Disruptive innovations in data management and analytics have led to the development of patient-centric Healthcare 4.0 from the hospital-centric Healthcare 3.0. This work presents an IoT-based monitoring systems for patients with cardiovascular abnormalities. IoT-enabled wearable ECG sensor module transmits the readings in real-time to the fog nodes/mobile app for continuous analysis. Deep learning/machine learning model automatically detect and makes prediction on the rhythmic anomalies in the data. The application alerts and notifies the physician and the patient of the rhythmic variations. Real-time detection aids in the early diagnosis of the impending heart condition in the patient and helps physicians clinically to make quick therapeutic decisions. The system is evaluated on the MIT-BIH arrhythmia dataset of ECG data and achieves an overall accuracy of 95.12% in classifying cardiac arrhythmia.
Top1. Introduction
The Healthcare 4.0 paradigm, at its core, involves providing highly personalized services to patients. For this to be realized in the healthcare industry, real time service is of the essence. Fog Computing enables computations to be performed at the edge, while simultaneously utilizing the cloud to store large amounts of data. Thus, it can be leveraged to meet the real-time servicing needs of Healthcare 4.0. We use an ECG to get the readings and use high performance computing (HPC) and ML algorithms for monitoring and analysis. Develop an efficient system based on the fog computing paradigm for the real-time monitoring and analysis of ECG data of users and decrease response time in case of emergencies. The Overall Objectives can be stated as
- ●
Accurate and Real-Time monitoring and analysis of ECG data.
- ●
Immediate notification to Doctors and Emergency Services in case of anomalies.
- ●
Performance enhancement of Healthcare 4.0 using Fog Computing, HPC and ML algorithms.
- ●
Decreasing the latency of monitoring and analysis by the utilization of fog architecture.
- ●
Creation of a mobile application for visualization of the data by doctors/patients.
The Deliverables can be achieved such as Machine Learning and Deep Learning Models for the analysis of ECG Data to detect anomalies, Fog Architecture on which the Model is deployed for decreased latency, a mobile application for patients and doctors to view the diagnosis and data.
The scope is limited to Detect anomalies in heartbeat based on ECG Reading, notify assigned doctors in case of emergency and/or when an anomaly is detected, Fog computation architecture for real time predictions and notifications, Mobile app for both doctor and patient to keep track of heart health history, prescriptions etc.
The motivation is to improve the efficiency of fog architecture for improved performance. Add support for tracking other health measures such as blood pressure etc. Continuous improvement of model accuracy, and addition of other models to predict health state based on various other factors such as blood pressure etc.
Top2. Literature Survey
The Healthcare 4.0 paradigm, at its core, involves providing highly personalized services to patients. For this to be realized in the healthcare industry, real time service is of the essence. Fog Computing enables computations to be performed at the edge, while simultaneously utilizing the cloud to store large amounts of data. Thus, it can be leveraged to meet the real-time servicing needs of Healthcare 4.0. Below, we give a brief history of the utilization of Fog Computing in the healthcare industry and how it can benefit our project. Several works on the prediction of heart anomalies based on ECG data are also explored.
(Kraemer et al., 2017) presents a survey on fog processing inside medical services informatics, and investigates, characterizes, and examines distinctive application use cases. Applications are ordered into utilization case classes and a stock of explicit tasks that can be taken care of by fog computing is recorded. The level of the organization such fog computing tasks can be executed are talked about and trade-offs as for prerequisites pertinent to medical care are given. The survey shows that: 1. there is a critical number of registering undertakings in medical care that require or can profit by fog computing standards 2. Handling on higher organization levels is needed because of requirements in remote gadgets and the need to total information 3. Protection concerns and constant forestall calculation tasks to be totally moved to the cloud these discoveries prove the requirement for an intelligible methodology toward fog computing in medical services, for which a rundown of suggested innovative work activities are introduced.