Low-Cost Internet of Things Platform for Epilepsy Monitoring Using Real-Time Electroencephalogram

Introduction

The urbanization and population growth in many developing countries are rising exponentially. Even in nations wherein the bulk of the populace lives in rural areas, the sources are focused on the cities. Thus, present patient care facilities cannot fulfill the demand of population needs and results in an imbalance in the healthcare demand and the health services. All nations face transportation and verbal exchange demanding situations, and a scarcity of physicians and other health specialists in rural and faraway areas. The effect on the health and health care services in rural and faraway areas is greatly influenced by confining investment and other sources are also restrained at the government level. As referred to above, in the growing countries, a huge percentage reside in poverty and confined opportunities, and sources to supply health care offerings. In many advanced countries, there’s a downward trend in investment and infrastructure to help health services in rural and faraway communities.

In cities where there are emergency departments in hospitals and ambulance facilities, the system is there to ‘save’ the life of people. In agrarian and remote areas, this cannot be taken conceded, or neglected. Access is the number one issue for rural health. Thus, people in agrarian and remote areas would be treated in their local environment. Along with the above-mentioned issue, one more issue is that people are facing the long queues and large distance travel to the hospitals that is very difficult in the older ages.

According to the World Health Organization, one billion people are affected by neurological dis- orders. The report shows that close to 50 million of them suffer from epilepsy and 24 million from Alzheimer’s and other dementias. It is estimated that over 6.8 million people die annually from neurological disorders. As the global population ages, neurological disorders will have an impact in both developed and developing countries. In the present scenario of COVID-19, it’s also risky to visit the doctors in the hospitals for those patients (Kaiser et al., 2021). Elderly people are also facing the difficulty in movements and Neurological disorders decease worsen the case (Memon et al., 2018).

Epilepsy affects over 50 million people worldwide and 25% of those affected cannot be completely controlled by current medical or surgical treatments (Zhang et al., 2018). Seizures are random, sudden, excessive and out of control neurological disturbances. This type of seizure includes spontaneous seizures, lasting from a few seconds to more than 2 minutes. An early seizure prediction system that predicts seizures minimize a person’s risk of death (Hassan and Subasi, 2016), Ulate-Campos et al. (2016).

The solution to the above problem is to provide the access to concentrated resources in their local environment. This can be done by the rapid development of Low-cost, Portable intelligent, and IoT- based healthcare systems. That can be deployed at a sub-center level that connects them through the information and connectivity infrastructures called the Internet of Healthcare Things (IoHT) systems. The proposed work in this paper focuses on Developing and Deploying a Low-cost (IoT) Internet of Things based Epilepsy Monitoring system that monitors and predicts Epilepsy in Real-time using EEG signals from remote locations that provide the counter-attack to resolve the issues raised above. The contribution of this work is itemized below.

• i)

  A machine-learning model designed and trained in Matlab that predicts epileptic cases from EEG signals.

• ii)

  Wireless channel was set up in ThingSpeak server for the uploading the result of the prediction from raspberry pi and NodeMCU 32S collects EEG signals that were finally sent to Matlab for analysis and feature extraction.

• iii)

  The wireless connected credit card size mini-controller the raspberry pi 3 was used for the deployment of the trained model for prediction as a portable standalone system.

• iv)

  This intelligent portable IoT-based system can be placed at a sub-center in the agrarian and remote areas that are connected to the resource centers in the cities.

• v)

  The data visualizations can be done on Android/IOS and Matlab desktop.