RR interval, QRS duration, and PR interval. The morphological feature pertains to one or more waveform characteristics of ECG components for classifying at least one of normal heart beat and abnormal heart beats, wherein the morphological feature may include at least one of R-peak amplitude, QRS slope, and ST segment elevation. The spatial feature pertains to a three-dimensional view of the one or more ECG signals, wherein the spatial feature may include at least one of electrical axis of the heart, multi-lead QRS amplitude variability and inter-lead correlation.
[0026] An aspect of the present disclosure relates to a method for QRS complex detection, beat classification, and arrhythmia detection in electrocardiogram (ECG) signals. The method include the steps of: receiving, by a system, ECG signals from a user and filtering noise including baseline wander and high-frequency artifacts from the one or more ECG signals to improve signal quality. The method include the steps of: segmenting, by the system, the ECG signals into fixed-length intervals of ECG segments and extracting relevant features. The method include the steps of: analysing, by the system, the extracted relevant features using predefined parameters for generating probability masks for detecting QRS complexes. The method include the steps of: identifying, by the system, R-peaks within the QRS complexes and assigning corresponding confidence scores. The method include the steps of: classifying, by the system, the identified QRS complexes into predefined heartbeat categories based on the corresponding confidence scores using a neural network model. The method include the steps of: displaying, by the system, the classified QRS complexes on a computing device associated with the user for clinical interpretation.
[0027] Various objects, features, aspects, and advantages of the present disclosure will become more apparent from the following detailed description of preferred embodiments, along with the accompanying drawing figures in which numerals represent like features.
[0028] Within the scope of this application, it is expressly envisaged that the various aspects, embodiments, examples, and alternatives set out in the preceding paragraphs, in the claims and/or in the following description and drawings, and in particular the individual features thereof, may be taken independently or in any combination. Features described in connection with one embodiment are applicable to all embodiments, unless such features are incompatible.

BRIEF DESCRIPTION OF THE DRAWINGS
[0029] In the figures, similar components and/or features may have the same reference label. Further, various components of the same type may be distinguished by following the reference label with a second label that distinguishes among the similar components. If only the first reference label is used in the specification, the description is applicable to any one of the similar components having the same first reference label irrespective of the second reference label.
[0030] FIG. 1 illustrates an exemplary network architecture of the proposed system for QRS complex detection, beat classification, and arrhythmia detection in electrocardiogram (ECG) signals, in accordance with an embodiment of the present disclosure.
[0031] FIG. 2 illustrates an exemplary block diagram architecture of the system for QRS complex detection in electrocardiogram (ECG) signals, in accordance with an embodiment of the present disclosure.
[0032] FIG. 3 illustrates a flow diagram for implementing a method for QRS complex detection, beat classification, and arrhythmia detection in electrocardiogram (ECG) signals, in accordance with an embodiment of the present disclosure.
[0033] FIG. 4 illustrates an exemplary computer system in which or with which embodiments of the present invention can be utilized in accordance with embodiments of the present disclosure.

DETAILED DESCRIPTION
[0034] In the following description, for the purposes of explanation, various specific details are set forth in order to provide a thorough understanding of embodiments of the present disclosure. It will be apparent, however, that embodiments of the present disclosure may be practiced without these specific details. Several features described hereafter can each be used independently of one another or with any combination of other features. An individual feature may not address all of the problems discussed above or might address only some of the problems discussed above. Some of the problems discussed above might not be fully addressed by any of the features described herein.
[0035] The ensuing description provides exemplary embodiments only, and is not intended to limit the scope, applicability, or configuration of the disclosure. Rather, the ensuing description of the exemplary embodiments will provide those skilled in the art with an enabling description for implementing an exemplary embodiment. It should be understood that various changes may be made in the function and arrangement of elements without departing from the spirit and scope of the invention as set forth.
[0036] In the following description, numerous specific details are set forth in order to provide a thorough understanding of embodiments of the present disclosure. It will be apparent to one skilled in the art that embodiments of the present disclosure may be practiced without some of these specific details.
