Indigenous Development of EEG Headset for Detection and Prevention of Epileptic Seizures

Project Title

Indigenous Development of EEG Headset for Detection and Prevention of Epileptic Seizures

Project Area of Specialization Electrical/Electronic EngineeringProject Summary

Epilepsy is a serious neurological disorder that affects around 50 million people worldwide. Despite technical advancements in current medical and diagnostic technology, there is a need for an effective seizure forecasting tool, particularly in nations such as Pakistan. This project aims to design, develop, and fabricate a novel low-cost headset and drug injection system for the diagnosis of epilepsy patients. The proposed hardware and computational tools aid in the early-stage identification, classification, and treatment of epileptic seizures based on abnormal neural activity. The overall tentative cost of the project is PKR 140,000, which, when completed, shall contribute to the noble cause of easing the lives of epilepsy patients and assisting pathologists with an improved diagnosis and on-time medical treatment. 

Project Objectives

Problem Statement:
Despite advancement in the technological area of modern medical and diagnostic devices, there is a need for a resourceful tool for seizure forecasting especially in developing countries like Pakistan. Traditional epilepsy detection is completed by neurosurgeons according to their own clinical experience by observing the electroencephalogram (EEG). This method not only takes a lot of time but also depends on the subjective judgment of neurosurgeons. Therefore, the realization of automatic high performance epilepsy detection is the main research direction for scholars. Also, commercially available headsets for EEG Headsets is around PKR 159,000 for EMOTIV EPOC 14-Channel Mobile EEG [1]. Such wearable multichannel devices are very costly due to the import taxes associated with them because of little to no local research in the field, rendering Pakistan unable to produce them locally.

Goals and Objectives
To solve the problems stated above, we set the following project goals,
i. To design and develop a low-cost sensor-based multi-channel headset capable of reading EEG signals from the brain.

ii. To automatically predict epileptic seizures based on abnormal neural activity at least 10 seconds prior to the occurrence of an epilepsy attack. The solution is aimed to assist the neurologists in making valid diagnosis while timely informing the medical staff when a seizure is likely to occur, thus adding to the Hospital Management System.

iii. To design and develop a method for automatic drug injection, aimed for quicker response to emergency situations.

Project Implementation Method

Designing of EEG Headset: The hybrid sensor is placed on the head of patient through head mask made up of rubber. Then the sensor is connected to to EEG amplifier. The EEG amplifier is the part of the data acquisition system responsible for accommodating, amplifying, and converting the analog electrical signals from the sensor into a digital signal that can be processed by the computer. EEG amplifier integrated with the Low pass filter. Low pass filter used to reduce the noise of a sensor reading coming from a electrode. Then the signal is passed with microcontroller where they interfere with the epileptic data of the patients.

Epilepsy Detection System: there are several steps for epileptic seizure detection from EEG analysis such as data acquisition, data pre-processing, feature extraction, model selection, classifier training and performance analysis/decision making.
The collected EEG signals are amplified, digitized, and then sent to a computer or mobile device for storage and data processing. Pre-processing in case of EEG data is applied to remove noise and artifacts from the data to get closer to the true neural signals. The pre-processed signal is sent to a feature extractor, whose purpose is to reduce the data by measuring certain “features” or “properties.” These features (or, more precisely, the values of these features) are then passed to a classifier that evaluates the evidence presented and makes a final decision as to the species. Then, a suite of different models such as Support Vector Machine (SVM), Quadratic Discriminant Classifier (QDC), Linear Discriminant Classifier (LDC) etc. are fit and evaluated on the problem, i.e. the model selection. Finally, we train our classifier with a 70:30 test-train split data, and then perform model assessment or classifier evaluation.

Epilepsy Prevention System: Status epilepticus requires quick medical intervention via drug injection through IV (Intravenous) Push method. For situations where IV access is not possible, Intramuscular (IM) method maybe used, but is not preferable due to slow response. Considering the mentioned 2 cases, two solutions can be taken into consideration. Our solution is for case 1 only.
The plan is to connect the processor that processes the EEG signals mentioned in the Epilepsy Detection System heading, to a relevant drug-filled capsule attached with an actuator and another regulator on the drip set which is set up on a hospitalized patient. This component is placed in between the regulator and injection side, on the tubing. The component on actuation, stops the flow of fluid from the volume chamber, and allows the drug filled in the capsule to flow. The actuator pushes the drug with carefully selected pressure into the IV-injection site.

