1. A real-time breast cancer classifier system (100), the system (100) comprising: a user device (102) configured to collect real-time medical imaging data and clinical parameters including mammography images, patient demographics, vital signs and relevant diagnostic metadata from a user through a user interface (104); a communication network (106) configured to establish a communication link for data transmission within the system (100); an edge processing gateway (108) connected to the user device (102) via the communication network (106), the edge processing gateway (108) configured to perform on-device inference for breast lesion analysis, wherein the edge processing gateway (108) further comprises: a data input module (110) configured to receive real-time data from the user device (102); a pre-processing module (112) configured to clean, normalize and reduce noise in the received data; a feature extraction module (114) configured to extract most relevant features from the pre-processed data including texture features, morphological characteristics, intensity patterns, edge information, and region-based features associated with the potential breast lesions; a breast lesion detection module (116) configured to detect and localise abnormal breast lesions within the medical imaging data based on the extracted features; a breast lesion classification module (118) configured to classify the detected breast lesions into benign, and malignant categories; a decision support module (122) configured to recommend appropriate clinical actions based on the classification results of the detected breast lesions for review by healthcare professionals; an output module (132) configured to transmit the detected breast lesions, classified breast lesions, supported decision to the user device (102).

2. The system (100) as claimed in claim 1, wherein the system (100) further comprises a cloud tier (134) configured to store de-identified breast lesion imaging data, classification results, and the system (100) performance metrics for enabling federated learning and periodic model retraining.

3. The system (100) as claimed in claim 1, wherein the breast lesion classification module (118) utilizes quantized convolutional neural networks and fine-tuned efficientnet architectures, to enable high-performance classification of breast lesions.

4. The system (100) as claimed in claim 1, wherein the edge processing gateway (108) further comprises a clinical interpretation module (120) configured to generate visual heatmaps and feature-importance maps, thereby assisting clinicians in understanding breast lesion classification results and supporting informed decision-making during clinical evaluation.

5. The system (100) as claimed in claim 1, wherein the edge processing gateway (108) further comprises a report generation module (124) configured to generate standardized clinical reports including digital imaging and communications in medicine structured reports, health level seven messages, and fast healthcare interoperability resources messages.

6. The system (100) as claimed in claim 1, wherein the edge processing gateway (108) further comprises an alert generation module (126) configured to generate alerts to healthcare professionals based on the classification results of detected breast lesions, including high-risk findings, thereby enabling timely clinical intervention at the point of care.

7. The system (100) as claimed in claim 1, wherein the edge processing gateway (108) further comprises a feedback module (128) configured to collect model performance metrics, inference results, and clinician feedback, operate in an offline-first mode to store data locally, and synchronize with the cloud tier (134) when connectivity is available, thereby ensuring continuous evaluation of the system (100) and reliability.

8. The system (100) as claimed in claim 1, wherein the edge processing gateway (108) further comprises an edge control module (130) configured to perform digitally signed over-the-air updates and maintain the system (100) performance logs, and provide rollback functionality, thereby ensuring secure, reliable, and uninterrupted operation of the breast cancer classification.

9. A method (200) for real-time breast cancer classifier system (100), the method (200) comprising: collecting real-time medical imaging data and clinical parameters including mammography images, patient demographics, vital signs and relevant diagnostic metadata from a user through a user interface (104) via the user device (102); establishing a communication link for data transmission within the system (100) via the communication network (106); performing on-device inference for breast lesion analysis via the edge processing gateway (108), comprising several modules; receiving real-time data from the user device (102) via the data input module (110); cleaning, normalize and reduce noise in the received data via the pre-processing module (112); extracting most relevant features from the pre-processed data including texture features, morphological characteristics, intensity patterns, edge information, and region-based features associated with the potential breast lesions via the feature extraction module (114); detecting and localise abnormal breast lesions within the medical imaging data based on the extracted features via the breast lesion detection module (116); classifying the detected breast lesions into benign and malignant categories via the breast lesion classification module (118); generating visual heatmaps and feature-importance maps, thereby assisting clinicians in understanding breast lesion classification results and supporting informed decision-making via the clinical interpretation module (120); recommending appropriate clinical actions based on the classification results of the detected breast lesions for review by healthcare professionals via the decision support module (122); generating standardized clinical reports including digital imaging and communications in medicine structured reports, health level seven messages, and fast healthcare interoperability resources messages via the report generation module (124); generating alerts to healthcare professionals based on the classification results of detected breast lesions, including high-risk findings, thereby enabling timely clinical intervention at the point of care via the alert generation module (126); collecting model performance metrics, inference results, and clinician feedback, operate in an offline-first mode to store data locally, and synchronize with the cloud tier (134) when connectivity is available, thereby ensuring continuous evaluation of the system (100) and reliability via the feedback module (128); performing digitally signed over-the-air updates and maintain the system (100) performance logs, and provide rollback functionality, thereby ensuring secure, reliable, and uninterrupted operation of the breast cancer classification via the edge control module (130); transmitting the detected breast lesions, classified breast lesions, supported decision to the user device (102) via the output module (132).

10. The method (200) as claimed in claim 9, wherein the breast lesion classification module (118) configured to performs on-device inference of medical imaging data in the range of 70 frames per second, enabling rapid and real-time classification of detected breast lesions without relying on cloud connectivity.

