Description:FORM 2
THE PATENTS ACT, 1970
(39 of 1970)
&
THE PATENT RULES, 2003
Complete Specification
(See section 10 and rule 13)
1.Title of the Invention ADVANCE SECURE HEALTHCARE SYSTEM BASED ON BLOCKCHAIN TECHNOLOGY ENTANGLED WITH FEDERATED LEARNING TECHNIQUE
2. Applicants
Name
Nationality
Address
Divyashree D
Sheela D V
Manjunath B
Shuaib Ahmed Shariff
Reva University
Indian
Indian
Indian
Indian
Indian
Assistant Professor, School of CSA, Reva University, Bangalore, Karnataka, India.
Assistant Professor, School of CSA, Reva University, Bangalore, Karnataka, India.
Assistant Professor, School of CSA, Reva University, Bangalore, Karnataka, India.
Assistant Professor, School of CSA, Reva University, Bangalore, Karnataka, India.
Bangalore, Karnataka, India.
3. Preamble to the Description:
The following specification particularly describes the invention and the manner in which it is to be performed.
FIELD OF THE INVENTION
This invention Invention relates to the healthcare industry is focusing on in-home healthcare services, which allow patients to conduct medical examinations in the comfort of their own home.
BACKGROUND OF THE INVENTION
The IoMT is an emerging technology that is rapidly expanding. The Internet of Things (IoT) enables the connection of numerous objects to collect data that may be used to enhance human health, productivity, and effectiveness. Smart cities, smart grids, and smart houses are well-established concepts that are transforming our daily lives. Among the most potential new technical approaches for addressing the global health equality gap is the use of an IoT-based monitoring system of patient health. Another name for these IoT technologies is the IoMT. In this study, the terms IoT and IoMT are used interchangeably, even though the study will focus on the healthcare industry. Furthermore, the Internet of Things has the potential to improve healthcare and public safety significantly. Individuals can obtain information on their lifestyles, physical and mental efficiency, and living surroundings, among other things, by connecting their beings to the Internet. This allows healthcare providers to monitor people's health remotely and in real-time. Furthermore, the data gathered could be used to promote evidence-based interventions for diseases, trauma, protection, early detection, and rehabilitation. In today's world, transferring patients from their homes to hospitals for routine check-ups is extremely difficult. There are several challenges, including queuing, travel time, and the possibility of patients contracting various viruses while traveling through this polluted environment.
SUMMARY OF THE INVENTION
In recent years, IoMT-based intelligent healthcare has been widely implemented to maximize the utilization of medical data to improve the accuracy of disease forecasting and healthcare therapy. With the rising amount and diversity of medical data, there is an urgent need for efficient data mining techniques to assess this information in order to facilitate disease identification, provide medical treatments, and improve patient care. Nonetheless, it confronts several difficulties, including patient privacy breaches and a variety of adversary threats to data transmission. In addition, the development of artificial intelligence and worldwide epidemics have hastened the adoption of intelligent healthcare while presenting problems
around information security, hostile cyberattacks, and service quality. A secure healthcare 5.0 system based on blockchain technology entangled with federated learning techniques is being explored to increase predictive performance. Numerous analytical frameworks have been employed to determine the viability of this particular argument. The proposed RTS-DELM method is extremely effective. The proposed technique has a rate of accuracy of 93.22 percent for disease prediction and 96.18 percent for intrusion detection. Furthermore, it is acknowledged that developing a basic method is less expensive and faster. We are pleased with the preliminary findings and intend to expand our research by evaluating additional sets of data in the future. The proposed system's computational complexity is limited by the growing number of hidden layers. Future research will aim to identify and quantify the factors with greater precision. To optimize the performance of various configurations, the learning system will be retrained on a more frequent basis.
BRIEF DESCRIPTION OF THE DRAWINGS
Fig.1: depicts the Block chain Structure.
Fig.2: depicts the A Federated Learning Technique.
Fig.3: depicts the A Proposed Federated Learning-based Healthcare Monitoring system for Healthcare 5.0.
Fig.4: depicts the NSL-KDD Dataset Structure.
Fig.5: depicts the Parkinson’s Disease Dataset Structure.
Table 1: depicts the Proposed RTS-DELM Entangled with Federated Learning Pseudo code (Server-side).
Table 2: depicts the Proposed Smart FML- RTS-DELM Pseudo code (Client-side).
