Találati lista | Tudóstér

001-es BibID:	BIBFORM115556
035-os BibID:	(Scopus)85180022388
Első szerző:	Oniga István László (villamosmérnök)
Cím:	Implementation considerations for an intelligent embedded e-Health system and experimental results for EEG-based activity recognition / Stefan Oniga, Iuliu Alexandru Pap, Tamas Majoros
Dátum:	2023
Megjegyzések:	The development of eHealth systems has seen a rise in the past years and eHealth applications are becoming more widespread while improving many healthcare fields such as remote patient monitoring, interactive patient care and various predictions and detections of specific health related problems. This chapter presents the decisions, challenges and solutions that might arise when implementing an eHealth data acquisition system. An advantageous solution for the implementation of an embedded system using a single board computer for the acquisition of biomedical signals from blood pressure sensors, pulse oximeter, airflow sensor, galvanic skin response sensor, temperature sensor, EEG helmet, etc. is presented in this chapter. The acquired data can be transmitted remotely, stored locally or in the cloud, processed and analyzed. The automatic processing and interpretation of the acquired data in order to recognize the health state and detect the deviation from normal patterns can be achieved using shallow and deep neural networks. During the application of automatic evaluation, several problems and questions arise that need to be solved. One of these is what preprocessing of the available data is needed. Another question is necessity of extracting features from the data, and if so, what features they should be. After that, one of the countless machine learning algorithms suitable for solving the task must be chosen. At this step of the procedure, other constraints, such as resource requirements and speed must be considered. Finally, the architecture and parameters of the applied method must be decided. This chapter presents an example of automatic recognition of motor imagery movement activities based on EEG signals using convolutional neural networks. The factors that influence the recognition rate of some activities are presented and four different convolutional neural networks architecture were compared from efficiency point of view.
ISBN:	9781032404172
Tárgyszavak:	Műszaki tudományok Informatikai tudományok könyvfejezet könyvrészlet eHealth Electroencephalography (EEG) Remote patient monitoring Machine learning Brain computer interface Activity recognition
Megjelenés:	Smart Embedded Systems: Advances and Applications / Arun Sinha, Abhishek Sharma, Luiz Alberto Pasini Melek, Daniele Caviglia. - p. 33-55. -
További szerzők:	Pap, Iuliu Alexandru Majoros Tamás (1991-) (mérnök, informatikus)
Internet cím:	Szerző által megadott URL DOI Intézményi repozitóriumban (DEA) tárolt változat
Borító:
Saját polcon: