Neural Correlates of Reach Planning and Execution

<p>Humans and other primates often interact with the world by reaching and grabbing objects with their hands. This seemingly simple activity is a challenging computational problem that requires the nervous system to transform sensory input into muscle activations that move the hand appropriately through space. This dissertation investigates neural activity in the dorsal premotor cortex of the macaque monkey while simple and complex reaching movements are planned and executed. A novel virtual-reality obstacle-avoidance task is used to decorrelate the direction of the initial segment of the trajectory from the direction of the final target. An unobstructed center-out task is used as a comparison. The firing rates of many neurons are modulated by kinematic factors including hand position and movement direction. During obstacle-avoidance reaching, both the initial segment and final target directions are represented in the firing of dorsal premotor neurons. Population decoders for position, velocity and target direction were built using the indirect optimal linear estimator method, a variant of the population vector algorithm. The decoding model constructed from the direct-reaching task ultimately predicts the direction of movement, not the final target, during the planning period before movement begins. A separate decoding model predicts the target direction when the hand must move elsewhere initially. The time course of neural activity during planning suggests that the two monkeys utilized different preparatory strategies during the obstacle-avoidance task, leading to differences in performance on a subset of trials. A position-based population decoder predicts the hand trajectory during movement, anticipating the real hand position by approximately 200 ms. These findings demonstrate that multiple kinematic parameters of hand movement are represented in dorsal premotor cortex during planning and execution of voluntary reaching behavior. A simple linear decoding scheme based on roughly cosine-tuned spiking activity can extract relevant information from the population of neurons. This work contributes to the overall understanding of the factors that influence dorsal premotor cortical activity during complex reaching movements.</p>

Name thesis:degree_name

Doctor of Philosophy (PhD)

Level thesis:degree_level

Dissertation

Discipline thesis:degree_discipline

Biology and Biomedical Sciences: Neurosciences

Year dc:date.available

2014

Author dc:creator

• Pearce, Thomas

Contributors dc:contributor

• Daniel W Moran

• cosine tuning
• dorsal premotor cortex
• motor planning
• neural encoding
• population vector

Language dc:language

English (en)

DOI

https://doi.org/10.7936/K7CC0XQ1

OAI identifier oai:identifier

oai:openscholarship.wustl.edu:etd-2259