Abstract:

Epilepsy monitoring has been one of most recognisable applications of wearable sensors as it can be used to mitigate everyday risks of having seizures. This problem based on statistical data affects children, which from development point of view must be accurately diagnosed and effectively treated. Any exceptions, delays can affect the child's health but especially its development increasing the risk of future conjugate diseases. Described in the paper tool is an attempt to provide a mobile remotely accessible monitoring tool which will alarm, notify and help to report health events which than can be automatically classified as seizures. The algorithms and NERVE tool have been designed not only to identify seizure but also recognise specific four classes of generalised seizures (clonic, tonic, myoclonic, atonic) including evaluation of the intensity of the seizure. This paper discusses the usage of wearable class multi-sensors introducing surface electromyography supplemented with limbs movement and activity evaluation. These data sources need to be selected with regard to not only detection accuracy but also mobility and usability features, which make the measuring process so cumbersome and difficult. To identify a seizure the NERVE tool is using a wireless, attachable, small biomedical sensor capable of autonomous EMG and IMU processing on Arduino platform. Based on the algorithmic signal filtering and evaluation the system identifies, a health treat (an episode) and wirelessly notifies the parent or caregiver. The assistance process allows not only to record the biomedical data but also supplement the episode with video recording and postictal survey describing the health event. Having such detail electronic documentation, epileptologist is capable of identifying the etiology of the disease, proposing a treatment or verifying the efficiency of treatment. The mobile application provides a novel approach in case of child seizures support based also on classification of the seizure based on the survey, which is able to increase the method’s accuracy, shorten communication time. This work also demonstrates a quantitative approach to specific physical state evaluation and describes functionality of designed system in the domain of hardware and software components.