Neurotechnology and Rehabilitation: Advances in neural devices and brain-machine interfaces for post-stroke rehabilitation

• DOI: https://doi.org/10.22533/at.ed.1595182510049

• Palavras-chave: Brain-Computer Interfaces; Neuromorphic Computing; Stroke;

• Keywords: Brain-Computer Interfaces; Neuromorphic Computing; Stroke;

• Abstract:

  Brain-computer interfaces (BCI) that allow people with severe motor disabilities to use their brain signals to directly control objects have attracted growing interest in rehabilitation. BCIs have been used to help regain lost motor control in a limb after a stroke. While neuromorphic computing (NC) uses neural models in hardware and software to replicate brain-like behaviors, it could help usher in a new era of medicine by providing low-power, low-latency, small-size solutions that are significant for neurorehabilitation. Stroke is the most common severe manifestation of cerebrovascular disease, therefore characterized by a neurological deficit of sudden onset, stroke is predominantly caused by cerebral ischemia, due to atherosclerosis of large arteries, cardioembolism, or affection of small vessels and, less frequently, by intracerebral hemorrhage. The aim of the article is to carry out a literature review on updates in stroke rehabilitation and the association with BCI and NC and efficacy. This original article constitutes a bibliographic review, with several original articles, which were observed in scientific databases such as Latin American and Caribbean Literature in Health Sciences (LILACS), Scientific Electronic Library Online (SCIELO), National Library of Medicine (NIH), Nature, MEDLINE. Post-stroke cognitive impairment and dementia (PSCID) is a major source of morbidity and mortality after stroke worldwide. PSCID occurs as a consequence of ischemic stroke (ISC), intracerebral hemorrhage (ICH) or subarachnoid hemorrhage. Cognitive impairment and dementia that manifest after a clinical stroke are categorized as vascular. BCIs based on motor imagery are widely used in rehabilitation training for stroke patients. By using motor imagery, patients can be trained to gain control over their brain signals, allowing them to activate devices that assist movement. This training approach is believed to enhance sensory inputs, leading to brain plasticity that improves motor function. While NC presents a promising solution by mimicking biological synaptic activity, offering an alternative to conventional architectures. Unlike von Neumann computers, which operate in sequential execution, neuromorphic systems
  integrate processing and memory in networks of neurons and artificial synapses, allowing for efficient, event-driven computing. This architecture allows for lower power consumption, real-time adaptability and enhanced processing of neural signals, making it a strong candidate for improving BCIs. It can be concluded that despite these advantages, large-scale experimental validation of neuromorphic integrated BCIs remains limited, and further research is needed to address challenges such as hardware constraints, signal fidelity and real-world implementation.