Description:REAL-TIME BREAST CANCER CLASSIFIER SYSTEM
FIELD OF DISCLOSURE
[0001] The present disclosure generally relates to medical imaging and artificial intelligence systems, more specifically, relates to a real-time breast cancer classifier system.
BACKGROUND OF THE DISCLOSURE
[0002] A real-time breast cancer classifier system that uses optimized deep learning models, including Convolutional Neural Networks and EfficientNets, deployed on Internet of Things edge devices to analyse medical images locally. It provides fast, accurate lesion classification, generates standardized clinical reports, and integrates seamlessly with hospital systems while preserving patient privacy.
[0003] Traditional breast cancer classification systems face several limitations. Many rely on cloud-based processing, which introduces latency, bandwidth dependence, and potential privacy risks, making them unsuitable for real-time decision-making at the point of care. Semi-automated workflows are often slow, particularly in low-resource clinical settings, and depend heavily on the availability of radiologists and pathologists, increasing the risk of delays and human error. Existing Artificial Intelligence models are resource-intensive to run efficiently on local edge devices, and integration with hospital systems such as Picture Archiving and Communication system, Radiology Information System, and Electronic Health Records is often unreliable. Furthermore, the models trained on one dataset may perform poorly across different hospitals, devices, and patient populations due to variability in imaging and labelling, limiting their generalizability. Many conventional systems also lack real-time feedback capabilities, preventing immediate clinical intervention. Additionally, they unable to provide standardized reporting outputs compatible with hospital workflows, creating extra manual work for clinicians. Maintenance and updates for these systems tends to be cumbersome, and they lack robust mechanisms for monitoring model performance, detecting drift over time, and reducing reliability. They also often require high computational resources, making deployment in low-resource environments difficult. Existing systems generally unable to adapt heterogeneous imaging workflows, resulting in inconsistent performance across clinical sites. Moreover, auditability and traceability of Artificial Intelligence decisions are limited, making regulatory compliance and clinical validation challenging. Many systems also lack explainable Artificial Intelligence features, which reduces clinician confidence and makes interpretation of results difficult. They may be vulnerable to unauthorized access, particularly when connected to external networks. Additionally, conventional systems often unable to prioritize critical findings automatically, limiting efficiency in high-volume settings. Finally, a limited number of prospective multi-site studies and evolving regulatory guidelines often hinder widespread adoption and clinician trust.
[0004] The present invention solves the problem of the prior art by providing a real-time breast cancer classifier system that performs automated lesion detection and classification directly on edge computing devices within the hospital environment, eliminating dependence on cloud-based processing and reducing latency, bandwidth requirements, and privacy risks. The invention leverages optimized deep learning models, including customized Convolutional Neural Networks and fine-tuned EfficientNets, enabling highly accurate and resource-efficient inference on local hardware. The invention integrates seamlessly with hospital systems such as Picture Archiving and Communication Systems, Radiology Information Systems, and Electronic Health Records, and produces standardized clinical reports in real time, reducing manual workload for clinicians. Additionally, the invention includes monitoring and drift detection mechanisms, secure over-the-air updates, and adaptive capabilities for heterogeneous imaging devices and workflows, ensuring consistent performance across sites. By providing real-time feedback, explainable Artificial Intelligence outputs, and secure operation, the invention enhances diagnostic speed, accuracy, and clinician trust, while maintaining compliance with regulatory standards.
SUMMARY OF THE DISCLOSURE
[0005] The following is a summary description of illustrative embodiments of the invention. It is provided as a preface to assist those skilled in the art to more rapidly assimilate the detailed design discussion which ensues and is not intended in any way to limit the scope of the claims which are appended hereto in order to particularly point out the invention.
[0006] According to illustrative embodiments, the present disclosure focuses on a real-time breast cancer classifier system which overcomes the above-mentioned disadvantages or provide the users with a useful or commercial choice.
[0007] An objective of the present disclosure is to enable automated lesion detection and classification using optimized deep learning models, including Convolutional Neural Networks and EfficientNets.
[0008] Another objective of the present disclosure is to perform edge-based Artificial Intelligence inference on local computing devices, reducing latency, bandwidth dependence, and privacy risks associated with cloud-based solutions.
[0009] Another objective of the present disclosure is to provide fast and accurate lesion classification, supporting timely clinical decision-making in high-volume and low-resource settings.
[0010] Another objective of the present disclosure is to include drift monitoring and performance tracking, enabling continuous assessment of Artificial Intelligence model accuracy over time.
[0011] Another objective of the present disclosure is to offer adaptability across heterogeneous imaging devices and workflows, ensuring consistent performance across different hospitals and clinics.
[0012] Another objective of the present disclosure is to maintain patient privacy by performing all processing locally on edge devices, minimizing the need to transfer sensitive data externally.
[0013] Another objective of the present disclosure is to allow real-time feedback and alerting, enabling clinicians to intervene immediately when critical findings are detected.
[0014] Another objective of the present disclosure is to support federated learning and secure collaborative model improvement without sharing identifiable patient data between institutions.
