A quantitative investigation of natural head movement and its contribution to spatial orientation perception.

Loading...
Thumbnail Image

Authors

Sinnott, Christian

Issue Date

2023

Type

Dissertation

Language

Keywords

Research Projects

Organizational Units

Journal Issue

Alternative Title

Abstract

Movement is ubiquitous in everyday life. As we exist in a physical world, we constantly account for our position in it relative to other physical features: both at a conscious, volitional level and an unconscious one. Our experience estimating our own position accumulates over the lifespan, and it is thought that this experience (often referred to as a prior) informs current perception of spatial orientation. Broadly, this perception of spatial orientation is rapidly performed by the nervous system by monitoring, interpreting, and integrated sensory information from multiple sense organs. To do this efficiently, the nervous system likely represents this sensory information in a statistically optimal manner. Some of the most important information for spatial orientation perception comes from visual and vestibular sensation, which rely on sensory organs located in the head. While statistical information about natural visual and vestibular stimuli have been characterized, natural head movement and position, which likely drives correlated dynamics across head-located senses, has not. Furthermore, sensory cues essential to spatial orientation perception are directly affected by head movement specifically. It is likely that measurement of these sensory cues taken during natural behaviors sample a significant portion of the total behaviors that comprise ones’ prior. In this dissertation, I present work quantifying characteristics of head orientation and heading, two dimensions of spatial orientation, over long-duration recordings of natural behavior in humans. Then, I use these to generate priors for Bayesian modeling frameworks which successfully predict observed patterns of orientation and heading perception bias. Given the ability to predict some patterns of bias (head roll and heading azimuth) particularly well, it is likely our data are representative of real behaviors that comprise previous experience the nervous system may have. Natural head orientation and heading distributions reveal several interesting trends that open future lines of research. First, head pitch demonstrates large amounts of inter-subject variability; likely this is due to biomechanical differences, but as these remain relatively stable over the lifespan these should bias head movements. Second, heading azimuth appears to vary significantly as a function of task. Heading azimuth distributions during low velocities (which predominantly consist of stationary activities like standing or sitting) are strongly multimodal across all subjects, while azimuth distributions during high velocities (predominantly consisting of locomotion) are unimodal with relatively low variance. Future work investigating these trends, as well as implications these trends and data have for sensory processing and other applications is discussed.

Description

Citation

Publisher

Journal

Volume

Issue

PubMed ID

DOI

ISSN

EISSN