Search for: All records

Abstract Objective.Spiking neural networks (SNNs) are powerful tools that are well suited for brain machine interfaces (BMI) due to their similarity to biological neural systems and computational efficiency. They have shown comparable accuracy to state-of-the-art methods, but current training methods require large amounts of memory, and they cannot be trained on a continuous input stream without pausing periodically to perform backpropagation. An ideal BMI should be capable training continuously without interruption to minimize disruption to the user and adapt to changing neural environments.Approach.We propose a continuous SNN weight update algorithm that can be trained to perform regression learning with no need for storing past spiking events in memory. As a result, the amount of memory needed for training is constant regardless of the input duration. We evaluate the accuracy of the network on recordings of neural data taken from the premotor cortex of a primate performing reaching tasks. Additionally, we evaluate the SNN in a simulated closed loop environment and observe its ability to adapt to sudden changes in the input neural structure.Main results.The continuous learning SNN achieves the same peak correlation ( $\rho =0.7$) as existing SNN training methods when trained offline on real neural data while reducing the total memory usage by 92%. Additionally, it matches state-of-the-art accuracy in a closed loop environment, demonstrates adaptability when subjected to multiple types of neural input disruptions, and is capable of being trained online without any prior offline training.Significance.This work presents a neural decoding algorithm that can be trained rapidly in a closed loop setting. The algorithm increases the speed of acclimating a new user to the system and also can adapt to sudden changes in neural behavior with minimal disruption to the user. 

Various breathing and cough simulators have been used to model respiratory droplet dispersion and viral droplets, in particular for SARS-CoV-2 modeling. However, limited data are available comparing these cough simulations to physiological breathing and coughing. In this study, three different cough simulators (Teleflex Mucosal Atomization Device Nasal (MAD Nasal), a spray gun, and GloGermTM MIST) that have been used in the literature were studied to assess their physiologic relevance. Droplet size, velocity, dispersion, and force generated by the simulators were measured. Droplet size was measured with scanning electron microscopy (SEM). Slow-motion videography was used to 3D reconstruct and measure the velocity of each simulated cough. A force-sensitive resistor was used to measure the force of each simulated cough. The average size of droplets from each cough simulator was 176 to 220 µm. MAD Nasal, the spray gun, and GloGermTM MIST traveled 0.38 m, 0.89 m, and 1.62 m respectively. The average velocities for the MAD Nasal, spray gun, and GloGermTM MIST were 1.57 m/s, 2.60 m/s, and 9.27 m/s respectively, and all yielded a force of <0.5 Newtons. GloGermTM MIST and the spray gun most closely resemble physiological coughs and breathing respectively. In conclusion, none of the simulators tested accurately modeled all physiologic characteristics (droplet size, 3-D dispersion velocity, and force) of a cough, while there were various strengths and weaknesses of each method. One should take this into account when performing simulations with these devices.