[0037] Various embodiments of the present disclosure will be explained in detail with respect to FIGs. 1-4.
[0038] The present disclosure addresses these challenges by introducing a deep learning model that leverages neural networks trained on large ECG datasets. The model enhances QRS detection accuracy, classifies different heartbeat types, and identifies arrhythmias with high precision, even in noisy and dynamic conditions. By learning inherent ECG patterns, the proposed system adapts to different patient profiles, heart conditions, and signal variations, making it more reliable and efficient for clinical applications and real-time monitoring in wearable devices.
[0039] FIG. 1 illustrates an exemplary network architecture 100 of the proposed system 102 for QRS complex detection, beat classification, and arrhythmia detection in electrocardiogram (ECG) signals, in accordance with an embodiment of the present disclosure.
[0040] In an embodiment, referring to FIG. 1, the system 102 for QRS complex detection, beat classification, and arrhythmia detection in electrocardiogram (ECG) signals may be connected to a network 104, which is further connected to at least one computing device 108-1, 108-2, … 108-N (collectively referred as computing device 108, herein) associated with one or more users 106-1, 106-2, … 106-N (collectively referred as user 106, herein). The system 102 may include at least one user equipment associated with the user 106 capable of display the at least one classified QRS complexes via a network 104.
[0041] In an embodiment, the system 102 may be configured to receive one or more ECG signals from at least one user 106.The system 102 may be configured to filter noise including baseline wander and high-frequency artifacts from the one or more ECG signals to improve signal quality. Further, the system 102 may be configured to segment the one or more ECG signals into a fixed-length interval of ECG segments and extract one or more relevant features. For instance, the system 102 may collect raw ECG signals segmented into 10-second intervals, sampled at 500 Hz, resulting in input sequences of 5000 samples per segment. The one or more features may be extracted from the one or more ECG signals by using a deep residual convolutional neural network (ResNet) along with a convolutional block attention module (CBAM). The CBAM module may inlcude at least one of a channel attention sub-module and a spatial attention sub-module configured to enhance one or more ECG signals.
[0042] In an embodiment, the system 102 may be configured to analyse the one or more extracted relevant features using on one or more predefined parameters to generate one or more probability masks for QRS complexes detection. The system 102 can analyse the extracted relevant features based on the predefined parameters, including duration, amplitude, and polarity, to improve the QRS complexes detection and beat classification. The system 102 may be configured to generate the one or more probability masks for one or more cardiac events which include, but not limited to: QRS complexes, normal beats, premature ventricular contractions (PVCs), and the like. Each of the one or more probability masks may represents the likelihood of a specific cardiac event occurring at each time point in the ECG signal. Further, the system 102 may enable alignment between the ECG signals and their corresponding masks for accurate training and evaluation. The one or more predefined parameters may include, but not limited to: a duration, an amplitude, a polarity, and the like. The one or more extracted relevant features may include, but not limited to: a temporal feature, a morphological feature, a spatial feature, and the like.
[0043] In an embodiment, the temporal feature pertains to the timing and duration of ECG events for analyzing the one or more ECG signals. The temporal feature may include, but not limited to: RR Interval, QRS Duration, PR Interval, and the like. The morphological feature pertains to one or more waveform characteristics of ECG components for classifying at least one of normal heartbeat and abnormal heartbeats. The morphological feature may include, but not limited to: R-peak amplitude, QRS slope, ST segment elevation, and the like. The spatial feature pertains to a three-dimensional view of the one or more ECG signals, wherein the spatial feature may include, but not limited to: electrical axis of the heart, multi-lead QRS amplitude variability, inter-lead correlation, and the like.
[0044] In an embodiment, the system 102 may be configured to identify one or more R-peaks within the QRS complexes and assign one or more corresponding confidence scores. The system 102 may be configured to apply a dynamic thresholding mechanism to detect the one or more R-peaks within the QRS complexes.