BRIEF DESCRIPTION OF THE INVENTION
The IoMT is an emerging technology that is rapidly expanding. The Internet of Things (IoT) enables the connection of numerous objects to collect data that may be used to enhance human health, productivity, and effectiveness. Smart cities, smart grids, and smart houses are well-established concepts that are transforming our daily lives. Among the most potential new technical approaches for addressing the global health equality gap is the use of an IoT-based monitoring system of patient health [8]. Another name for these IoT technologies is the IoMT. In this study, the terms IoT and IoMT are used interchangeably, even though the study will focus on the healthcare industry. Furthermore, the Internet of Things has the potential to improve healthcare and public safety significantly. Individuals can obtain information on their lifestyles, physical and mental efficiency, and living surroundings, among other things, by connecting their beings to the Internet. This allows healthcare providers to monitor people's health remotely and in real-time. Furthermore, the data gathered could be used to promote evidence-based interventions for diseases, trauma, protection, early detection, and rehabilitation. In today's world, transferring patients from their homes to hospitals for routine check-ups is extremely difficult. There are several challenges, including queuing, travel time, and the possibility of patients contracting various viruses while traveling through this polluted environment. As a result, the healthcare industry is focusing on in-home healthcare services, which allow patients to conduct medical examinations in the comfort of their own homes. A smart health monitoring system is designed to help patients who live in remote areas contact doctors in urban areas. This technology acts as a bridge between patients and clinicians. It keeps track of vital signs like heart rate, electrocardiogram (ECG), blood pressure, temperature, and whether or not a person has fallen. The system collects this data and sends it to the application for further analysis via a wireless connection. Effective mining algorithms are urgently needed to analyse medical information to help in disease discovery, offer medical treatment, and enhance patient care. Machine learning is a sophisticated computational technique that has been used in a variety of domains such as image recognition, language processing, and health care. Nonetheless, machine learning models acquire great accuracy only with a vast quantity of training set, which is crucial in healthcare, where precision may sometimes mean the difference between saving or losing a patient's life. In most cases, centralized training strategies include acquiring a big volume of information from a robust cloud server, which might lead to major consumer privacy violations, especially in the medical field. As an open and accountable data protection mechanism, the
development of blockchain technology paves the way for new ways to address key issues of privacy, security, and ethics in a smart healthcare system. However, blockchain has achieved excellent success as a backbone of cybersecurity architecture for a variety of smart healthcare technologies, such as the control of patient record access, information distribution, etc. It is justifiable to incorporate blockchain in intelligent domestic networks as it is autonomous of heterogeneous protocols that are often used in smart systems. However, notwithstanding increased interest in smart healthcare technology, current work is distributed across diverse fields of study. Such a timely study is also being undertaken to close the void and have practical insights on blockchain technology, techniques, and their application in the field of the intelligent healthcare system. The IoMT has generated security flaws in the healthcare system. Unified networks enable hackers to reach the linked devices for malicious purposes. The big issue is hack attempts on medical records and healthcare devices; not only are they the key components of a smart healthcare system but they can also be used for malicious purposes: delivering phishing and spam mail. Because of unencrypted wireless keys, smart healthcare devices are often the central target for DDoS attacks, particularly because they are immediately turned on to provide smarter solutions such as patient medical records and the automatic updating of false medical reports. These challenges originate from a centralized IoMT system structure, and privacy issues are growing as the IoMT revolution progresses, namely record forgery and manipulation, device interference, and infiltration of illegal IoMT devices via attacks against server and gateway networks.
These issues can be addressed through the deployment of blockchain-based systems and unified "cloud- like" computer networks. Blockchain architecture is illustrated in Figure 1. It is a series of blocks that are connected with simple cryptography. Inflexibility, decentralization, and transparency are the three main concepts in the functioning of Blockchain technologies. The three responsibilities have been highly productive and opened their doors to a broad array of technologies relevant to the virtual currency, such as the functioning of autonomous vehicles, such as smartphones, and embedded devices. Whereas Blockchain technology is safe and enables anonymity, there are still certain drawbacks to it in the current stage of deployment is a decentralized machine learning platform for the IoT that enables numerous devices to collaboratively learn machine learning models without exchanging their actual data. Figure 2 depicts FL's architecture. This enhances the intelligent healthcare system by avoiding the leaking of patient information. Figure 2 shows a federated learning-based healthcare system in which embedded sensors gather medical information
from healthcare providers, multiple edge devices collaborate on federated learning algorithms, and machine learning techniques assess the patient's well-being and, if needed, seek immediate assistance in the cloud. Federated learning is a recently popular paradigm due to the incredible assurance it provides for studying fragmented sensitive information. Rather than combining data from disparate sources or relying on the traditional find-then-replicate strategy, it enables training of a common global model on a centralized server while retaining data in the relevant organizations. This method, known as federated learning, allows individual locations to work together to train a global model. Federated learning is the process of pooling training data from multiple sources to develop a global model without directly exchanging datasets. This ensures that patient privacy is maintained across sites.
