Description:FIELD OF THE INVENTION
[0001] The present invention relates to the field of biomedical signal processing and healthcare technology. More particularly, the present invention relates to a spectrum weighting system for electrocardiogram-based stress identification.

DESCRIPTION OF THE RELATED ART
[0002] The following description of the related art is intended to provide background information pertaining to the field of 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] Stress disorder identification remains a critical healthcare challenge globally. This is particularly true in developing countries and rural communities. Traditional stress assessment methods predominantly rely on subjective questionnaires and clinical observations. These conventional approaches pose significant accessibility barriers and social stigma. They also lead to delayed diagnosis and treatment of stress-related conditions. The dependence on specialized psychiatric expertise limits healthcare access in underserved areas. This creates disparities in mental health service delivery. Therefore, there is a need to develop objective, automated alternatives for accessible stress disorder identification.
[0004] Existing electrocardiogram-based stress detection systems operate with fixed feature extraction methods. They do not consider the varying of different frequency components that significantly influence stress pattern recognition. These conventional systems fail to adapt their analysis to discriminate between stress-indicative and non-stress frequency patterns. This results in poor sensitivity and inconsistent classification accuracy across different patient populations.
[0005] Recent developments in machine learning-based biomedical signal processing have shown promise. Research demonstrates the potential of deep learning to extract meaningful patterns from physiological signals. However, current implementations have several limitations. They utilize uniform frequency feature treatment without considering the relative of spectral components. They employ standard convolutional architectures without optimization for stress-specific frequency patterns. These systems also lack integration with learnable weights for enhancing discriminative capabilities. The absence of spectrum weighting modules limits their accuracy from 72.43% baseline performance. Additionally, inadequate frequency component optimization prevents achieving the precision required for reliable clinical stress screening.
[0006] Therefore, there exists a requirement for an improved spectrum weighting system for electrocardiogram-based stress identification. This system should incorporate frequency analysis modules including signal conversion circuitry for extracting frequency domain features. These modules enable comprehensive spectral analysis of ECG signals. The system should implement weight optimization modules with gradient-based learning. These modules dynamically compute learnable weights to discriminate stress patterns from non-stress patterns. The system should utilize weighted feature generation units with multiplier circuits for combining frequency features with weights. It should integrate CNN processing modules or any other classifier for hierarchical feature extraction. The system should also employ control units for coordinating sequential signal flow from ECG acquisition through classification. This comprehensive approach provides an objective, accurate, and scalable solution for stress identification achieving 91.2% accuracy across diverse clinical settings.

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] An object of the present disclosure is to provide a spectrum weighting system for electrocardiogram-based stress identification and method thereof which employs at least one user equipment configured with electrocardiogram electrodes in physical contact with a user, enabling continuous acquisition of physiological signals while maintaining proper electrode-skin interface for reliable signal quality.
[0009] An object of the present disclosure is to provide a spectrum weighting system for electrocardiogram-based stress identification and method thereof which utilizes a frequency analysis module including signal conversion circuitry electrically connected to the user equipment to extract frequency domain features from acquired electrocardiogram signals, thereby enabling comprehensive spectral analysis of cardiac electrical activity.
[0010] An object of the present disclosure is to provide a spectrum weighting system for electrocardiogram-based stress identification and method thereof which implements a weight optimization module where frequency domain features undergo gradient-based optimization to compute learnable weights that discriminate between stress and non-stress patterns in the electrocardiogram frequency spectrum.
[0011] An object of the present disclosure is to provide a spectrum weighting system for electrocardiogram-based stress identification and method thereof which incorporates a weighted feature generation unit including multiplier circuits to generate weighted frequency features by multiplying frequency domain features with corresponding learnable weights dynamically adjusted based on stress pattern characteristics.
[0012] An object of the present disclosure is to provide a spectrum weighting system for electrocardiogram-based stress identification and method thereof which operates as an integrated processing pipeline by continuously acquiring electrocardiogram signals, extracting frequency features, computing weights, and generating weighted features for stress classification, thereby forming a complete operational flow from signal acquisition to stress identification.
[0013] An object of the present disclosure is to provide a spectrum weighting system for electrocardiogram-based stress identification and method thereof which integrates a CNN processing module or any other classifier for processing weighted frequency features. In the present implementation, CNN uses a first convolutional layer with 128 filters and second convolutional layer with 64 filters.
