学术文献库

In today's world, many patients with cognitive impairments and motor dysfunction seek the attention of experts to perform specific conventional therapies to improve their situation. However, due to a lack of neurorehabilitation professionals, patients suffer from severe effects that worsen their condition. In this paper, we present a technological approach for a novel robotic neurorehabilitation training system. It relies on a combination of a rehabilitation device, signal classification methods, supervised machine learning models for training adaptation, training exercises, and socially interactive agents as a user interface. Together with a professional, the system can be trained towards the patient's specific needs. Furthermore, after a training phase, patients are enabled to train independently at home without the assistance of a physical therapist with a socially interactive agent in the role of a coaching assistant.