Neural Substrates of Trial-and-Error Motor Learning

Jangid, Abhishek (2023) Neural Substrates of Trial-and-Error Motor Learning. Masters thesis, Indian Institute of Science Education and Research Kolkata.

Text (MS dissertation of Abhishek Jangid (18MS152)) Thesis_18MS152.pdf - Submitted Version Restricted to Repository staff only Download (4MB)

Abstract

In our daily lives,learning by trial-and-error or remforcement learning plays a vital role in acquiring new skills and adapting to changing environments. From learning to ride a bike to developing expertise in a particular profession, learning to dance or leanring to play guitar, trial-and-error learning 1s what takes us to expertise. For example, when learning to play a new instrument, we might try different finger placements or techniques until we find what works best for us and emploting trial-and-error here ts a fundamental mechanism that enables us to refine our actions and make progress towards our goals. Variability is one of the main components of trial-and-error learning and it depends on the rewards received 1n the past. Although our understanding of how the brain implements trial-and-error learning comes primarily from decision-making tasks with a limited number of options. motor learning in real life involves improving movement features. such as them kinematics or timing of our movements, to achieve specific performance goals. This requires the brain to learn in continuous high-dimensional action spaces where the number of "choices" 1s infinite. Further, despite the significance of this type of learning in our daily lives, we have limited knowledge about its neural underpinnings. My goal is to uncover the neural substrates of trial-and-error motor learning using long-term, high-throughput experiments on rats. specifically using a 2D lever-pressing task. To achieve this, I employ the use of behavioural experiments and lesioning of midbrain motor areas. Additionally. I will analyse electrophysiological recordings taken from rats, specifically targeting the motor cortex and basal ganglia in a timed lever press task that requires them to make two presses with a specific inter-press interval (IPI) to receive a reward. The IPI is the component which the rats have to figure out using trial-and-error. The The electrophvsiological recordings should contain information about how the rats modify their actions depending on whether they are getting reward or not. We predict that this information will be encoded in form of spike rate, which we can analyse. By analysing spike activity modulation and understanding how past outcomes affect current performance, I hope to gain insight into the fundamental principles underlying this type of learning.

Actions (login required)

View Item