Description:FIELD OF THE INVENTION
The present invention relates to the field of medical diagnostics and artificial intelligence. More specifically, it concerns an IoT-enabled system and method for lung nodule detection using a machine learning classification model. The invention integrates cloud, fog, and IoT frameworks with a Support Vector Machine-based diagnostic pipeline to enable efficient, accurate, and real-time identification of lung nodules from medical imaging data.
BACKGROUND OF THE INVENTION
Lung nodule detection is currently aided by a number of commercial products and existing solutions that use imaging and machine learning models approaches like Random Forest, Decision Tree, Logistic Regression etc. AI-driven modules for automated lung cancer detection have been incorporated into tools like Siemens Syngo.via, IBM Watson Health, and Google's Deep Mind. Clinical decision support tools such as GE Healthcare's Thoracic VCAR and Philips IntelliSpace Portal help radiologists identify nodules and determine their risk of cancer. Furthermore, research frequently uses open-source resources such as the LUNA16 and LIDC-IDRI datasets to create and evaluate lung nodule detection algorithms. In order to help radiologists evaluate scans more quickly, commercial practice usually uses Computer-Aided Diagnosis (CAD) systems integrated into PACS (Picture Archiving and Communication Systems) in hospitals. These technologies lower the danger of supervision, indicate malignancy scores, and reveal possible nodules. Despite their usefulness, a lot of the current solutions rely significantly on deep learning, which necessitates a lot of processing power and sizable annotated datasets. By helping radiologists recognize, categorize, and assess nodules based on CT scans, these tools are incorporated into radiology processes to increase diagnostic efficiency and accuracy.
US11730387B2: A method of detecting and diagnosing cancers characterized by the presence of at least one nodule/neoplasm from an imaging scan is presented. To detect nodules in an imaging scan, a 3D CNN using a single feed forward pass of a single network is used. After detection, risk stratification is performed using a supervised or an unsupervised deep learning method to assist in characterizing the detected nodule/neoplasm as benign or malignant. The supervised learning method relies on a 3D CNN used with transfer learning and a graph regularized sparse MTL to determine malignancy. The unsupervised learning method uses clustering to generate labels after which label proportions are used with a novel algorithm to classify malignancy. The method assists radiologists in improving detection rates of lung nodules to facilitate early detection and minimizing errors in diagnosis.
Lung cancer remains a leading cause of mortality, where early identification of lung nodules significantly increases survival rates. Current diagnostic methods often depend on radiologists’ manual interpretation, which is time-consuming and prone to variability. Existing automated systems demand large annotated datasets, high computational resources, or are prone to misclassifying non-nodular structures as nodules, resulting in false positives and increased clinical workload. Furthermore, many systems lack adaptability across different datasets, resolution levels, and patient demographics. The present invention addresses these issues by providing an IoT-enabled framework that distributes computational load across fog and cloud resources, leverages Support Vector Machine models for efficient classification of nodules, reduces false positives, and delivers interpretable results for clinical decision support.
SUMMARY OF THE INVENTION
This summary is provided to introduce a selection of concepts, in a simplified format, that are further described in the detailed description of the invention.
This summary is neither intended to identify key or essential inventive concepts of the invention and nor is it intended for determining the scope of the invention.
The invention proposes a distributed IoT-enabled framework for automated lung nodule detection using a Support Vector Machine model. Medical imaging data, obtained through computed tomography scans, is acquired through IoT-enabled devices and preprocessed for noise reduction, normalization, and lung segmentation. Gateway and fog nodes perform initial data processing, while cloud resources manage computationally intensive tasks under the supervision of cloud controllers and protection modules.
The Support Vector Machine classifier processes candidate regions of interest, distinguishing nodules from non-nodular structures. A false positive reduction mechanism further refines the output, ensuring only verified nodules are highlighted on original scans. The distributed architecture integrates fog nodes, cloud data centers, service directors, and security supervisors, ensuring real-time analysis, scalability, and secure communication.
By combining IoT deployment, machine learning, and distributed computing, the invention provides a robust, adaptive, and resource-efficient solution. It improves diagnostic precision, reduces radiologists’ workload, and ensures accessibility for both advanced and resource-constrained healthcare settings.
To further clarify advantages and features of the present invention, a more particular description of the invention will be rendered by reference to specific embodiments thereof, which is illustrated in the appended drawings. It is appreciated that these drawings depict only typical embodiments of the invention and are therefore not to be considered limiting of its scope. The invention will be described and explained with additional specificity and detail with the accompanying drawings.
The architecture, design, implementation, and operational technique of the proposed invention are described in detail here. The various components that make up this proposed invention's design are seen in Figure 1 and discussed in more detail below. The suggested study integrates cloud computing, fog, and IoT techniques for optimal predictive analytics.