[0014] An object of the present disclosure is to provide a spectrum weighting system for electrocardiogram-based stress identification and method thereof which incorporates a control unit that coordinates signal flow through sequential processing stages from electrocardiogram acquisition through frequency analysis, weighting, and classification, thereby ensuring synchronized operation for real-time stress monitoring in clinical settings.

SUMMARY
[0015] This section 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 a spectrum weighting system for electrocardiogram-based stress identification. The system uses weight optimization modules to compute learnable weights in real-time based on frequency domain features extracted from ECG signals. Weighted feature generation units produce optimized frequency features with dynamic adjustment. Classification modules enable hierarchical stress pattern recognition.
[0017] The system includes user equipment with electrocardiogram electrodes in physical contact with users connected to a frequency analysis module. A control unit executes signal processing through: frequency domain feature extraction, weight computation, and weighted feature generation. These modules generate stress-discriminative features based on ECG spectral analysis. A CNN processing module or any classifier performs hierarchical feature extraction. Real-time signal flow coordination ensures continuous stress monitoring. The control unit enables clinical deployment with enhanced accuracy.
[0018] The method includes acquiring electrocardiogram signals through electrode-equipped user equipment and extracting frequency domain features. Weight optimization modules compute learnable weights through gradient-based optimization to discriminate stress patterns. The weighted feature generation unit multiplies frequency features with weights. Classification through CNN or any other classifiers generates stress classification output. The process operates continuously for real-time stress identification in healthcare settings.
[0019] Various objects, features, aspects and advantages of the inventive subject matter will become more apparent from the following detailed description of preferred embodiments, along with the accompanying drawing figures in which like numerals represent like components.

BRIEF DESCRIPTION OF DRAWINGS
[0020] The accompanying drawings are included to provide a further understanding of the present disclosure and are incorporated in and constitute a part of this specification. The drawings illustrate exemplary embodiments of the present disclosure and, together with the description, serve to explain the principles of the present disclosure. The diagrams are for illustration only, which thus is not a limitation of the present disclosure.
[0021] FIG. 1 illustrates an exemplary representation of a network environment implementing a spectrum weighting system for electrocardiogram-based stress identification, in accordance with an embodiment of the present disclosure.
[0022] FIG. 2 illustrates an exemplary block diagram of the spectrum weighting system depicting input components, control unit, and output components, in accordance with an embodiment of the present disclosure.
[0023] FIG. 3 illustrates an exemplary representation of the control unit architecture including processing engine modules, in accordance with an embodiment of the present disclosure.
[0024] FIG. 4 illustrates an exemplary flow diagram depicting a method for electrocardiogram-based stress identification using spectrum weighting, in accordance with an embodiment of the present disclosure.

DETAILED DESCRIPTION
[0025] 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.
Definitions:
Electrocardiogram (ECG): A physiological signal representing the electrical activity of the heart captured through electrodes placed on the skin surface, containing frequency domain information indicative of autonomic nervous system responses to stress conditions.
Learnable weights: Adaptive numerical coefficients computed through gradient-based optimization modules that quantify the relative significance of each frequency component in the ECG spectrum for discriminating between stress and non-stress patterns.
Spectrum Weighting: A signal processing methodology that applies differential to frequency components through element-wise multiplication of frequency domain features with optimized weight vectors to enhance stress pattern recognition.
Module: As used herein, the term "module" refers to hardware components including processing circuitry, associated memory, and interface circuits that implement specific functionalities within the system architecture.
[0026] An aspect of the present disclosure relates to a spectrum weighting system for electrocardiogram-based stress identification including at least one user equipment with ECG electrodes for signal acquisition. The system includes at least one frequency analysis module extracting frequency domain features. The system includes at least one weight optimization module computing learnable weights. The system includes at least one weighted feature generation unit producing optimized features. The system includes classification modules implementing hierarchical pattern recognition. The system includes control units coordinating signal flow. The system includes multiplier circuits performing element-wise operations. The system includes parallel processing circuits enabling real-time classification. The system includes clinical deployment capabilities achieving enhanced accuracy for stress monitoring across diverse healthcare settings.
[0027] Various embodiments of the present disclosure are described using FIGs. 1 to 4.