[0045] In an embodiment, the system 102 may be configured to classify the identified QRS complexes into one or more predefined heartbeat categories based on the one or more corresponding confidence scores using a neural network model. The one or more predefined heartbeat categories may include, but not limited to: normal beats, premature ventricular contractions (PVCs), bundle branch blocks (BBBs), and the like.
[0046] In an embodiment, the system 102 may include a Bidirectional Gated Recurrent Unit (Bi-GRU) layer is configured to interpret the one or more ECG signals by using at least one of: temporal dependencies and long-term dependencies.
[0047] In an embodiment, the system 102 may be configured to display the classified QRS complexes on at least one computing device 108 associated with the at least one user 106 for clinical interpretation.
[0048] In an embodiment, the system 102 may include the ResNet with the CBAM integrated with the Bi-GRU layer for enhanced QRS complex detection in the one or more ECG signals. For instance, feature extraction using the ResNet with the CBAM may include an initial convolutional layer which may extract the one or more ECG features using 1D convolution with batch normalization and LeakyReLU activation for improved stability. Further, one or more residual blocks with multiple convolutional layers (filters: 16, 32, 64, 128) may be used for hierarchical feature extraction, with skip connections to mitigate vanishing gradients. The CBAM enhances feature relevance through channel and spatial attention mechanisms, refining feature maps for better discrimination.
[0049] In an embodiment, the system 102 may include the temporal feature learning via the Bi-GRU layer may be configured to capture the long-term dependencies in a forward direction and a backward direction, improving the model’s ability to analyze ECG sequences dynamically. In an embodiment, the system 102 may be configured to perform upsampling and output generation. The upsampling layers may restore the original temporal dimensions after downsampling. The final Conv1D layer with sigmoid activation generates a three-channel output, predicting probability masks for QRS complexes, normal beats, and premature ventricular contractions (PVCs).
[0050] In an embodiment, the system 102 may include the neural network model training and optimization process involves compiling, training, and monitoring the deep learning model for QRS complex detection in the one or more ECG signals. The compilation may be performed using an Adam optimizer with a predefined learning rate (e.g., 0.001) and binary cross-entropy as the loss function for effective binary classification. Further, training begins by splitting the dataset into at least one of training sets and testing sets, reshaping input data (e.g., batch_size, 5000, 1), and using optimized hyperparameters such as batch size (16) and epochs (25). To enhance performance, early stopping is implemented with a patience parameter to prevent overfitting. Further, the system 102 may include one or more callbacks and monitoring mechanisms such as EarlyStopping track validation loss and stop training when improvements plateau, while the best model weights are saved for future inference.
[0051] In an exemplary embodiment, the computing device 108 may include, but not be limited to, a computer enabled device, a mobile phone, a smartphone, a tablet, a laptop, a display device, a surveillance camera, an automatic teller machine, and a point of sale, a kiosk, and a smart doorbell, a smart home device, a Augmented Reality/Virtual Reality/Mixed Reality (AR/VR/MR), an imaging device, a display projector, a Remote Detection Service (Detection Device) enabled devices such as iBeacon technologies, or some combination thereof. A person of ordinary skill in the art will understand that the at least one computing device 108 may be individually referred to as a computing device and collectively referred to as computing devices 108.
[0052] In an exemplary embodiment, the network 104 may include, but not be limited to, at least a portion of one or more networks having one or more nodes that transmit, receive, forward, generate, buffer, store, route, switch, process, or a combination thereof, etc. one or more messages, packets, signals, waves, voltage or current levels, some combination thereof, or so forth. In an exemplary embodiment, the network 104 may include, but not be limited to, a wireless network, a wired network, an internet, an intranet, a public network, a private network, a packet-switched network, a circuit-switched network, an ad hoc network, an infrastructure network, a Public-Switched Telephone Network (PSTN), a cable network, a cellular network, a satellite network, a fiber optic network, or some combination thereof.