[0015] Another objective of the present disclosure is to reduce manual workload for clinicians by automating repetitive tasks such as image annotation, scoring, and reporting.
[0016] Another objective of the present disclosure is to provide a reliable, scalable, and maintainable system architecture that integrates seamlessly with existing hospital infrastructure and complies with regulatory standards.
[0017] Another objective of the present disclosure is to facilitate integration with clinical dashboards and clinician user interfaces, improving workflow efficiency.
[0018] Another objective of the present disclosure is to provide secure encryption of all patient data stored locally, preventing unauthorized access.
[0019] Another objective of the present disclosure is to support offline-first operation, even in low-connectivity network environments.
[0020] Another objective of the present disclosure is to facilitate cross-institutional collaboration through federated learning, improving model generalizability across diverse populations.
[0021] Another objective of the present disclosure is to enable scalable deployment across multiple clinical sites without significant reconfiguration.
[0022] Another objective of the present disclosure is to improve overall patient care and diagnostic efficiency by combining speed, accuracy, and clinical workflow integration.
[0023] Yet another objective of the present disclosure is to allow prioritization of high-risk cases, enabling clinicians to act quickly on critical findings.
[0024] In light of the above, in one aspect of the present disclosure, a real-time breast cancer classifier system is disclosed herein. The system comprises a user device configured to collect real-time medical imaging data and clinical parameters including mammography images, patient demographics, vital signs and relevant diagnostic metadata from a user through a user interface. The system includes a communication network configured to establish a communication link for data transmission within the system. The system also includes an edge processing gateway connected to the user device via the communication network, the edge processing gateway configured to perform on-device inference for breast lesion analysis, wherein the edge processing gateway further comprises a data input module configured to receive real-time data from the user device, a pre-processing module configured to clean, normalize and reduce noise in the received data, a feature extraction module configured to extract most relevant features from the pre-processed data including texture features, morphological characteristics, intensity patterns, edge information, and region-based features associated with the potential breast lesions, a breast lesion detection module configured to detect and localise abnormal breast lesions within the medical imaging data based on the extracted features, a breast lesion classification module configured to classify the detected breast lesions into benign, and malignant categories, a decision support module configured to recommend appropriate clinical actions based on the classification results of the detected breast lesions for review by healthcare professionals, an output module configured to transmit the detected breast lesions, classified breast lesions, supported decision to the user device.
[0025] In one embodiment, the cloud tier configured to store de-identified breast lesion imaging data, classification results, and the system performance metrics for enabling federated learning and periodic model retraining.
[0026] In one embodiment, the breast lesion classification module utilizes quantized convolutional neural networks and fine-tuned efficientnet architectures, to enable high-performance classification of breast lesions.
[0027] In one embodiment, the clinical interpretation module configured to generate visual heatmaps and feature-importance maps, thereby assisting clinicians in understanding breast lesion classification results and supporting informed decision-making during clinical evaluation.
[0028] In one embodiment, the report generation module configured to generate standardized clinical reports including digital imaging and communications in medicine structured reports, health level seven messages, and fast healthcare interoperability resources messages.
[0029] In one embodiment, the alert generation module configured to generate alerts to healthcare professionals based on the classification results of detected breast lesions, including high-risk findings, thereby enabling timely clinical intervention at the point of care.
[0030] In one embodiment, the edge control module configured to perform digitally signed over-the-air updates and maintain the system performance logs, and provide rollback functionality, thereby ensuring secure, reliable, and uninterrupted operation of the breast cancer classification.
[0031] In light of the above, in one aspect of recent disclosure a method of making a real-time breast cancer classifier system is disclosed herein. The method comprises collecting real-time medical imaging data and clinical parameters including mammography images, patient demographics, vital signs and relevant diagnostic metadata from a user through a user interface via the user device. The method includes establishing a communication link for data transmission within the system via the communication network. The method also includes performing on-device inference for breast lesion analysis via the edge processing gateway. The method includes receiving real-time data from the user device via the data input module. The method also includes cleaning, normalize and reduce noise in the received data via the pre-processing module. The method also includes extracting most relevant features from the pre-processed data including texture features, morphological characteristics, intensity patterns, edge information, and region-based features associated with the potential breast lesions via the feature extraction module. The method further includes detecting and localise abnormal breast lesions within the medical imaging data based on the extracted features via the breast lesion detection module. The method also includes classifying the detected breast lesions into benign and malignant categories via the breast lesion classification module. The method further configured to performs on-device inference of medical imaging data in the range of 70 frames per second, enabling rapid and real-time classification of detected breast lesions without relying on cloud connectivity. The method includes generating visual heatmaps and feature-importance maps, thereby assisting clinicians in understanding breast lesion classification results and supporting informed decision-making via the clinical interpretation module. The method also includes recommending appropriate clinical actions based on the classification results of the detected breast lesions for review by healthcare professionals via the decision support module. The method includes generating standardized clinical reports including digital imaging and communications in medicine structured reports, health level seven messages, and fast healthcare interoperability resources messages via the report generation module. The method also includes generating alerts to healthcare professionals based on the classification results of detected breast lesions, including high-risk findings, thereby enabling timely clinical intervention at the point of care via the alert generation module. The method includes collecting model performance metrics, inference results, and clinician feedback, operate in an offline-first mode to store data locally, and synchronize with the cloud tier when connectivity is available, thereby ensuring continuous evaluation of the system and reliability via the feedback module. The method also includes performing digitally signed over-the-air updates and maintain the system performance logs, and provide rollback functionality, thereby ensuring secure, reliable, and uninterrupted operation of the breast cancer classification via the edge control module. Furthermore, the method also includes transmitting the detected breast lesions, classified breast lesions, supported decision to the user device via the output module.