[0028] FIG. 1 illustrates an exemplary representation of a network environment (100) implementing a spectrum weighting system for electrocardiogram-based stress identification, in accordance with an embodiment of the present disclosure.
[0029] In an embodiment, referring to FIG. 1, the network architecture (100) can include multiple user equipment devices (108-1, 108-2, 108-N) representing ECG monitoring systems which may be configured to connect to a network (104), which is further connected to a system (102), a centralized server (110), and a healthcare dashboard (112). In an implementation, each user equipment (108-1, 108-2, 108-N) equipped with ECG electrodes can be configured to acquire a plurality of physiological signals from users (106-1, 106-2, 106-N) for stress identification. The plurality of physiological signals can include electrocardiogram waveforms containing stress-indicative frequency patterns. Healthcare providers can monitor and analyze stress levels through the network architecture.
[0030] In an exemplary embodiment, the users (106-1, 106-2, 106-N) may be patients or individuals undergoing stress assessment through ECG monitoring. These users may include, but not limited to, individuals with stress disorders, regular meditators participating in wellness programs, patients in clinical settings, employees in high-stress occupations, and participants in mental health screening programs. The centralized server (110) may aggregate data from multiple monitoring systems and provide population-level stress analytics.
[0031] 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, or data streams. 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 clinical information system network, a telemedicine network, or some combination thereof.
[0032] In another exemplary embodiment, the system (102) may include or include, by way of example but not limitation, one or more of: a stand-alone server, a server blade, a server rack, a bank of servers, a server farm, hardware supporting a part of a cloud service or system, hardware running a virtualized server, one or more processors executing code to function as a server, one or more machines performing server-side functionality as described herein. In an embodiment, the control unit (212) of each user equipment may be coupled to the system (102) through communication modules. In another embodiment, the system (102) may also be operatively coupled to the healthcare dashboard (112) for stress pattern visualization and clinical decision support.
[0033] In an embodiment, the spectrum weighting system (102) can include at least one user equipment (108-1) including electrocardiogram electrodes in physical contact with a user to acquire electrocardiogram signals; a frequency analysis module (314) including signal conversion circuitry electrically connected to the at least one user equipment (108-1), the frequency analysis module (314) configured to receive the acquired electrocardiogram signals from the at least one user equipment (108-1) and extract frequency domain features from the acquired electrocardiogram signals; an weight optimization module (316) electrically connected to the frequency analysis module (314), the weight optimization module (316) configured to receive the frequency domain features from the frequency analysis module (314) and compute learnable weights for frequency components to discriminate between stress and non-stress patterns in the electrocardiogram signals; a weighted feature generation unit (204-5) including multiplier circuits electrically connected to receive the frequency domain features from the frequency analysis module (314) and the learnable weights from the weight optimization module (316), the weighted feature generation unit (204-5) configured to generate weighted frequency features by multiplying the frequency domain features with corresponding learnable weights; a CNN processing module (318) including parallel processing circuits electrically connected to the weighted feature generation unit (204-5), the CNN processing module (318) configured to receive the weighted frequency features from the weighted feature generation unit (204-5) and process the weighted frequency features to generate stress classification output; a control unit (212) electrically connected to the frequency analysis module (314), the weight optimization module (316), the weighted feature generation unit (204-5), and the CNN processing module (318), the control unit (212) configured to coordinate signal flow from the at least one user equipment (108-1) through each connected module sequentially to transform the physiological electrocardiogram signals into the stress classification output, enabling real-time stress monitoring in clinical settings.
[0034] FIG. 2 illustrates an exemplary block diagram of the spectrum weighting system (200) depicting input components, control unit, and output components, in accordance with an embodiment of the present disclosure.
[0035] In an aspect, referring to FIG. 2, the system (202) may include one or more control unit(s) (212). The one or more control unit(s) (212) may be implemented as one or more microprocessors, microcomputers, microcontrollers, digital signal processors, central processing units, logic circuitries, and/or any devices that process data based on operational instructions. Among other capabilities, one or more control unit(s) (212) may be configured to fetch and execute computer-readable instructions stored in the memory (304) of the system (202). The memory (304) 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 process ECG signals and compute weights. The memory (304) 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 (EEPROM), flash memory, and the like.