BRIEF DESCRIPTION OF THE DRAWINGS
The illustrated embodiments of the subject matter will be understood by reference to the drawings, wherein like parts are designated by like numerals throughout. The following description is intended only by way of example, and simply illustrates certain selected embodiments of devices, systems, and methods that are consistent with the subject matter as claimed herein, wherein:
Figure.1: Block diagram of proposed innovation
Figure.2: Block diagram of proposed Methodology
The figures depict embodiments of the present subject matter for the purposes of illustration only. A person skilled in the art will easily recognize from the following description that alternative embodiments of the structures and methods illustrated herein may be employed without departing from the principles of the disclosure described herein.
DETAILED DESCRIPTION OF THE INVENTION
The detailed description of various exemplary embodiments of the disclosure is described herein with reference to the accompanying drawings. It should be noted that the embodiments are described herein in such details as to clearly communicate the disclosure. However, the amount of details provided herein 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 as defined by the appended claims.
It is also to be understood that various arrangements may be devised that, although not explicitly described or shown herein, embody the principles of the present disclosure. Moreover, all statements herein reciting principles, aspects, and embodiments of the present disclosure, as well as specific examples, are intended to encompass equivalents thereof.
The terminology used herein is for the purpose of describing particular embodiments only and is not intended to be limiting of example embodiments. As used herein, the singular forms “a",” “an” and “the” are intended to include the plural forms as well, unless the context clearly indicates otherwise. It will be further understood that the terms “comprises,” “comprising,” “includes” and/or “including,” when used herein, specify the presence of stated features, integers, steps, operations, elements and/or components, but do not preclude the presence or addition of one or more other features, integers, steps, operations, elements, components and/or groups thereof.
It should also be noted that in some alternative implementations, the functions/acts noted may occur out of the order noted in the figures. For example, two figures shown in succession may, in fact, be executed concurrently or may sometimes be executed in the reverse order, depending upon the functionality/acts involved.
In addition, the descriptions of "first", "second", “third”, and the like in the present invention are used for the purpose of description only, and are not to be construed as indicating or implying their relative importance or implicitly indicating the number of technical features indicated. Thus, features defining "first" and "second" may include at least one of the features, either explicitly or implicitly.
Unless otherwise defined, all terms (including technical and scientific terms) used herein have the same meaning as commonly understood by one of ordinary skill in the art to which example embodiments belong. It will be further understood that terms, e.g., those defined in commonly used dictionaries, should be interpreted as having a meaning that is consistent with their meaning in the context of the relevant art and will not be interpreted in an idealized or overly formal sense unless expressly so defined herein.
The architecture, design, implementation, and operational technique of the proposed invention are described in detail here. The various components that make up this proposed invention's design are seen in Figure 1 and discussed in more detail below. The suggested study integrates cloud computing, fog, and IoT techniques for optimal predictive analytics.
Components Used:
The IoT endpoints, gateway devices, computer, fog worker nodes, and cloud data center nodes are among the hardware components needed for the suggested operation. After IoT end devices identify scanning data, gateway devices receive and handle the information. Patient information that is entered via a gateway device—a computer or smart phone—will be sent to the system's master or innovation nodes. The gateway hardware operates similarly to the fog device. In order to generate output, the computer takes work requests from gateway devices in one of two ways: either directly through a trained Machine Learning (Support vector Machine) model or as data distribution to accessible invention nodes via an information director. As the Fog worker nodes reach their maximum capacity, the computer becomes a gateway device, effectively directing traffic to other Cloud-data-center nodes while being monitored by a Cloud controller that guarantees additional load handling. When a computer or gateway device requests information, a fog worker node processes the data and generates results based on a machine learning (Support Vector Machine) model it has discovered. This research project uses Raspberry Pi devices to install worker nodes. Anytime during operational requirements, users can connect to Cloud resources thanks to the Cloud-data-center node. An information director, a service director, a protection supervisor, a cloud controller, and a service observer with a machine learning-based model are among the software components that make up the suggested innovation. The information director system's capabilities allows for the analysis of data from IoT devices. This system has the capacity to combine data from many sources and adjust the speed at which data is transferred. The next Fog worker nodes where data will exchange information are chosen by the information director. Setting aside adequate funds for the program is one of the duties of the service director. The compute server's service observer looks at the resource status of each computer in the system as well as each Fog worker node.
The system uses the warehouse services app catalog to assess the needs of various programs. After receiving the information, the service director deploys the required Cloud and Fog worker node infrastructure. To ensure that users can access the system, the computer protection supervisor verifies that the gateway device authentication credentials are valid. The protection supervisor for the Fog worker node examines how the node transmits secure communications to different entities while engaging in computer operations. When a cloud storage and resource request is made, the framework is notified to display cloud-based instances, such as virtual devices and containers. Programs are executed in real time, and the service observer, which manages resource allocation duties, checks their accuracy against implementation requirements.