[0032] These and other advantages will be apparent from the present application of the embodiments described herein.
[0033] The preceding is a simplified summary to provide an understanding of some embodiments of the present invention. This summary is neither an extensive nor exhaustive overview of the present invention and its various embodiments. The summary presents selected concepts of the embodiments of the present invention in a simplified form as an introduction to the more detailed description presented below. As will be appreciated, other embodiments of the present invention are possible utilizing, alone or in combination, one or more of the features set forth above or described in detail below.
[0034] These elements, together with the other aspects of the present disclosure and various features are pointed out with particularity in the claims annexed hereto and form a part of the present disclosure. For a better understanding of the present disclosure, its operating advantages, and the specified object attained by its uses, reference should be made to the accompanying drawings and descriptive matter in which there are illustrated exemplary embodiments of the present disclosure.
BRIEF DESCRIPTION OF THE DRAWINGS
[0035] To describe the technical solutions in the embodiments of the present disclosure or in the prior art more clearly, the following briefly describes the accompanying drawings required for describing the embodiments or the prior art. Apparently, the accompanying drawings in the following description merely show some embodiments of the present disclosure, and a person of ordinary skill in the art can derive other implementations from these accompanying drawings without creative efforts. All of the embodiments or the implementations shall fall within the protection scope of the present disclosure.
[0036] The advantages and features of the present disclosure will become better understood with reference to the following detailed description taken in conjunction with the accompanying drawing, in which:
[0037] FIG.1 illustrates a block diagram of a real-time breast cancer classifier system, in accordance with an embodiment of the present disclosure;
[0038] FIG.2 illustrates a detailed block diagram showing the internal functionality of the real-time breast cancer classifier system, in accordance with an embodiment of the present disclosure;
[0039] FIG.3 illustrates a flowchart of a method, outlining the sequential steps for the real-time breast cancer classifier system, in accordance with an embodiment of the present disclosure;
[0040] Like reference, numerals refer to like parts throughout the description of several views of the drawing.
[0041] The real-time breast cancer classifier system is illustrated in the accompanying drawings, which like reference letters indicate corresponding parts in the various figures. It should be noted that the accompanying figure is intended to present illustrations of exemplary embodiments of the present disclosure. This figure is not intended to limit the scope of the present disclosure. It should also be noted that the accompanying figure is not necessarily drawn to scale.
DETAILED DESCRIPTION OF THE DISCLOSURE
[0042] The following is a detailed description of embodiments of the disclosure depicted in the accompanying drawings. The embodiments are in such detail as to communicate the disclosure. However, the amount of detail offered is not intended to limit the anticipated variations of embodiments; on the contrary, the intention is to cover all modifications, equivalents, and alternatives falling within the scope of the present disclosure.
[0043] In the following description, numerous specific details are set forth in order to provide a thorough understanding of the embodiments of the present disclosure. It may be apparent to one skilled in the art that embodiments of the present disclosure may be practiced without some of these specific details.
[0044] Various terms as used herein are shown below. To the extent a term is used, it should be given the broadest definition persons in the pertinent art have given that term as reflected in printed publications and issued patents at the time of filing.
[0045] The terms “a” and “an” herein do not denote a limitation of quantity, but rather denote the presence of at least one of the referenced items.
[0046] The terms “having”, “comprising”, “including”, and variations thereof signify the presence of a component.
[0047] Referring now to FIG. 1 to FIG. 3 to describe various exemplary embodiments of the present disclosure. FIG. 1 illustrates a block diagram of a real-time breast cancer classifier system 100, in accordance with an embodiment of the present disclosure.
[0048] The system 100 may include a user device 102. The system 100 may include a user interface 104. The system 100 may include a communication network 106. The system 100 may include an edge processing gateway 108, further comprises a data input module 110, a pre-processing module 112, a feature extraction module 114, a breast lesion detection module 116, a breast lesion classification module 118, a clinical interpretation module 120, a decision support module 122, a report generation module 124, an alert generation module 126, a feedback module 128, an edge control module 130, an output module 132.The system 100 may include a cloud tier 134.
[0049] The user device 102 configured to collect real-time medical imaging data and clinical parameters including mammography images, patient demographics, vital signs and relevant diagnostic metadata from a user through a user interface 104. The user device 102 may include a workstation, desktop computer, tablet, mobile device, and imaging console that allows clinicians to enter patient information, upload imaging data, and provide feedback.
[0050] In one embodiment of the present invention, the user interface 104 further includes graphical elements such as forms, buttons, and display panels, allowing users to confirm patient demographics, annotate regions of interest in images, and provide feedback on Artificial Intelligence outputs.
[0051] In one embodiment of the present invention, the user interface 104 may support real-time visualization of imaging data, enabling immediate quality checks and adjustments.