[0036] Referring to FIG. 2, the system (202) may include various interfaces for component connectivity. The interfaces may include a variety of connections, for example, interfaces for ECG signal input from the ECG sensor (204-1), PPG sensor (204-2), and respiratory sensor (204-3), power input (210), output connections to the weighted layer processing unit (204-5) and CNN classifier module (204-6), and communication pathways to the wireless network module (104). The interfaces may facilitate data flow to/from the control unit (212). The interfaces may also provide communication pathways for multiple components of the system (202). Examples of such components include but are not limited to, the processing unit within the control unit (212) and the memory (304).
[0037] In an embodiment, the control unit (212) may be implemented as a combination of hardware and programming (for example, programmable instructions) to implement one or more functionalities of the spectrum weighting modules. In the examples described herein, such combinations of hardware and programming may be implemented in several different ways. For example, the programming for the control unit (212) may be processor-executable instructions stored on a non-transitory machine-readable storage medium and the hardware for the control unit (212) may include a processing resource (for example, the processor (302)), 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 control unit (212). In such examples, the system (202) may include the machine-readable storage medium storing the instructions and the processing resource to execute the instructions. In other examples, the control unit (212) may be implemented by electronic circuitry.
[0038] In an embodiment, the memory (304) may include data that may be either stored or generated as a result of functionalities implemented by any of the components of the processing engine (308) or the control unit (212). The stored data may include frequency domain features, weight vectors, historical ECG recordings, stress classification models, and clinical thresholds. In an embodiment, the memory (304) may be integrated within the control unit (212).
[0039] In an exemplary embodiment, the control unit (212) may include one or more modules selected from any of a receiving module (312), an FFT analysis module (314), a weighting module (316), a CNN processing module (318), and other modules (320) having functions that may include but are not limited to signal preprocessing, artifact removal, baseline correction, and classification result
[0040] FIG. 3 illustrates an exemplary representation of the control unit architecture (300) including processing engine modules, in accordance with an embodiment of the present disclosure.
[0041] In an embodiment, referring to FIG. 3, the control unit (212) can include the processor(s) (302) with memory (304) executing multiple processing modules within the processing engine (308). The receiving module (312) can acquire ECG signals from connected sensors and perform initial signal quality assessment. The FFT analysis module (314) can transform time-domain ECG signals into frequency domain representations extracting spectral features. The weighting module (316) can compute and apply learnable weights through iterative optimization. The CNN processing module (318) can implement hierarchical feature extraction through convolutional layers. The other module(s) (320) can manage auxiliary functions including data logging and system diagnostics. The interface(s) (306) can route signals between input sensors and output components.
[0042] In an embodiment, the frequency analysis module (314) integrated within the control unit (212) may implement Fast Fourier Transform modules converting time-domain ECG signals into frequency domain representations. A person of ordinary skill in the art will understand that the FFT computation may utilize radix-2 or radix-4 modules for efficient spectral analysis.
[0043] In an embodiment, the system (202) may segment the acquired electrocardiogram signals into three-minute intervals before frequency domain analysis. The segmentation strategy may capture stress-indicative temporal patterns while providing sufficient spectral resolution. The three-minute window may balance computational efficiency with physiological relevance, enabling detection of both acute stress responses and sustained stress states. The overlapping window approach may ensure continuous monitoring without missing transitional stress events.
[0044] In an embodiment, the weight optimization module (316) may implement gradient-based learning modules including stochastic gradient descent or adaptive moment estimation. The module may iteratively adjust the learnable weights through backpropagation of classification errors. The optimization process may maximize discrimination between stress-indicative frequency components and non-stress frequency components by amplifying discriminative spectral features while suppressing noise. The convergence criteria may ensure stable weight values achieving optimal stress pattern separation.
[0045] In an embodiment, the weighted feature generation unit (204-5) can perform element-wise multiplication between frequency domain features and learnable weights. The multiplication operation can be expressed mathematically as x'_i = w_i × x_i where x_i represents the i-th frequency component, w_i represents the corresponding weight, and x'_i represents the weighted feature. The selective amplification can improve signal-to-noise ratio for stress-relevant spectral information.
[0046] In an embodiment, the CNN processing module (318) can implement a multi-layer convolutional architecture for hierarchical feature extraction from weighted frequency features. The first convolutional layer having 128 filters can extract primary stress-related features capturing local spectral patterns. The second convolutional layer having 64 filters can extract hierarchical stress patterns combining primary features into complex representations. The convolutional operations can utilize kernel size of 4 with ReLU activation functions introducing non-linearity for pattern discrimination.
[0047] In an embodiment, each convolutional layer can be followed by max-pooling operations reducing spatial dimensions while preserving salient features. The pooling layers can implement subsampling factor of 2 progressively condensing feature representations. The batch normalization layers can stabilize training by normalizing intermediate activations. The dropout layers can prevent overfitting by randomly deactivating neurons during training. The architectural choices can balance model capacity with generalization performance.
[0048] In an embodiment, the fully connected layer following the convolutional stages can integrate extracted features into unified representations for final classification. The layer can contain 10 neurons providing sufficient capacity for complex feature combinations while avoiding excessive parameterization. The dense connections can enable global feature interactions capturing holistic stress patterns. The final output layer can implement softmax activation producing probability distributions over stress and non-stress classes. Alternatively, a sigmoid activation function with one output node also may be used instead of the softmax activation.
[0049] In an embodiment, the control unit (212) can update the learnable weights based on classification performance feedback. The adaptive learning mechanism can enable personalization to user-specific stress patterns over time. The update process can utilize exponential moving averages preventing abrupt weight changes while allowing gradual adaptation. The personalization can improve long-term monitoring accuracy by accounting for individual physiological variations in stress responses.
[0050] In an embodiment, the weight optimization module (316) can include a optimization module to dynamically adjust the weights associated with each spectral feature to enhance the stress identification accuracy of the CNN processing module (318). The dynamic weight adjustment is done as an optimal solution to finding the minimum of the loss function using gradient module. The closed-loop architecture can implement continuous learning paradigms refining weights based on accumulated evidence.
[0051] In an embodiment, the system can achieve stress identification accuracy improvement from 72.43% to 91.2% through the adaptive weighting of frequency components. The baseline accuracy of 72.43% can represent performance without frequency weighting using standard CNN architectures. The enhanced accuracy of 91.2% can demonstrate the effectiveness of learnable weights in amplifying discriminative spectral features. The 18.77% absolute improvement can validate the spectrum weighting approach for clinical deployment.
[0052] In an embodiment, the control unit (212) can identify stress patterns specific to rural healthcare environments by adapting the learnable weights based on population-specific electrocardiogram characteristics. The adaptation can account for demographic factors including age distribution, occupational stress sources, and lifestyle patterns prevalent in rural communities. The population-specific optimization can enhance screening effectiveness in resource-limited settings where access to specialized mental health services remains constrained.
[0053] In an embodiment, the system can support integration with existing clinical ECG equipment through standardized interfaces. The compatibility can enable deployment without replacing established medical infrastructure. The integration can utilize standard ECG lead configurations including 3-lead or 12-lead systems. The signal acquisition can maintain clinical-grade quality with sampling rates of 200 Hz or higher ensuring adequate temporal resolution for frequency analysis.
[0054] In an embodiment, the power input (210) can accept standard medical-grade power supplies ensuring electrical safety compliance. The isolation transformers can prevent ground loops protecting both patients and equipment. The electromagnetic compatibility can meet medical device standards preventing interference with other clinical instruments. The power management can implement low-noise designs minimizing artifacts in sensitive ECG measurements.
[0055] In an embodiment, the status indicator (216) can provide visual feedback regarding system operational states. The indicators can display signal quality metrics alerting operators to electrode connection issues. The classification confidence levels can be visualized enabling clinical interpretation of results. The real-time feedback can guide proper system usage ensuring reliable stress assessments.
[0056] In an embodiment, the communication module (204-7) can implement secure data transmission protocols protecting patient privacy. The encryption standards can comply with healthcare regulations including HIPAA requirements. The data anonymization can enable population studies while preserving individual confidentiality. The secure communication can support telemedicine applications extending specialist consultations to remote locations.
[0057] In an embodiment, the database (310) can store historical ECG recordings enabling longitudinal stress pattern analysis. The temporal tracking can identify stress trend changes supporting early intervention strategies. The data retention policies can balance storage requirements with clinical value. The indexed storage can enable rapid retrieval for comparative analyses.
[0058] In an embodiment, the system can implement artifact detection modules identifying and excluding corrupted signal segments. The motion artifacts from physical movement can be detected through baseline wandering patterns. The electrical interference can be identified through characteristic frequency signatures. The automatic quality control can ensure only clean signals contribute to stress classification maintaining diagnostic reliability.