Benefits of the Project

Epilepsy patients experience challenges in daily life due to precautions they have to take in order to cope with this condition. When a seizure occurs, it might cause injuries or endanger the life of the patients or others, especially when they are using heavy machinery or driving. The project benefits the epilepsy patients directly and the family as well as medical staff concerned with the subjects indirectly. With the help of the proposed model, epilepsy-affected patients will get more time for proper medication required for preventing the seizure before it actually occurs. Therefore, the proposed project is of great importance.

Nearly one-third of patients with epilepsy continue to have seizures despite optimal medication management. The proposed project has the potential to improve outcomes in these patients by allowing more tailored therapies and might, additionally, have a role in accident and Sudden Unexpected Death in Epilepsy (SUDEP) prevention. 

The proposed project will allow the detection of seizures prior to their clinical onset. Furthermore, this project might be used in accident prevention and seizure tracking and could further be useful in closed loops to facilitate seizure abortion. Beyond its use in immediate patient care, the project may allow for increased granularity of neuroepidemiologic data, thereby permitting improved seizure prediction and risk factor assessment.

Technical Details of Final Deliverable

Final deliverables involve the prototype including the EEG headset device which via processing of incoming signals from the headset, detect the epilepsy attack and inject the anti-epileptic medicine to the patient.

EEG Headset and Signal cleaning via Hardware:

EEG headset consists of the integration of hardware and software. Firstly, the patient is allowed to place a rubber-based EEG headset mask on their head which supports the hybrid electrode sensor. Sahara Hybrid electrode is used for the purpose, which is a high quality hybrid electrode sensor and deals with acquiring brain signals. Amplification is done inside the electrode clip which is connected with the sensor.  The electrode clip cable requires 1.5 kV protection for the arrival of brain signal to filters. Then brain signal is passed through the amplifier circuit which amplifies the signal. Low pass filter and noise filter increases the strength of the signal while removing noise, such as artifacts, machine noise, etc.

Signal Processing:

After signal acquisition and noise cleaning, the fine signal then passes through the EEG device where the signal is processed by a microcontroller which interlinks with the Matlab coding. For this we used a 1.5GHz quad-core 64-bit ARM Cortex-A72 CPU (Raspberry pi 4). The latest version of MATLAB is used i.e., R 2022a. The code identifies the disorder brain signal by supervised machine learning algorithms (Linear Discrimination Classifier LDC, Quadratic Discriminant Classifier QDC, Support Vector Machine SVM). 

Drug Injection System:

The epileptic prevention system relates to a micro controller, which injects the specific anti-epileptic drug in the patient’s body via IV route. The system involves a simple drip regulator which is mechanical in function, a low-torque mini-motor which moves the regulator’s dial when the microcontroller signals the motor on detection of seizure, and a mini-linear actuator (Xeryon XLA-3 or JC35-N), which will push the drug from a full-capsule into the IV-line. The mechanical regulator prevents the flow of fluid from volume chamber (fluid which is being infused to the patient already), and allows the anti-epileptic drug being pushed by actuator.

Final Deliverable of the Project HW/SW integrated systemCore Industry HealthOther Industries Medical Core Technology Artificial Intelligence(AI)Other Technologies NeuroTech, Wearables and Implantables, OthersSustainable Development Goals Good Health and Well-Being for People, Industry, Innovation and Infrastructure, Life on Land, Partnerships to achieve the GoalRequired Resources

Item Name	Type	No. of Units	Per Unit Cost (in Rs)	Total (in Rs)
			Total in (Rs)	75210
Sahara Hybrid Electrodes	Equipment	4	5000	20000
Raspberry Pi 4B	Equipment	1	38000	38000
Low Pass Filter	Equipment	1	5210	5210
Noise Filters	Equipment	1	2000	2000
Head Cover, Injection, Wires	Miscellaneous	1	10000	10000