[0052] The communication network 106 configured to establish a communication link for data transmission within the system 100. The communication network 106 enabling secure transmission of real-time clinical parameters, physiological data and health information.
[0053] In one embodiment of the present invention, the communication network 106 may include but not limited to wired and wireless networks.
[0054] In one embodiment of the present invention, the communication network 106 includes wireless networks including but not limited to Bluetooth and wireless fidelity.
[0055] The edge processing gateway 108 connected to the user device 102 via the communication network 106, the edge processing gateway 108 configured to perform on-device inference for breast lesion analysis. The edge processing gateway 108 ensures rapid analysis, data privacy, and reliable operation even when cloud connectivity is limited, thereby supporting timely and informed clinical decision-making.
[0056] The data input module 110 configured to receive real-time data from the user device 102. The data input module 110 responsible for collecting, validating, and organizing incoming data, including mammography images, patient demographic information, diagnostic metadata, and other relevant clinical inputs required for breast lesion analysis.
[0057] The pre-processing module 112 configured to clean, normalize and reduce noise in the received data. The pre-processing module 112 including image resizing, intensity normalization, contrast enhancement, and removal of artifacts and irrelevant background information.
[0058] The feature extraction module 114 configured to extract most relevant features from the pre-processed data including texture features, morphological characteristics, intensity patterns, edge information, and region-based features associated with the potential breast lesions. The feature extraction module 114 analyses the processed mammography images to identify distinctive patterns and structural characteristics that may indicate abnormal tissue regions.
[0059] The breast lesion detection module 116 configured to detect and localise abnormal breast lesions within the medical imaging data based on the extracted features. The breast lesion detection module 116 may utilize deep learning-based detection techniques to scan the medical images and determine the location, boundaries, and spatial coordinates of suspected lesions.
[0060] In one embodiment of the present invention, the breast lesion detection module 116 may be represented using segmentation masks, along with confidence scores indicating the abnormality
[0061] The breast lesion classification module 118 configured to classify the detected breast lesions into benign, and malignant categories. The breast lesion classification module 118 evaluates characteristics such as lesion shape, margin irregularity, texture distribution, and intensity patterns to distinguish between benign and malignant findings and generate confidence scores associated with each detected lesion.
[0062] In one embodiment of the present invention, the breast lesion classification module 118 utilizes quantized convolutional neural networks and fine-tuned efficientnet architectures, to enable high-performance classification of breast lesions.
[0063] In one embodiment of the present invention, the breast lesion classification module 118 further includes convolutional neural networks are configured to automatically learn hierarchical feature representations from the detected lesion regions, thereby capturing complex patterns related to tissue structure and abnormal growth characteristics.
[0064] In one embodiment of the present invention, the breast lesion classification module 118 further includes EfficientNet architecture, which is optimized through compound scaling of network depth, width, and resolution, improves classification accuracy while maintaining computational efficiency.
[0065] In one embodiment of the present invention, the breast lesion classification module 118 may be trained and fine-tuned using annotated breast imaging datasets to enhance its ability to accurately differentiate between benign and malignant lesions by utilizing these optimized deep learning techniques. The breast lesion classification module 118 achieves fast, reliable, and resource-efficient inference, thereby supporting real-time breast lesion analysis within the system 100.
[0066] In one embodiment of the present invention, the breast lesion classification module 118 further includes INT8 (8-bit-integer) quantized to reduce computational complexity and memory consumption, thereby enabling efficient real-time inference on the edge processing gateway.
[0067] The clinical interpretation module 120 configured to generate visual heatmaps and feature-importance maps, thereby assisting clinicians in understanding breast lesion classification results and supporting informed decision-making during clinical evaluation. The clinical interpretation module 120 analyses the outputs generated by the breast lesion classification module 118 and produces visual representations that highlight the regions of the medical image that contributed most significantly to the classification decision.
[0068] In one embodiment of the present invention, the clinical interpretation module 120 further includes heatmaps may be overlaid on the original mammography images to visually indicate areas of higher diagnostic relevance, while the feature-importance maps illustrate the relative contribution of different image features influencing the classification outcome.
[0069] The decision support module 122 configured to recommend appropriate clinical actions based on the classification results of the detected breast lesions for review by healthcare professionals. The decision support module 122 may include suggested follow-up imaging, biopsy, referral to a specialist, and monitoring schedules.
[0070] The report generation module 124 configured to generate standardized clinical reports including digital imaging and communications in medicine structured reports, health level seven messages, and fast healthcare interoperability resources messages. The report generation module 124 seamlessly integrated into clinical workflows, facilitating efficient review, improved communication among healthcare professionals, and timely patient management decisions.
[0071] In one embodiment of the present invention, the report generation module 124 further includes Digital Imaging and Communications in Medicine Structured Reports for image-linked documentation, Health Level Seven messages for secure communication between hospital information systems, and Fast Healthcare Interoperability Resources messages for modern web-based healthcare data exchange.
[0072] The alert generation module 126 configured to generate alerts to healthcare professionals based on the classification results of detected breast lesions, including high-risk findings, thereby enabling timely clinical intervention at the point of care. The alert generation module 126 automatically triggers notifications, warnings, and priority flags when lesions are classified as malignant.