[0059] In an embodiment, the healthcare dashboard (112) can provide comprehensive visualization interfaces for clinical users. The displays can show real-time ECG waveforms with overlaid stress indicators. The frequency spectrum visualizations can highlight weighted components enabling understanding of classification decisions. The historical trend graphs can track stress levels over days, weeks, or months supporting treatment monitoring. The population analytics can identify community-level stress patterns informing public health interventions.
[0060] In an embodiment, the system can support batch processing modes analyzing multiple ECG recordings simultaneously. The parallel processing can enable efficient screening programs processing hundreds of participants. The automated report generation can produce standardized clinical documentation. The batch analytics can identify high-risk individuals requiring immediate intervention.
[0061] In an embodiment, the system can integrate with electronic health record systems through standard interfaces. The integration can automatically populate patient charts with stress assessment results. The interoperability can support care coordination between primary care and mental health providers. The seamless data flow can improve clinical workflow efficiency.
[0062] FIG. 4 illustrates an exemplary flow diagram depicting a method (400) for electrocardiogram-based stress identification using spectrum weighting, in accordance with an embodiment of the present disclosure.
[0063] As illustrated, method (400) includes, at block (402), acquiring electrocardiogram signals from a user through at least one user equipment (108-1) including electrocardiogram electrodes in physical contact with the user. The acquisition can involve proper electrode placement following standard ECG protocols. The skin preparation can ensure low-impedance connections maximizing signal quality. The electrode positioning can capture cardiac electrical activity from multiple vectors. The signal acquisition can continue for sufficient duration capturing stress-related variations.
[0064] Continuing further, method (400) includes, at block (404), extracting frequency domain features from the acquired electrocardiogram signals using a frequency analysis module (314) including signal conversion circuitry electrically connected to the at least one user equipment (108-1). The extraction can apply windowing functions preventing spectral leakage. The FFT computation can generate frequency bins with appropriate resolution for heart rate variability analysis. The power spectral density can be calculated identifying energy distribution across frequencies. The feature vector can encompass relevant frequency ranges for stress assessment.
[0065] Continuing further, method (400) includes, at block (406), computing learnable weights for frequency components using an weight optimization module (316) electrically connected to the frequency analysis module (314), where the learnable weights discriminate between stress and non-stress patterns in the electrocardiogram signals. The computation can initialize weights randomly or using prior knowledge. The optimization iterations can adjust weights minimizing classification loss. The gradient calculations can determine weight update directions. The learning rate can control convergence speed balancing stability with adaptation rate.
[0066] Continuing further, method (400) includes, at block (408), generating weighted frequency features using a weighted feature generation unit (204-5) including multiplier circuits, where the weighted feature generation unit (204-5) receives the frequency domain features from the frequency analysis module (314) and the learnable weights from the weight optimization module (316), and multiplies the frequency domain features with corresponding learnable weights. The multiplication can amplify discriminative frequencies while attenuating irrelevant components. The weighted features can enhance stress-specific spectral signatures. The feature scaling can normalize magnitudes preventing numerical instabilities.
[0067] Continuing further, method (400) includes, at block (410), processing the weighted frequency features using a CNN processing module (318) including parallel processing circuits electrically connected to the weighted feature generation unit (204-5) to generate stress classification output. The processing can apply learned convolutional filters detecting stress patterns. The hierarchical feature extraction can build complex representations from simple patterns. The non-linear transformations can capture intricate stress manifestations. The classification layer can produce confidence scores for clinical interpretation.
[0068] Continuing further, method (400) includes, at block (412), coordinating signal flow through a control unit (212) electrically connected to the frequency analysis module (314), the weight optimization module (316), the weighted feature generation unit (204-5), and the CNN processing module (318), where the control unit (212) transforms the physiological electrocardiogram signals into the stress classification output through sequential processing, enabling real-time stress monitoring in clinical settings. The coordination can synchronize module operations maintaining processing pipeline efficiency. The data buffering can handle varying processing speeds. The result aggregation can combine intermediate outputs. The final classification can be transmitted to clinical interfaces.
[0069] In an embodiment, throughout method (400), the system can continuously monitor signal quality metrics ensuring reliable operation. The quality indicators can detect electrode disconnections triggering alerts. The adaptive thresholds can accommodate individual physiological variations. The continuous operation can provide uninterrupted stress monitoring supporting clinical observations. The real-time processing can enable immediate intervention when acute stress is detected.
