| Summary: | In this manuscript, the work conducted with passive and active brain-computer interfaces (BCI) within the Health 4.0 framework is reported. Such systems are feasible for a real-time personalized rehabilitation. No-tably, passive BCIs are studied for detecting the engage-ment of patients, while active BCIs allow for an alternative mean of communication and control. Machine learning was exploited to classify the EEG signals associated with subjects’ brain activity. Results show that up to 61.4% and 73.7% classification accuracy can be achieved in emotional and cognitive engagement detection respectively, while channel and training time reduction guarantees wearabil-ity of the system and less stress for the patient.
|
|---|