[0053] In an embodiment, the system 102 may be used in real-time ECG monitoring, clinical diagnostics, wearable health devices, and remote patient monitoring applications to detect and classify QRS complexes efficiently. The system 102 may play a pivotal role in cardiac arrhythmia detection, automated ECG interpretation, and decision support for healthcare professionals by providing accurate QRS complex analysis and classification. The users may include, but not limited to: healthcare professionals, cardiologists, biomedical researchers, developers of AI-powered medical devices, and the like, who require reliable ECG signal processing solutions. The input data may include the one or more ECG signals collected from single-lead or multi-lead electrodes, along with annotated training datasets for deep learning model optimization and validation.
[0054] In an embodiment, the system 102 includes one or more processors (refer FIG. 2), and a memory (refer FIG. 2) coupled to the one or more processors, where said memory stores instructions which when executed by the one or more processors cause the system 102 to receive the input data from the computing device 108 associated with the users 106.
[0055] FIG. 2 illustrates an exemplary block diagram architecture of the system for providing real-time cognitive load assessment and dynamic learning management, in accordance with an embodiment of the present disclosure.
[0056] In an aspect, referring to FIG. 2, the system 102 may include one or more processor(s) 202. The one or more processor(s) 202 may be implemented as one or more microprocessors, microcomputers, microcontrollers, edge or fog microcontrollers, digital signal processors, central processing units, logic circuitries, and/or any devices that process data based on operational instructions. Among other capabilities, the one or more processor(s) 202 may be configured to fetch and execute computer-readable instructions stored in a memory 204 of the system 102. The memory 204 may be configured to store one or more computer-readable instructions or routines in a non-transitory computer readable storage medium, which may be fetched and executed to create or share data packets over a network service. The memory 204 may include any non-transitory storage device including, for example, volatile memory such as Random Access Memory (RAM), or non-volatile memory such as Erasable Programmable Read-Only Memory (EPROM), flash memory, and the like.
[0057] Referring to FIG. 2, the system 102 may include an interface(s) 206. The interface(s) 206 may include a variety of interfaces, for example, interfaces for data input and output devices, referred to as I/O devices, storage devices, and the like. The interface(s) 206 may facilitate communication to/from the system 102. The interface(s) 206 may also provide a communication pathway for one or more components of the system 102. Examples of such components include, but are not limited to, processing unit/engine(s) 208 and a local database 210.
[0058] In an embodiment, the processing unit/engine(s) 208 may be implemented as a combination of hardware and programming (for example, programmable instructions) to implement one or more functionalities of the processing engine(s) 208. In examples described herein, such combinations of hardware and programming may be implemented in several different ways. For example, the programming for the processing engine(s) 208 may be processor-executable instructions stored on a non-transitory machine-readable storage medium and the hardware for the processing engine(s) 208 may include a processing resource (for example, one or more processors), to execute such instructions. In the present examples, the machine-readable storage medium may store instructions that, when executed by the processing resource, implement the processing engine(s) 208. In such examples, the system 102 may include the machine-readable storage medium storing the instructions and the processing resource to execute the instructions, or the machine-readable storage medium may be separate but accessible to the system 102 and the processing resource. In other examples, the processing engine(s) 208 may be implemented by electronic circuitry.
[0059] In an embodiment, the local database 210 may include data that may be either stored or generated as a result of functionalities implemented by any of the components of the processor 202 or the processing engines 208. In an embodiment, the local database 210 may be separate from the system 102.
[0060] In an exemplary embodiment, the processing engine 208 may include one or more engines selected from any of a ECG acquisition module 212, a feature extraction module 214, a peak detection module 216, a classification module 218, and other modules 220 having functions that may include but are not limited to testing, storage, and peripheral functions, such as a wireless communication unit for remote operation, an audio unit for alerts, and the like.