[0070] In an embodiment, the method can be extended to support multi-modal physiological monitoring incorporating additional sensors. The PPG sensor (204-2) can provide pulse rate variability complementing ECG analysis. The respiratory sensor (204-3) can detect breathing patterns associated with stress states. The sensor fusion can improve classification robustness through complementary information. The multi-modal approach can enhance diagnostic confidence in challenging cases.
[0071] In an embodiment, the method can implement calibration procedures personalizing system performance for individual users. The baseline recordings can establish normal stress levels during relaxed states. The stress induction protocols can capture elevated stress responses. The calibration can optimize weights for individual physiological characteristics. The personalization can improve long-term monitoring accuracy accounting for inter-individual variations.
[0072] The described disclosure presents an advanced spectrum weighting system for electrocardiogram-based stress identification offering multiple innovations advancing mental health screening technology. The learnable weights can optimize frequency component contributions responding to complex stress patterns. The deep learning-based processing can extract hierarchical features surpassing traditional analysis methods. The adaptive optimization can personalize detection modules improving individual monitoring accuracy. The objective ECG-based approach can eliminate subjective assessment limitations. The real-time processing can enable immediate clinical interventions. The rural healthcare focus can address underserved population needs. These combined innovations can transform stress screening practices, providing accessible, accurate, and scalable mental health assessment supporting healthcare objectives in diverse global settings.
[0073] 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 disclosure. These and other changes in the preferred embodiments of the disclosure 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 disclosure and not as limitation.
[0074] If the specification states a component or feature “may”, “can”, “could”, or “might” be included or have a characteristic, that particular component or feature is not required to be included or have the characteristic.
[0075] As used in the description herein and throughout the claims that follow, the meaning of “a,” “an,” and “the” includes plural reference unless the context clearly dictates otherwise. Also, as used in the description herein, the meaning of “in” includes “in” and “on” unless the context clearly dictates otherwise.
[0076] Moreover, in interpreting the specification, all terms should be interpreted in the broadest possible manner consistent with the context. In particular, the terms “comprises” and “comprising” should be interpreted as referring to elements, components, or steps in a non-exclusive manner, indicating that the referenced elements, components, or steps may be present, or utilized, or combined with other elements, components, or steps that are not expressly referenced. Where the specification claims refer to at least one of something selected from the group consisting of A, B, C ….and N, the text should be interpreted as requiring only one element from the group, not A plus N, or B plus N, etc.
[0077] While the foregoing describes various embodiments of the proposed disclosure, other and further embodiments of the proposed disclosure may be devised without departing from the basic scope thereof. The scope of the proposed disclosure is determined by the claims that follow. The proposed disclosure is not limited to the described embodiments, versions or examples, which are included to enable a person having ordinary skill in the art to make and use the invention when combined with information and knowledge available to the person having ordinary skill in the art.

ADVANTAGES OF THE PRESENT DISCLOSURE
[0078] The present disclosure provides a spectrum weighting system for electrocardiogram-based stress identification that achieves stress identification accuracy improvement from 72.43% to 91.2% through adaptive weighting of frequency components that discriminate between stress and non-stress patterns in ECG signals.
[0079] The present disclosure provides a spectrum weighting system that enables objective stress identification using ECG signals processed through learnable weights, eliminating dependence on subjective questionnaires and psychiatric expertise required by traditional assessment methods.
[0080] The present disclosure provides a spectrum weighting system that adapts learnable weights based on classification performance feedback and population-specific electrocardiogram characteristics, enabling deployment in rural healthcare environments where specialized mental health services are limited.