Healthcare 5.0 is a network architecture that requires fifth-generation communication as the core network architecture for linking healthcare equipment. IoT will create data that AI can utilize, contributing to the advancement of digital wellbeing by concentrating not just on patient well-being and quality of life, but also on the well-being and quality of life of people worldwide. The primary issues in healthcare 4.0 are flawless information transfer with little or no information leakage. Automation and AI are two emerging healthcare 5.0 technologies that have the potential to transform any type of employment. Intelligence in healthcare 5.0 encompasses ideas such as accurate and systematic disease diagnosis, virtual patient monitoring and detection, remote surgery, and intelligent treatment, that comprises online training for patients suffering from anxiety. Artificial Intelligence (AI) is a broad concept that encompasses all intelligent advancements. It refers to the ability of machine-learning methods to predict final results without the involvement of humans. As a result of the tremendous expansion and advancement of technology in healthcare, the term "smart health system" has been coined. In light of these capabilities, numerous solutions have been proposed for use in numerous industries, including smart homes, Industrial Internet of Things (IoT), and smart healthcare. However, as privacy threats become increasingly sophisticated, there are still a number of obstacles to overcome when implementing blockchain-based federated learning in healthcare 5.0:
We design an Intrusion Detection system (IDS) in healthcare 5.0 system that enhances the security and privacy by detecting intrusions and attack patterns.
The proposed method secures the FL-based healthcare 5.0 Network by accurately assessing its reliability with regard to the most important security goals of secrecy, validity, and
accessibility. To support the claim that the overhead caused by the proposed technique is minimal compared to the value of its security and privacy, the study assesses the suggested method's capacity to securely protect sensitive data while consuming quite minimal resources. The purpose of this study is to examine a system model based on a Real-Time Deep Extreme Learning System (RTS-DELM) for intelligent disease prediction and intrusion detection in healthcare
5.0 that achieve the maximum possible level of accuracy. The remaining portions of this research study can be broken down into the following categories. Section 2 offers an overview of relevant research. Section 3 enlightens the proposed methodology. Section 4 presents the simulation and findings of the suggested method. Finally, section 5 discusses the study's findings.
BLOCKCHAIN MODULE IMPLEMENTATION
A blockchain was initially designed about and developed by Satoshi Nakamoto in the year. The block comprises a vast amount of transaction information, including the block id, a hash of the previous block, transaction details, nonce, and time stamps. In a blockchain scheme, where miners determine the correct hash to add a block, the winners would first search the existing block before searching for a new block. The proof of work methodology is used to verify whether or not a given block of transactions is legitimate. The steps listed below describe the core components of blockchain technology. In a smart healthcare system, any node that is connected to the Internet must communicate with a repository of storage data, along with miners in a blockchain framework. The blockchain holds all unprocessed transactions pending an opening to a new block to validate it. Many of the transactions are first reviewed and then shortly analyzed by the Merkle tree. The new smart healthcare system connectivity ecosystem will be created by blockchain technology, as it is versatile and compliant with healthcare IoMT applications.
RTS-DELM MODULE IMPLEMENTATION
RTS-DELM is a data analytics platform that automates data analysis tasks and provides insightful information. A RTS-DELM approach analyses real-time data using the Deep Extreme Learning Machine (DELM). The DELM can be used to determine energy consumption levels, inventory services, and specify transportation operations, to name a few applications [40]. The RTS-DELM method can be used to modify datasets in healthcare
networks, implying that any errors can be excluded. It employs a novel approach to disease prediction and diagnosis. The goal of this study is to evaluate a system model based on the RTS- DELM for the most accurate adaptive forecasting of healthcare monitoring.
FEDERATED LEARNING MODULE IMPLEMENTATION
Federated learning is a recently popular paradigm due to the incredible assurance it provides for learning with fragmented sensitive data. Rather than combining data from multiple sources or relying on the traditional find-then-replicate strategy, it allows for the training of a common global model on a central server while retaining data in the appropriate organizations. Federated learning, as it is known, allows individual sites to contribute to the training of a global model. Federated learning is the process of combining training data from multiple sources to build a global model without directly sharing datasets. This ensures that patient privacy is preserved across locations.