[0061] In an embodiment, the system 102 may include the ECG acquisition module 212 which may be configured to collecting one or more raw ECG signals from single-lead or multi-lead electrodes placed on the user’s skin. The ECG acquisition module 212 may be configured to converts analog bioelectrical signals into digital format using an analog-to-digital converter (ADC) for further processing. The ECG acquisition module 212 may apply sampling techniques (e.g., 250–500 Hz) to ensure high-resolution ECG signal capture. Additionally, the ECG acquisition module 212 may perform initial noise filtering to remove baseline wander and power line interference.
[0062] In an embodiment, the system 102 may include the feature extraction module 214 can be configured to processes one or more raw ECG signals from the ECG acquisition module 212 to extract spatial, temporal and morphological features for QRS complex detection. The feature extraction module 214 may apply convolutional layers in a deep residual network to capture spatial features and bidirectional GRU layers to identify temporal dependencies. Key extracted features include RR intervals, QRS duration, R-peak amplitude, and QRS slope characteristics. Additionally, the feature extraction module 214 may incorporate the Convolutional Block Attention Modules (CBAM) to enhance feature relevance.
[0063] In an embodiment, the system 102 may include the peak detection module 216 which may be configured to identify one or more R-peaks within the at least one QRS complex using a dynamic thresholding mechanism. The peak detection module 216 may first enhance signal peaks using adaptive filtering and then applies a moving window-based thresholding technique to detect prominent R-peaks. The peak detection module 216 may continuously update threshold based on one or more recent ECG signal characteristics to adapt to heart rate variability. The peak detection module 216 may also integrates false peak rejection mechanisms to minimize detection errors in noisy or irregular ECG signals.
[0064] In an embodiment, the system 102 may include the classification module 218 which may be configured to categorize the at least one QRS complexes based on the one or more extracted features and the one or more detected R-peaks. The classification module 218 may employ a multi-class deep learning model trained on labeled ECG datasets to differentiate between normal QRS complexes, premature ventricular contractions (PVCs), and bundle branch blocks (BBBs). The classification model 218 may utilize a combination of convolutional neural networks (CNNs) and recurrent neural networks (RNNs) to improve accuracy. Additionally, the classification module 218 may assign a confidence score to each classification result, ensuring reliability in real-time cardiac event detection and diagnosis.
[0065] FIG. 3 illustrates a flow diagram 300 for implementing for QRS complex detection, beat classification, and arrhythmia detection in electrocardiogram (ECG) signals, in accordance with an embodiment of the present disclosure.
[0066] In an embodiment, at step 302, the method 300 may include receiving, by a system 102, one or more ECG signals from at least one user 106 and filtering noise including baseline wander and high-frequency artifacts from the one or more ECG signals to improve signal quality.
[0067] In an embodiment, at step 304, the method 300 may include segmenting, by the system 102, the one or more ECG signals, from step 302, into a fixed-length interval of ECG segments and extracting one or more relevant features.
[0068] In an embodiment, at step 306, the method 300 may include analysing, by the system 102, the one or more extracted relevant features, from step 304, using one or more predefined parameters for generating one or more probability masks for detecting QRS complexes.
[0069] In an embodiment, at step 308, the method 300 may include identifying, by the system 102, one or more R-peaks within the QRS complexes, from step 306, and assigning one or more corresponding confidence scores.
[0070] In an embodiment, at step 310, the method 300 may include classifying, by the system 102, the identified QRS complexes into one or more predefined heartbeat categories based on the one or more corresponding confidence scores, from step 308, using a neural network model.
[0071] In an embodiment, at step 312, the method 300 may include displaying, by the system 102, the classified QRS complexes, from step 310, on at least one computing device 108 associated with the at least one user 106 for clinical interpretation.
[0072] FIG. 4 illustrates an exemplary computer system 500 in which or with which embodiments of the present invention can be utilized in accordance with embodiments of the present disclosure.