, Claims:1. A spectrum weighting system (102) for electrocardiogram-based stress identification, the system comprising:
at least one user equipment (108-1) comprising electrocardiogram electrodes in physical contact with a user to acquire electrocardiogram signals;
a frequency analysis module (314) comprising signal conversion circuitry electrically connected to the at least one user equipment (108-1), the frequency analysis module (314) positioned to receive the acquired electrocardiogram signals from the at least one user equipment (108-1) and extract frequency domain features from the acquired electrocardiogram signals;
an weight optimization module (316) electrically connected to the frequency analysis module (314), the weight optimization module (316) positioned to receive the frequency domain features from the frequency analysis module (314) and compute learnable weights for frequency components to discriminate between stress and non-stress patterns in the electrocardiogram signals;
a weighted feature generation unit (204-5) comprising multiplier circuits electrically connected to receive the frequency domain features from the frequency analysis module (314) and the learnable weights from the weight optimization module (316), the weighted feature generation unit (204-5) positioned to generate weighted frequency features by multiplying the frequency domain features with corresponding learnable weights;
a CNN processing module (318) comprising parallel processing circuits electrically connected to the weighted feature generation unit (204-5), the CNN processing module (318) positioned to receive the weighted frequency features from the weighted feature generation unit (204-5) and process the weighted frequency features to generate stress classification output;
a control unit (212) electrically connected to the frequency analysis module (314), the weight optimization module (316), the weighted feature generation unit (204-5), and the CNN processing module (318), the control unit (212) positioned to coordinate signal flow from the at least one user equipment (108-1) through each connected module sequentially to transform the physiological electrocardiogram signals into the stress classification output, enabling real-time stress monitoring in clinical settings.
2. The spectrum weighting system (102) as claimed in claim 1, wherein the frequency analysis module (314) is positioned to segment the acquired electrocardiogram signals into three-minute intervals before extracting the frequency domain features to capture stress-indicative temporal patterns.
3. The spectrum weighting system (102) as claimed in claim 1, wherein the weight optimization module (316) is positioned to iteratively adjust the learnable weights through gradient-based optimization that maximizes discrimination between stress-indicative frequency components and non-stress frequency components.
4. The spectrum weighting system (102) as claimed in claim 1, wherein the CNN processing module (318) comprises:
a first convolutional layer having 128 filters positioned to extract primary stress-related features from the weighted frequency features; and
a second convolutional layer having 64 filters positioned to extract hierarchical stress patterns from the primary stress-related features.
5. The spectrum weighting system (102) as claimed in claim 1, wherein the weighted feature generation unit (204-5) is positioned to enhance frequency components between 0.04 Hz to 0.15 Hz associated with heart rate variability while suppressing non-discriminative frequency components.
6. The spectrum weighting system (102) as claimed in claim 1, wherein the control unit (212) is further positioned to update the learnable weights based on classification performance feedback to adapt to user-specific stress patterns over time.
7. The spectrum weighting system (102) as claimed in claim 1, wherein the weight optimization module (316) comprises a feedback loop electrically connected to the CNN processing module (318) to receive classification accuracy data for dynamic weight adjustment.
8. The spectrum weighting system (102) as claimed in claim 1, wherein the system achieves stress identification accuracy improvement from 72.43% to 91.2% through the adaptive weighting of frequency components.
9. The spectrum weighting system (102) as claimed in claim 1, wherein the control unit (212) is positioned to identify stress patterns specific to rural healthcare environments by adapting the learnable weights based on population-specific electrocardiogram characteristics.
10. A method for electrocardiogram-based stress identification using spectrum weighting, the method comprising:
acquiring (402) electrocardiogram signals from a user through at least one user equipment (108-1) comprising electrocardiogram electrodes in physical contact with the user;
extracting (404) frequency domain features from the acquired electrocardiogram signals using a frequency analysis module (314) comprising signal conversion circuitry electrically connected to the at least one user equipment (108-1);
computing (406) learnable weights for frequency components using an weight optimization module (316) electrically connected to the frequency analysis module (314), wherein the learnable weights discriminate between stress and non-stress patterns in the electrocardiogram signals;
generating (408) weighted frequency features using a weighted feature generation unit (204-5) comprising multiplier circuits, wherein the weighted feature generation unit (204-5) receives the frequency domain features from the frequency analysis module (314) and the learnable weights from the weight optimization module (316), and multiplies the frequency domain features with corresponding learnable weights;
processing (410) the weighted frequency features using a CNN processing module (318) comprising parallel processing circuits electrically connected to the weighted feature generation unit (204-5) to generate stress classification output; and
coordinating (412) signal flow through a control unit (212) electrically connected to the frequency analysis module (314), the weight optimization module (316), the weighted feature generation unit (204-5), and the CNN processing module (318), wherein the control unit (212) transforms the physiological electrocardiogram signals into the stress classification output through sequential processing, enabling real-time stress monitoring in clinical settings.