Framework Design and Implementation:
iFog Sim is used to model IoT and fog settings in order to ascertain features pertaining to latency, congestion, energy consumption, and overall expenses. Using iFog Sim, developers may test cost performance metrics, network use, and perceived latency. In order to bridge the gap between cloud systems, IoT devices, and fog settings, Fog Bus acts as a connector. Developers can construct IoT interfaces that are independent of certain platforms by using Fog Bus.
Working Principle:
Thesuggested innovation makes use of multiple computational processes. In this suggested invention, fog worker nodes are the slaves and the computer is the master. The same network is used by devices like the PC, fog worker nodes, and gateway equipment. Using the computer alone, the computer and the Fog worker nodes, or the Cloud node alone are the three methods of communication. In the first scenario, the computer finishes the task and provides the outcome; in the second scenario, the Fog worker node does this. They act as a gate way device, forwarding to the Clouds when the computer and the fog worker nodes are overloaded because of a lack of resources.
Support Vector Machine Algorithm using Machine Learning:
The figure 2 shows a thorough machine learning pipeline for lung nodule detection that makes use of the Support Vector Machine (SVM) technique. A Computed Tomography (CT) scanner
takes precise interior images of the patient's chest during the scanning part of the procedure. High-resolution cross-sectional views from these scans are crucial for spotting anomalies in the lung tissues. After scanning, the CT scan pictures are gathered and saved in a digital format, usually DICOM, during the data acquisition phase. This step provides a volumetric picture of the thoracic cavity by capturing several image slices that correspond to various axial views of the lungs.
Data pre-processing is a fundamental process that enhances the quality and standardization of images at the time of data collection. Noise reduction, image normalization, scaling, and enhancement of contrast are some examples of pre-processing techniques. These operations are carried out to pre-condition the data for precise analysis by removing irrelevant data, emphasizing relevant features, and making data homogeneous. At image processing at the lung segmentation level, the lung areas alone are extracted from the remaining chest region. Thresholding and morphological operations are utilized for this purpose. Lung segmentation provides benefits such as the reduction of the volume of computing workload and focused exploration of areas where nodules are likely to occur. Once segmentation is performed, the algorithm infers probable regions of interest (ROIs) with nodules during lung nodule detection. Candidate nodules are chosen on intensity, shape, and texture features of segmented lung areas. However, this process results in a large number of false-positive regions that are nodule-like but contain no nodule. To address this problem, the process proceeds to classification by utilizing Support Vector Machines (SVM). The SVM algorithm, a supervised machine learner, is known for its high rate of performance in binary classification tasks. By observing the features of candidate nodules through labeled training data, it builds the ability to differentiate between real nodules and false positives.
Both linear and non-linear data distributions are handled by the SVM classifier using kernels, such as linear or radial basis functions. A false positive reduction step is used after categorization. Through the removal of non-nodular structures that the SVM model misclassified, this step further narrows down the candidate pool. To improve specificity and lower diagnostic mistakes, methods like secondary classifiers or post-processing filters may be applied. Nodule identification, where the correctly recognized nodules are highlighted on the original CT scans, marks the pipeline's conclusion. Radiologists are then shown these outputs for diagnostic and clinical assessment. To sum up, this diagram shows a methodical and effective way to use machine learning for automated lung nodule diagnosis. From scanning to final detection, every stage is essential for maintaining accuracy, lowering false positives, and assisting radiologists in making an accurate and timely diagnosis of lung cancer. The incorporation of SVM improves the system's accuracy.
Since lung nodules are still one of the most prevalent malignancies that people worldwide encounter, early identification is essential for both male and female survival rates. The accuracy of lung nodule identification has significantly increased thanks to contemporary artificial intelligence (AI) and machine learning approaches. One of the finest strategies for accurately identifying breast cancer is the combination of Support Vector Machines (SVM) with enhanced segmentation techniques.
The invention provides an IoT-enabled framework for automated lung nodule detection using a Support Vector Machine classifier. The system begins with IoT end devices that capture medical imaging data, such as CT scan results, which are transmitted to gateway devices. These gateways, which may include computers or smartphones, forward data to fog worker nodes or cloud resources for further analysis.
Fog worker nodes act as intermediate processors, reducing latency by handling initial computations locally. When fog resources reach capacity, the system seamlessly directs computational tasks to cloud data centers, ensuring scalability and continuous performance. A cloud controller oversees workload distribution and resource allocation, while a service director ensures deployment of appropriate infrastructure. A protection supervisor manages authentication and data security during transmission between devices, fog, and cloud resources.
Data preprocessing includes noise removal, normalization, and image enhancement to standardize CT scan quality. Lung segmentation techniques isolate lung regions from surrounding anatomy using thresholding and morphological operations. This reduces computational complexity and ensures that classification is focused on relevant regions. Candidate regions are then extracted based on intensity, shape, and texture features.
