Heine, Dean (2024) Wearable Seizure Detector with Smart Phone App for both In/Outpatients (Hardware System). [USQ Project]
Abstract
Epilepsy is a condition that affects 50 million people worldwide. It is a brain condition that is manifest by seizures. Seizures can affect partial areas or the whole of the brain. Seizures can manifest in a way that causes uncontrollable body movements or can manifest in a way that is not detectable by a casual observer. Typical seizures last for a duration of 1 - 3 minutes. Seizures of 5 minutes or more can have serious health consequences and be life threatening. These seizures require immediate intervention and medical attention is important.
This research study aims to investigate a portable wearable seizure detector for an epilepsy sufferer that can be worn by an in or outpatient. The seizure detector would be designed to detect a patient having a seizure with no outward evidence of an episode known as an absence seizure. The wearable device would be connected to a smartphone. An application on the smartphone would be used to alert any caretakers of the patient using the seizure detector device. The application is not a part of the design for this research.
The research investigated using the Texas Instruments ADS1299 bio-potential chip as the microcontroller for the wearable device. The ADS1299 chip has the functionality to perform an EEG service. EEG is a method that can be used to detect seizures in the brain. An EEG was found to be described as the most useful method to detect epilepsy. An EEG is a non-invasive procedure, so it is typically associated with being comfortable and safe.
The background literature research showed that there are different types of electrodes that can be used for EEG testing. Considering the advantages and disadvantages of each type it was decided that dry electrodes would be the most suitable to use with the wearable device mainly due to its ease of setup. It was found that there are diagrams of connections of electrodes to the scalp of the patient called montages. An internationally recognised pattern is called the 10/20 montage, and it is this diagram which will be referred to when conducting testing in this research. It was found that electrodes can be incorporated into a cap that can make setting up. Due to cost a cap was not considered for use in the experiments for this research. A cap would be recommended in the future as part of the final product.
The background research discovered that there are two main methods in implementing an EEG test. They are bipolar (differential) and referential testing and have a specific structure to how the electrodes are connected from the scalp to the EEG’s instrument channel amplifiers. These methods were used when testing with the ADS1299 chip. A connection scheme was found during the literature research for these two methods, and they were used for the experiments conducted.
The Texas Instruments ADS1299 chip is available in a development kit, and this was purchased to conduct the testing. Other hardware was also purchased to complete the accessories required to do the tests. There were compromises made due to the budget available. Software for the development kit is available from the Texas Instruments website and was downloaded and used in the tests.
Bipolar and referential tests were conducted using the development kit and the analysis from the tests were included in the report. Two parameters that were needed to be considered when conducting the experiments were the sample per second rate and the number of samples recorded. The development kit does not provide real time analysis, but data is captured for consideration after an acquisition of signals are taken. Numerous tests were conducted, and it was found the best results were obtained with a sample rate between 1 – 4kSPS for the bipolar tests. Other rates produced very little signal or were polluted with a lot of noise. The sample rate for the referential tests had no effect on obtaining a signal. Although some brain signals were measured due to the limitation of the budget the quality and performance of the ADS1299 could not be confirmed. It could only be confirmed that the ADS1299 is functional as an EEG device. It was noticed that the mains supply influenced all the tests. This could be filtered with post-processing of the raw data.
Due to resources and time the scope of the project limited creating a completed wearable seizure detector. Some areas to follow up were described at the end of the report. Further considerations explained would be the type of case used and how it would be worn. Testing various iterations of the wearable device would be needed. To communicate between the wearable device and the smartphone the main control board of the device would need to be paired with a communication board. The ADS1299 chip uses SPI as a communication protocol and research showed that this could be paired with several different wireless protocols. It was decided weighing the advantages and disadvantages of each the preferred protocol for the device would be BLE.
There are other considerations that would require further research. The device has only 8 channels and it would be needed to be determined if this was sufficient since the patient being monitored may have life threatening seizures. It would also need to be determined if there is only one configuration of the electrodes needed or if the configuration depends on a particular patient’s condition. Also, if the device is to be worn by the patient at home the healthcare practitioners would need to ensure the caretaker is competent and confident in using the system so training may be required. As the device has a medical function advice would need to be sought as to whether the device would need to be certified.
|
Statistics for this ePrint Item |
| Item Type: | USQ Project |
|---|---|
| Item Status: | Live Archive |
| Faculty/School / Institute/Centre: | Current – Faculty of Health, Engineering and Sciences - School of Engineering (1 Jan 2022 -) |
| Supervisors: | Wen, Paul |
| Qualification: | Bachelor of Engineering (Honours) (Electrical and Electronic) |
| Date Deposited: | 09 Mar 2026 03:50 |
| Last Modified: | 09 Mar 2026 03:50 |
| Uncontrolled Keywords: | epilepsy; seizure detector |
| URI: | https://sear.unisq.edu.au/id/eprint/53049 |
Actions (login required)
 |
Archive Repository Staff Only |
