A 3D-printed  Fat-I BC-enabled prosthetic  arm: Control  based on brain neuronal data

Abstract: Neuroprosthetics enhance a variety of artificial devices that can beused to improve the quality of motor, sensory defects that emerge frombrain-injuries. Brain-Computer interface, as the name suggests,en compasses the interaction between brain and environment. The purpose of this thesis is to study the electrophysiological signals for controlling a prosthetic arm using fat intra-body communication(fat-IBC), in which the fat tissue inside the body is used as atransmission medium for low power microwaves. With the help of a standardized data-set of the recordings of UtahMicro-electrode Array, from BlackRock Microsystems, it was possible to study the different techniques of Spike Sorting which helped in mapping the recorded action potential to a neuron. In order to perform the published spike sorting process, signal analysis and machinelearning techniques were used to best quantize the neural signals. One objective of this thesis to determine the quality of isolation of neurons for which the spikes detection algorithm was implemented at acertain threshold level which was further classified using theclustering techniques to identify the putative neurons. The results of the classification of neurons indicated a drop in the isolationquality because of background noise or electrode drift. However, the framework used in this thesis could be used for future works with a different threshold value.