[0073] In one embodiment of the present invention, the alert generation module 126 delivered through the user device 102, hospital information systems, mobile applications, and secure messaging platforms.
[0074] The feedback module 128 configured to collect model performance metrics, inference results, and clinician feedback, operate in an offline-first mode to store data locally, and synchronize with the cloud tier 134 when connectivity is available, thereby ensuring continuous evaluation of the system 100 and reliability.
[0075] In one embodiment of the present invention, the feedback module 128 may include prediction accuracy, confidence scores, detected model drift, and clinician annotations. The feedback module 128 supports the system 100 reliability, ongoing performance monitoring, and future model fine-tuning, ensuring that the Artificial Intelligence models remain accurate, robust, and clinically trustworthy over time.
[0076] The edge control module 130 configured to perform digitally signed over-the-air updates and maintain the system 100 performance logs, and provide rollback functionality, thereby ensuring secure, reliable, and uninterrupted operation of the breast cancer classification. The edge control module 130 manages the deployment of updated Artificial Intelligence models, configuration changes, and software patches to the edge processing gateway 108 while verifying digital signatures to prevent unauthorized modifications.
[0077] In one embodiment of the present invention, the edge control module 130 continuously monitors the system 100 performance metrics and maintains logs to enable auditability, troubleshooting, and compliance with regulatory standards.
[0078] The outputs module 132 configured to transmit the detected breast lesions, classified breast lesions, supported decision to the user device 102. The output module 132 configured to directly integrate these outputs into hospital systems, such as the Picture Archiving and Communication System, Radiology Information System, and Electronic Health Record, in compliance with standards such as Digital imaging and communications in medicine Structured Reports, Health Level Seven, and Fast Healthcare Interoperability Resources.
[0079] The cloud tier 134 configured to store de-identified breast lesion imaging data, classification results, and the system 100 performance metrics for enabling federated learning and periodic model retraining. The cloud tier 134 collects information across different clinical sites, aggregating non-personally identifiable data to ensure patient privacy while enabling large-scale Artificial Intelligence model improvement.
[0080] In one embodiment of the present invention, the cloud tier 134 supports periodic model retraining, ensuring that the breast lesion classification models remain accurate, robust, and generalizable to diverse imaging devices, patient populations, and clinical conditions.
[0081] FIG.2 illustrates a detailed block diagram showing the internal functionality of the real-time breast cancer classifier system, in accordance with an embodiment of the present disclosure;
[0082] At step 202, the process begins with a medical imaging device capturing raw patient data and generates a digital image.
[0083] At step 204, the system 100 transmits this data by using Digital Imaging and Communications in Medicine Composite Storage and Digital Imaging and Communications in Medicine web.
[0084] At step 206, the edge ingestion receives the high-resolution files directly at the edge of the network.
[0085] At step 208, the pre-processing module 112 cleans the data by normalizing pixel values and resizing images.
[0086] At step 210, the optimized convolutional neural network analyses the pre-processed image to identify potential abnormalities.
[0087] At step 212, the metrics and drift monitor pushes de-identified metadata to a cloud tier to monitor for model drift and facilitates federated Artificial Intelligence.
[0088] At step 214, the device manager handles over-the-air updates, maintains logs and messages software rollbacks to ensure the system 100 stays online and secure.
[0089] At step 216, the system 100 integrates the artificial intelligence findings into the hospital existing databases, such as picture archiving communication system, radiology information system and electronic health record.
[0090] At step 218, the result writer converts the artificial intelligence findings into standardized clinical reports.
[0091] At step 220, the system delivers the data via Digital Imaging and Communications in Medicine-Structural Reporting, Health Level Seven, and Fast Healthcare Interoperability Resources directly to the clinical user interface for final review.
[0092] FIG.3 illustrates a flowchart of a method, outlining the sequential steps for the real-time breast cancer classifier system, in accordance with an embodiment of the present disclosure;
[0093] At step 302, collecting the real-time medical imaging data and clinical parameters is collected including mammography images, patient demographics, vital signs and relevant diagnostic metadata from a user through a user interface 104 via the user device 102.
[0094] At step 304, the communication link is established for data transmission within the system 100 via the communication network 106.
[0095] At step 306, the on-device inference is performed for breast lesion analysis via the edge processing gateway 108, comprising several modules;
[0096] At step 308, the real-time data is received from the user device 102 via the data input module 110.
[0097] At step 310, the received data is cleaned, normalize and reduce noise via the pre-processing module 112.
[0098] At step 312, the most relevant features is extracted from the pre-processed data including texture features, morphological characteristics, intensity patterns, edge information, and region-based features associated with the potential breast lesions via the feature extraction module 114.
[0099] At step 314, the abnormal breast lesions is detected and localise within the medical imaging data based on the extracted features via the breast lesion detection module 116.
[0100] At step 316, the detected breast lesions is classified into benign and malignant categories via the breast lesion classification module 118.
[0101] In one embodiment of the present invention, the breast lesion classification module 118 configured to performs on-device inference of medical imaging data in the range of 70 frames per second, enabling rapid and real-time classification of detected breast lesions without relying on cloud connectivity.
[0102] At step 318, the visual heatmaps is generated and feature-importance maps, thereby assisting clinicians in understanding breast lesion classification results and supporting informed decision-making via the clinical interpretation module 120.