The Support Vector Machine algorithm is applied to classify candidate regions. Trained with labeled datasets, it distinguishes true nodules from false candidates. The classifier is capable of handling both linear and non-linear data distributions through the use of kernel functions. To further reduce false positives, a refinement step eliminates non-nodular structures misclassified during initial detection.
Correctly classified nodules are highlighted on the original CT images and presented to radiologists for diagnostic evaluation. The pipeline thus combines automated machine learning with human verification for reliable outcomes.
The invention employs iFogSim and FogBus frameworks during development to simulate fog and IoT environments. This enables testing of metrics such as latency, bandwidth utilization, energy consumption, and cost-effectiveness. By modeling these parameters, the system ensures performance across varied clinical scenarios.
The distributed nature of the system ensures that computations can be performed at multiple layers. In one configuration, fog nodes alone perform detection for smaller workloads. In another, both fog and cloud resources collaborate for large datasets. Alternatively, cloud-only processing may be used when fog resources are unavailable. This flexibility ensures resilience and scalability.
The invention provides secure communications across devices and nodes. Authentication credentials are validated by protection supervisors, while secure protocols ensure data confidentiality during transmission. This is critical in medical applications where sensitive patient information is processed.
By employing Support Vector Machine models, the invention achieves reliable classification with smaller, well-preprocessed datasets, reducing reliance on massive annotated datasets that conventional deep learning models demand. This increases applicability in underfunded healthcare settings.
The system improves diagnostic precision by reducing false positives that typically arise when blood vessels or artifacts are mistaken for nodules. It also enhances interpretability by presenting clear visual outputs to radiologists.
The architecture is adaptable for real-time applications, enabling faster decision-making by medical experts. Integration with PACS systems in hospitals ensures seamless incorporation into existing clinical workflows.
Furthermore, the invention is cost-effective, as fog computing reduces dependence on high-performance centralized servers. This makes the system suitable for healthcare centers with limited resources while still providing advanced diagnostic support.
Overall, the invention integrates IoT connectivity, distributed computing, secure communication, and machine learning to provide a comprehensive solution for early and accurate lung nodule detection.
BEST METHOD OF WORKING
The best method of working involves deploying the system in a hospital environment where CT scans are routinely performed. Imaging data is captured through IoT-enabled devices and transmitted to gateway nodes, which forward it to fog worker nodes for initial preprocessing. The Support Vector Machine classifier, installed on fog or cloud servers, processes candidate regions to identify nodules. If fog resources are overloaded, the system automatically transfers tasks to cloud data centers under the supervision of a cloud controller. Results are securely transmitted back to hospital PACS systems, where identified nodules are highlighted on original scans for radiologist review. This configuration provides real-time detection, reduces false positives, and ensures accurate and efficient clinical decision-making.
, Claims:1. A system for automated lung nodule detection comprising:
IoT-enabled end devices configured to capture medical imaging data;
gateway devices configured to transmit imaging data to computational nodes;
fog worker nodes configured to preprocess imaging data and perform initial classification;
cloud data center nodes configured to handle large-scale computational tasks;
a cloud controller configured to manage workload distribution and scalability;
a protection supervisor configured to authenticate devices and secure communication;
a service director configured to allocate computing resources; and
a Support Vector Machine classifier configured to detect and classify lung nodules from preprocessed medical images.
2. The system as claimed in claim 1, wherein the fog worker nodes and cloud nodes collaboratively process medical imaging data based on workload requirements.
3. The system as claimed in claim 1, wherein the Support Vector Machine classifier employs kernel functions for classifying linear and non-linear data distributions.
4. The system as claimed in claim 1, wherein the preprocessing module includes noise reduction, normalization, lung segmentation, and enhancement of contrast.
5. The system as claimed in claim 1, wherein the protection supervisor validates gateway authentication credentials and ensures secure transmission.
6. The system as claimed in claim 1, wherein the service director deploys fog and cloud infrastructure in response to computational demands.
7. The system as claimed in claim 1, wherein the classifier includes a false positive reduction module to eliminate non-nodular regions misclassified as nodules.
8. The system as claimed in claim 1, wherein the output highlights detected nodules on original medical scans for radiologist review.
9. A method for automated lung nodule detection comprising:
capturing medical imaging data using IoT-enabled end devices;
transmitting the data through gateway devices to fog or cloud computational nodes;
preprocessing the imaging data using normalization, noise reduction, and lung segmentation;
applying a Support Vector Machine classifier to detect and classify nodules;
reducing false positives through post-classification refinement; and
presenting identified nodules on original medical scans for radiologist evaluation.
10. The method as claimed in claim 9, wherein workload distribution is managed between fog and cloud nodes under supervision of a cloud controller to ensure scalability and efficiency.