[0073] As shown in FIG. 4, the computer system 400 may include an external storage device 410, a bus 420, a main memory 430, a read only memory 440, a mass storage device 450, a communication port 460, and a processor 470. A person skilled in the art will appreciate that the computer system 400 may include more than one processor and communication ports. The processor 470 may include various modules associated with embodiments of the present invention. The communication port 460 may be any of an RS-232 port for use with a modem-based dialup connection, a 10/100 Ethernet port, a Gigabit or 10 Gigabit port using copper or fiber, a serial port, a parallel port, or other existing or future ports. The communication port 460 may be chosen depending on a network, such a Local Area Network (LAN), Wide Area Network (WAN), or any network to which computer system connects. The memory 430 may be a Random Access Memory (RAM), or any other dynamic storage device commonly known in the art. The read-only memory 440 may be any static storage device(s) e.g., but not limited to, a Programmable Read Only Memory (PROM) chips for storing static information e.g., start-up or BIOS instructions for the processor 470. The mass storage 450 may be any current or future mass storage solution, which may be used to store information and/or instructions. Exemplary mass storage solutions include, but are not limited to, Parallel Advanced Technology Attachment (PATA) or Serial Advanced Technology Attachment (SATA) hard disk drives or solid-state drives (internal or external, e.g., having Universal Serial Bus (USB) and/or Firewire interfaces), one or more optical discs, Redundant Array of Independent Disks (RAID) storage, e.g. an array of disks (e.g., SATA arrays).
[0074] The bus 420 may communicatively couple the processor(s) 470 with the other memory, storage and communication blocks. The bus 420 may be, e.g. a Peripheral Component Interconnect (PCI)/PCI Extended (PCI-X) bus, Small Computer System Interface (SCSI), USB or the like, for connecting expansion cards, drives and other subsystems as well as other buses, such a front side bus (FSB), which connects processor 470 to a software system.
[0075] Optionally, operator and administrative interfaces, e.g. a display, keyboard, and a cursor control device, may also be coupled to the bus 420 to support direct operator interaction with the computer system 400. Other operator and administrative interfaces may be provided through network connections connected through the communication port 460. The external storage device 410 may be any kind of external hard-drives, floppy drives, Compact Disc - Read Only Memory (CD-ROM), Compact Disc-Re-Writable (CD-RW), Digital Video Disk-Read Only Memory (DVD-ROM). Components described above are meant only to exemplify various possibilities. In no way should the aforementioned exemplary computer system limit the scope of the present disclosure.
[0076] While considerable emphasis has been placed herein on the preferred embodiments, it will be appreciated that many embodiments can be made and that many changes can be made in the preferred embodiments without departing from the principles of the invention. These and other changes in the preferred embodiments of the invention will be apparent to those skilled in the art from the disclosure herein, whereby it is to be distinctly understood that the foregoing descriptive matter to be implemented merely as illustrative of the invention and not as limitation.

ADVANTAGES OF THE PRESENT DISCLOSURE
[0077] The present disclosure provides a system and a method for QRS complex detection, beat classification, and arrhythmia detection in electrocardiogram (ECG) signals.
[0078] The present disclosure provides a non-invasive, cost-effective, reliable and an enhanced QRS detection accuracy which improves the precision of QRS complex detection, even in noisy or low-amplitude ECG signals.
[0079] The present disclosure employs an adaptive thresholding mechanism which adjusts based on real-time ECG variations, reducing false positives and false negatives in R-peak detection.
[0080] The present disclosure provides a robust, real-time solution which supports multi-lead ECG analysis, leveraging spatial features to improve detection accuracy and providing a comprehensive view of cardiac activity.
[0081] The present disclosure detects and classifies QRS complexes into normal and abnormal types in real-time cardiac monitoring and diagnostics environments.
[0082] The present disclosure provides a low-latency processing, enabling real-time ECG monitoring in wearable health devices, clinical settings, and ambulatory ECG systems.