[0103] At step 320, the appropriate clinical actions is recommended based on the classification results of the detected breast lesions for review by healthcare professionals via the decision support module 122.
[0104] At step 322, the standardized clinical reports is generated including digital imaging and communications in medicine structured reports, health level seven messages, and fast healthcare interoperability resources messages via the report generation module 124.
[0105] At step 324, the healthcare professionals is generated alerts based on the classification results of detected breast lesions, including high-risk findings, thereby enabling timely clinical intervention at the point of care via the alert generation module 126.
[0106] At step 326, the model performance metrics, inference results, and clinician feedback is collected, operate in an offline-first mode to store data locally, and synchronize with the cloud tier 134 when connectivity is available, thereby ensuring continuous evaluation of the system 100 and reliability via the feedback module 128.
[0107] At step 328, the digitally signed over-the-air updates is performed and maintain the system 100 performance logs, and provide rollback functionality, thereby ensuring secure, reliable, and uninterrupted operation of the breast cancer classification via the edge control module 130.
[0108] At step 330, the detected breast lesions, classified breast lesions, supported decision is transmitted to the user device 102 via the output module 132.
[0109] In the best mode of operation of the present invention, the system 100 operates by utilizing the user device 102 to capture real-time medical imaging data and associated clinical parameters including mammography images, patient demographics, vital signs, and relevant diagnostic metadata through the user interface 104. The collected data is transmitted through the communication network 106 to the edge processing gateway 108, wherein the on-device inference is performed for breast lesion analysis. The data input module 110 receives the transmitted data and forwards it to the pre-processing module 112, which performs data cleaning, normalization, and noise reduction to improve data quality. Subsequently, the feature extraction module 114 extracts relevant features including texture patterns, morphological characteristics, intensity distributions, edge information, and region-based features associated with potential breast lesions. The extracted features are analysed by the breast lesion detection module 116 to detect and localize abnormal breast lesions within the medical imaging data. The breast lesion classification module 118 then classifies the detected breast lesions into benign and malignant categories by performing optimized on-device inference in the range of 70 frames per second. Furthermore, the clinical interpretation module 120 generates visual heatmaps and feature-importance maps to assist clinicians in understanding the classification results. Based on the classification outcomes, the decision support module 122 recommends appropriate clinical actions for review by healthcare professionals. The report generation module 124 generates standardized clinical reports including Digital Imaging and Communications in Medicine structured reports, Health Level Seven messages, and Fast Healthcare Interoperability Resources messages. Additionally, the alert generation module 126 generates alerts for healthcare professionals in the case of high-risk findings to enable timely medical intervention. The feedback module 128 collects model performance metrics, inference results, and clinician feedback, operates in an offline-first mode for local storage, and synchronizes with the cloud tier 134 when connectivity becomes available. The edge control module 130 performs digitally signed over-the-air updates, maintains performance logs, and provides rollback functionality to ensure secure and reliable the system 100 operation. Finally, the output module 132 transmits the detected lesions, classification results, and decision support information back to the user device 102 for clinical review and further medical assessment.
[0110] While the invention has been described in connection with what is presently considered to be the most practical and various embodiments, it will be understood that the invention is not to be limited to the disclosed embodiments, but on the contrary, is intended to cover various modifications and equivalent arrangements included within the scope of the appended claims.
[0111] A person of ordinary skill in the art may be aware that, in combination with the examples described in the embodiments disclosed in this specification, units and algorithm steps may be implemented by electronic hardware, computer software, or a combination thereof.
[0112] The foregoing descriptions of specific embodiments of the present disclosure have been presented for purposes of illustration and description. They are not intended to be exhaustive or to limit the present disclosure to the precise forms disclosed, and many modifications and variations are possible in light of the above teaching. The embodiments were chosen and described to best explain the principles of the present disclosure and its practical application, and to thereby enable others skilled in the art to best utilize the present disclosure and various embodiments with various modifications as are suited to the particular use contemplated. It is understood that various omissions and substitutions of equivalents are contemplated as circumstances may suggest or render expedient, but such omissions and substitutions are intended to cover the application or implementation without departing from the scope of the present disclosure.
[0113] Disjunctive language such as the phrase “at least one of X, Y, Z,” unless specifically stated otherwise, is otherwise understood with the context as used in general to present that an item, term, etc., may be either X, Y, or Z, or any combination thereof (e.g., X, Y, and/or Z). Thus, such disjunctive language is not generally intended to, and should not, imply that certain embodiments require at least one of X, at least one of Y, or at least one of Z to each be present.
[0114] In a case that no conflict occurs, the embodiments in the present disclosure and the features in the embodiments may be mutually combined. The foregoing descriptions are merely specific implementations of the present disclosure but are not intended to limit the protection scope of the present disclosure. Any variation or replacement readily figured out by a person skilled in the art within the technical scope disclosed in the present disclosure shall fall within the protection scope of the present disclosure. Therefore, the protection scope of the present disclosure shall be subject to the protection scope of the claims.