Several firms or institutions collaborate to solve a machine-learning problem using federated learning, which is managed by a central server or service provider. As a result, a deep learning model is hosted and optimized by a central server. The model is trained by dispersing itself across remote centralized datacentres, which may include hospitals or other medical organizations, preserving data localization at these locations. Throughout the training process, no data from any participant is exchanged or transferred. Rather than sending data to a single server, as in traditional deep learning, the server maintains a globally shared architecture that is accessible to all institutions. Following that, each organization develops its model based on patient data. Following that, each institute communicates with the server using the error gradient of the model. The central server collects all participant feedback and modifies the global model according to predefined criteria. The predefined criteria allow the model to evaluate the excellence of the response and thus include only information that adds value. As a result, feedback from institutions reporting poor or unusual results may be overlooked. This method creates a single federated learning cycle that is repeated until the global model is acquired.
DATA FUSION MODULE IMPLEMENTATION
Data fusion techniques combine information from multiple sensors to produce more precise observations than a single, independent sensor could. The process of obtaining data from disparate but likely linked sources and combining it to have the greatest impact is known as
information extraction. A security framework includes several network security sensors, and it is challenging to gain a comprehensive, wide-angle perspective of the security system's dynamic security situation. Additionally, equipment spread across a vast region may be challenging to handle effectively. In order to enhance the model's efficacy and give a comprehensive analysis of the system's protection situation, it is also essential to efficiently and smartly combine the sensor data. In comparison, because the information comes from multiple sources, multiple data sources can provide a higher level of consistency in terms of trustworthiness.
Sensors contribute to the Internet of Things by serving as critical tools for evaluating smart healthcare systems, ecosystems, and consumers. The following devices fall under this category: Healthcare, Cameras, and Interactivity are just a few of the terms that come to mind when thinking about Computers processing the data collected by sensors. For example, in conjunction with the healthcare system, the heart monitor sensor regulates the cardiac rate. The application layer is a component of the IoMT network system that includes closed-circuit devices, wearables, and so on.
INTRUSION DETECTION SYSTEM (IDS) MODULE IMPLEMENTATION
A comprehensive intrusion detection system is necessary to fully assess the healthcare system. Any type of data can be analyzed using the machine learning method termed RTS-DELM.
The autonomous dataflow architecture used by this machine learning software allows it to track the information flow and identify trends of infiltration and assault. It's crucial to develop robust and adaptable algorithms to manage the continuously evolving smart blockchain-based applications. The suggested strategy provides a safe smart healthcare network design that would solve both present issues with centralized security of healthcare networks and potential future threats. In the work described here, a RTS-DELM technique will be employed to create a smarter, safer healthcare system employing sensors driven by the IoMT that are more effective as shown in Figure 3. The main contributions of this study are a thorough analysis of scientific advancements feasible to blockchain-based healthcare 5.0 systems powered by RTS-DELM, a fresh perspective on various employments (like data exchange between healthcare systems), and assistance from the most recent stages of technological development.
DATASET DESCRIPTION
For disease prediction analysis, we employed the publicly accessible Parkinson's disease dataset, and for intrusion detection, we employed the publicly available NSL-KDD dataset. Error! Reference source not found. Lists all protection protocols falling under their identity group. The NSL-KDD dataset is a refined edition of KDD 99 that includes several enhancements related to the original KDD 99 data collection. The NSL-KDD data collection contains 41 functions for each record. Error! Reference source not found. offers a proper description of the functions. This study includes 195 prolonged vowel phonations from 31 individuals, 23 of whom were diagnosed with Parkinson's disease. The primary objective of data processing is to distinguish between healthy individuals and those with Parkinson's disease, based on the "status" attribute, which is set to non-PD for healthy individuals and PD for those with Parkinson's disease; this is a two-decision classification issue. Error! Reference source not found. depicts the attributes of the data set.
WORKING OF THE PROPOSED MODEL
RTS-DELM computational technology may be used to make federated learning-based systems more intelligent. When using the RTS-DELM distributed blockchain technology, it is possible to increase the level of data secrecy. Additionally, by transferring new information and accelerating comprehension, RTS-DELM may be utilized to enhance understanding. It provides the platform and network architecture for the development of a decentralized blockchain application. The deployment architecture of the advanced system RTS-DELM is examined in this paper. Utilizing sensors, mobile devices, and IoMT systems as sources of information are the right way to employ this technology to gather intelligence. These strategies produce knowledge that is employed in smart applications. Despite this, real-time data analysis uses the RTS-DELM approach to assess and make predictions.
