RESUMO
This study evaluates the performance of two convolutional neural networks (CNNs) in a brain-machine interface (BMI) based on motor imagery (MI) by using a small dataset collected from five participants wearing a lower-limb exoskeleton. To address the issue of limited data availability, transfer learning was employed by training models on EEG signals from other subjects and subsequently fine-tuning them to specific users. A combination of common spatial patterns (CSP) and linear discriminant analysis (LDA) was used as a benchmark for comparison. The study's primary aim is to examine the potential of CNNs and transfer learning in the development of an automatic neural classification system for a BMI based on MI to command a lower-limb exoskeleton that can be used by individuals without specialized training.Clinical Relevance- BMI can be used in rehabilitation for patients with motor impairment by using mental simulation of movement to activate robotic exoskeletons. This can promote neural plasticity and aid in recovery.
Assuntos
Interfaces Cérebro-Computador , Exoesqueleto Energizado , Humanos , Eletroencefalografia , Redes Neurais de Computação , Aprendizado de MáquinaRESUMO
Graphene derivatives such as reduced graphene oxide (rGO) are used as components of novel biomaterials for their unique electrical properties. Electrical conductivity is a crucial factor for muscle cells, which are electrically active. This study reports the development of a new type of semi-interpenetrated polymer network based on two biodegradable FDA-approved biomaterials, sodium alginate (SA) and polycaprolactone (PCL), with Ca2+ ions as SA crosslinker. Several drawbacks such as the low cell adhesion of SA and weak structural stability can be improved with the incorporation of PCL. Furthermore, this study demonstrates how this semi-IPN can be engineered with rGO nanosheets (0.5% and 2% wt/wt rGO nanosheets) to produce electroactive nanohybrid composite biomaterials. The study focuses on the microstructure and the enhancement of physical and biological properties of these advanced materials, including water sorption, surface wettability, thermal behavior and thermal degradation, mechanical properties, electrical conductivity, cell adhesion and myogenic differentiation. The results suggest the formation of a complex nano-network with different interactions between the components: bonds between SA chains induced by Ca2+ ions (egg-box model), links between rGO nanosheets and SA chains as well as between rGO nanosheets themselves through Ca2+ ions, and strong hydrogen bonding between rGO nanosheets and SA chains. The incorporation of rGO significantly increases the electrical conductivity of the nanohybrid hydrogels, with values in the range of muscle tissue. In vitro cultures with C2C12 murine myoblasts revealed that the conductive nanohybrid hydrogels are not cytotoxic and can greatly enhance myoblast adhesion and myogenic differentiation. These results indicate that these novel electroactive nanohybrid hydrogels have great potential for biomedical applications related to the regeneration of electroactive tissues, particularly in skeletal muscle tissue engineering.