, Claims:1. A system (102) for QRS complex detection, beat classification, and arrhythmia detection in electrocardiogram (ECG) signals, the system (102) comprising:
one or more processors (202); and
a memory (204) coupled to the one or more processors (202), wherein the memory (204) storing instructions executable by the one or more processors (202) cause the system (102) to:
receive one or more ECG signals from at least one user and filter noise to remove baseline wander and high-frequency artifacts from the one or more ECG signals;
divide the one or more ECG signals into fixed-length intervals of ECG segments and extract one or more relevant features;
analyse the one or more extracted relevant features based on one or more predefined parameters to generate one or more probability masks for detecting QRS complexes;
identify one or more R-peaks within the QRS complexes and assign one or more corresponding confidence scores;
classify the identified QRS complexes into one or more predefined heartbeat categories based on the one or more corresponding confidence scores using a neural network model; and
display the classified QRS complexes on at least one computing device 108 associated with the at least one user 106 for clinical interpretation.

2. The system (102) as claimed in claim 1, wherein the system (102) is configured to apply a dynamic thresholding mechanism to detect the one or more R-peaks within at least one QRS complex.

3. The system (102) as claimed in claim 1, wherein the system 102 is configured to analyze the one or more extracted relevant features based on the one or more predefined parameters to improve the QRS complexes detection and beat classification,
wherein the one or more predefined parameters comprises at least one of a duration, an amplitude, and a polarity.

4. The system (102) as claimed in claim 1, wherein the one or more relevant features are extracted from the one or more ECG signals by using a deep residual convolutional neural network (ResNet) along with a convolutional block attention module (CBAM).

5. The system (102) as claimed in claim 4, wherein the CBAM module comprises at least one of a channel attention sub-module and a spatial attention sub-module configured to enhance one or more ECG signals.

6. The system (102) as claimed in claim 1, wherein the system (102) comprises:
a Bidirectional Gated Recurrent Unit (Bi-GRU) layer is configured to interpret the one or more ECG signals by using at least one of: temporal dependencies and long-term dependencies.

7. The system (102) as claimed in claim 1, wherein the one or more predefined heartbeat categories comprises at least one of: normal beats, premature ventricular contractions (PVCs), and bundle branch blocks (BBBs).

8. The system (102) as claimed in claim 1, wherein the one or more extracted relevant features comprises at least one of a temporal feature, a morphological feature, and a spatial feature.

9. The system (102) as claimed in claim 8, wherein the temporal feature pertains to timing and duration of ECG events for analyzing the one or more ECG signals, wherein the temporal feature comprising at least one of RR Interval, QRS Duration, and PR Interval,
wherein the morphological feature pertains to one or more waveform characteristics of ECG components for classifying at least one of normal heart beat and abnormal heart beats, wherein the morphological feature comprising at least one of R-peak Amplitude, QRS Slope, and ST Segment Elevation,
wherein the spatial feature pertains to a three-dimensional view of the one or more ECG signals, wherein the spatial feature comprising at least one of Electrical Axis of the Heart, Multi-Lead QRS Amplitude Variability and Inter-Lead Correlation.

10. A method (300) for QRS complex detection, beat classification, and arrhythmia detection in electrocardiogram (ECG) signals, the method comprising the steps of:
receiving, by a system (102), one or more ECG signals from at least one user and filtering noise to remove baseline wander and high-frequency artifacts from the one or more ECG signals to improve signal quality;
segmenting, by the system (102), the one or more ECG signals into fixed-length intervals of ECG segments and extracting one or more relevant features;
analysing, by the system (102), the one or more extracted relevant features using one or more predefined parameters for generating one or more probability masks for detecting QRS complexes;
identifying, by the system (102), one or more R-peaks within the QRS complexes and assigning one or more corresponding confidence scores;
classifying, by the system (102), the identified QRS complexes into one or more predefined heartbeat categories based on the one or more corresponding confidence scores using a neural network model; and
displaying, by the system (102), the classified QRS complexes on at least one computing device (108) associated with the at least one user (106) for clinical interpretation.