The creation of data for study reduces data mistakes such as repetition, missing information factors, malfunctions, and interference. When a little fraction of a data collection is all that is needed, the RTS-DELM method performs effectively. The architecture is capable of supporting a broad variety of applications in several fields, including fraud detection and/or prevention. As shown in Figure 3, the proposed RTS-DELM system employs a large number of hidden layers, hidden neurons, and various activating mechanisms to optimize the healthcare monitoring system. Data capture, preparation, and assessment are the three separate phases that make up the suggested approach for breaking down data analysis.
Prediction and performance make up the two sub-layers that make up the evaluation layer. Accurate data is gathered from sensors and actuators for analysis. The collecting layer uses the data that is provided as raw data. To eliminate inconsistencies in the preprocessing layer, a thorough approach for data cleaning and preparation is used. The foremost objectives of this approach are:
 To implement the RTS-DELM algorithm to build an optimal approach for healthcare monitoring at the client side.
 In this study, an intelligent algorithm is proposed to identify and monitor healthcare in patients and
 Identify the severity of the patients.
 To implement an Intrusion Detection System (IDS) to determine the flow of data to detect intrusions and attack patterns
To implement and improve IoMT in existing applications to sustain the medical standards. Furthermore, to increase the model's efficiency, a federated learning approach is implemented as shown in table 2 & table 3. Further evaluation is carried out using disease datasets and the freshly trained layer is applied to the pre-trained framework to increase the efficacy of the system.
 To preserve records of the local data to preserve the patient's privacy.
 To test the utility of a trained model by assessing it on a real-time dataset. with federated learning, the implementation of the prediction system is evaluated. In addition, model output is evaluated with other models of machine learning.
 To prevent humans from getting chronic diseases by the spreading of the disease.
 To estimate the performance of the proposed methodology by assessing different datasets.
 Furthermore, test and evaluate the datasets on different machine learning algorithms to justify the proposed approach.
This will provide a thorough analysis of technical advancements relevant to RTS-DELM-enabled federated learning-based systems to provide a fresh perspective on various implementations (such as healthcare data exchange). The main aim of this study is to build an intelligent algorithm proposed for healthcare monitoring among patients. The working of the proposed model is described as follows;
Hospitals identify and assign training tasks to a local model, which is subsequently uploaded to a centralized server, where it is disseminated to all IoMT devices as a global model. Additionally, the training phase comprises three layers: the sensing layer, the preprocessing layer, and the application layer. Medical data obtained by IoMT devices may contain missing or erroneous data. To reduce noisy data, the preprocessing layer uses moving average and normalization to address missing values. Clinical data is transmitted to the Application layer following preparation. Additionally, the Application Layer is separated into two sections: Prediction Layer and Performance Layer. Following the prediction layer, the results of the forecast layer are transmitted to the performance layer, which uses the accuracy and miss rate of the prediction layer to determine whether or not the learning requirements are met. In the case of ‘No’ the model will be retrained but in the case of ‘Yes’, the locally trained model is exported to the cloud as a global model.
Medical organizations receive a copy of a global model and train it on local data. Following that, each institution upgrades its model on the cloud without exchanging datasets directly. This assures the continuity of patient privacy between medical organizations.
The cloud aggregates all updated parameters provided by participating institutions to create a new global model, which is subsequently disseminated to all participating organizations. As opposed to outcomes produced from a single information source, the data fusion technique can give results that are more dependable and stable. Then the data requester launches a data sharing request to the centralized server. Upon receiving the request, the centralized server verifies the access. After authorizing access, the nodes train a global data model jointly by federated learning. Once the model is trained, the data requester gets the corresponding sharing results. The input layer parameters will be detected during the validation phase and forwarded to the evaluation phase for healthcare monitoring.

, Claims:CLAIMS
1. The in-blockchain features of the model can potentially be used by attackers to determine the original secret healthcare data.
2. To address the aforementioned problems, this study blends FL and sophisticated encryption to provide a safe and privacy-preserving healthcare 5.0 system. The following are the most significant contributions made by this study;
3. We present a blockchain-based FL framework for healthcare 5.0 that not only builds an accurate multidisciplinary model based on multiple edge devices, but also supervises the whole training process.
4. The proposed method provides an additional degree of security to blockchain-based FL, we present an RTS-DELM method that balances privacy and model accuracy by adjusting noise depending to the training process.
5. The proposed system considers multiple medical organizations in the proposed model because the locally trained model of other medical organizations may enhance the capability of the healthcare 5.0 system by sharing a global model.
6. To implemented the federated learning approach in healthcare 5.0 to improve the learning process of clinical data by locally training models.
7. To provide an intelligent hybrid approach to enhance secure communication and effective healthcare monitoring.
Dated this.
Senthil Kumar B
Agent for the applicant
IN/PA-1549