, Claims:I/We Claim:
1. A real-time breast cancer classifier system (100), the system (100) comprising:
a user device (102) configured to collect real-time medical imaging data and clinical parameters including mammography images, patient demographics, vital signs and relevant diagnostic metadata from a user through a user interface (104);
a communication network (106) configured to establish a communication link for data transmission within the system (100);
an edge processing gateway (108) connected to the user device (102) via the communication network (106), the edge processing gateway (108) configured to perform on-device inference for breast lesion analysis, wherein the edge processing gateway (108) further comprises:
a data input module (110) configured to receive real-time data from the user device (102);
a pre-processing module (112) configured to clean, normalize and reduce noise in the received data;
a feature extraction module (114) configured to extract most relevant features from the pre-processed data including texture features, morphological characteristics, intensity patterns, edge information, and region-based features associated with the potential breast lesions;
a breast lesion detection module (116) configured to detect and localise abnormal breast lesions within the medical imaging data based on the extracted features;
a breast lesion classification module (118) configured to classify the detected breast lesions into benign, and malignant categories;
a decision support module (122) configured to recommend appropriate clinical actions based on the classification results of the detected breast lesions for review by healthcare professionals;
an output module (132) configured to transmit the detected breast lesions, classified breast lesions, supported decision to the user device (102).
2. The system (100) as claimed in claim 1, wherein the system (100) further comprises a cloud tier (134) configured to store de-identified breast lesion imaging data, classification results, and the system (100) performance metrics for enabling federated learning and periodic model retraining.
3. The system (100) as claimed in claim 1, wherein the breast lesion classification module (118) utilizes quantized convolutional neural networks and fine-tuned efficientnet architectures, to enable high-performance classification of breast lesions.
4. The system (100) as claimed in claim 1, wherein the edge processing gateway (108) further comprises a clinical interpretation module (120) configured to generate visual heatmaps and feature-importance maps, thereby assisting clinicians in understanding breast lesion classification results and supporting informed decision-making during clinical evaluation.
5. The system (100) as claimed in claim 1, wherein the edge processing gateway (108) further comprises a report generation module (124) configured to generate standardized clinical reports including digital imaging and communications in medicine structured reports, health level seven messages, and fast healthcare interoperability resources messages.
6. The system (100) as claimed in claim 1, wherein the edge processing gateway (108) further comprises an alert generation module (126) configured to generate alerts to healthcare professionals based on the classification results of detected breast lesions, including high-risk findings, thereby enabling timely clinical intervention at the point of care.
7. The system (100) as claimed in claim 1, wherein the edge processing gateway (108) further comprises a feedback module (128) configured to collect model performance metrics, inference results, and clinician feedback, operate in an offline-first mode to store data locally, and synchronize with the cloud tier (134) when connectivity is available, thereby ensuring continuous evaluation of the system (100) and reliability.
8. The system (100) as claimed in claim 1, wherein the edge processing gateway (108) further comprises an edge control module (130) configured to perform digitally signed over-the-air updates and maintain the system (100) performance logs, and provide rollback functionality, thereby ensuring secure, reliable, and uninterrupted operation of the breast cancer classification.
9. A method (200) for real-time breast cancer classifier system (100), the method (200) comprising:
collecting real-time medical imaging data and clinical parameters including mammography images, patient demographics, vital signs and relevant diagnostic metadata from a user through a user interface (104) via the user device (102);
establishing a communication link for data transmission within the system (100) via the communication network (106);
performing on-device inference for breast lesion analysis via the edge processing gateway (108), comprising several modules;
receiving real-time data from the user device (102) via the data input module (110);
cleaning, normalize and reduce noise in the received data via the pre-processing module (112);
extracting most relevant features from the pre-processed data including texture features, morphological characteristics, intensity patterns, edge information, and region-based features associated with the potential breast lesions via the feature extraction module (114);
detecting and localise abnormal breast lesions within the medical imaging data based on the extracted features via the breast lesion detection module (116);
classifying the detected breast lesions into benign and malignant categories via the breast lesion classification module (118);
generating visual heatmaps and feature-importance maps, thereby assisting clinicians in understanding breast lesion classification results and supporting informed decision-making via the clinical interpretation module (120);
recommending appropriate clinical actions based on the classification results of the detected breast lesions for review by healthcare professionals via the decision support module (122);
generating standardized clinical reports including digital imaging and communications in medicine structured reports, health level seven messages, and fast healthcare interoperability resources messages via the report generation module (124);
generating alerts to healthcare professionals based on the classification results of detected breast lesions, including high-risk findings, thereby enabling timely clinical intervention at the point of care via the alert generation module (126);
collecting model performance metrics, inference results, and clinician feedback, operate in an offline-first mode to store data locally, and synchronize with the cloud tier (134) when connectivity is available, thereby ensuring continuous evaluation of the system (100) and reliability via the feedback module (128);
performing digitally signed over-the-air updates and maintain the system (100) performance logs, and provide rollback functionality, thereby ensuring secure, reliable, and uninterrupted operation of the breast cancer classification via the edge control module (130);
transmitting the detected breast lesions, classified breast lesions, supported decision to the user device (102) via the output module (132).
10. The method (200) as claimed in claim 9, wherein the breast lesion classification module (118) configured to performs on-device inference of medical imaging data in the range of 70 frames per second, enabling rapid and real-time classification of detected breast lesions without relying on cloud connectivity.