Spatial orientation is achieved by integrating sensory information from different modalities in order to estimate both body orientation and self-motion. Multiple sources of relevant sensory information are often available, so the question becomes, how should the nervous system combine them? If the goal is to achieve the single most probable combined estimate, then the rule is simple; each information source should be weighted based on its relative reliability. This kind of statistically optimal combination rule is known as Bayesian estimation, and it has been shown to accurately characterize perception in a variety of situations.

One aim of our research was to develop a comprehensive Bayesian model for spatial orientation perception. In order to develop the best possible model, we conducted a thorough review of existing spatial orientation models (MacNeilage et al. 2008). We focused on statistically optimal models of spatial orientation that are dynamic, meaning that the model input and output are continuous and vary over time. We concluded that the particle filter technique is best suited to modeling perception of spatial orientation because it is statistically optimal, the architecture is distributed rather than assuming particular feedback circuits, and most importantly, it is capable of implementing the non-linear system dynamics that are required for spatial orientation computations. Our review has provided key insights that will guide future modeling efforts. We have discussed these issues in some detail with other NSBRI investigators developing online systems that can predict spatial disorientation and alert crew members in situations when it is likely to occur (Project title: "Modeling and Mitigating Spatial Disorientation in Low-Gravity Environments"; Principal investigator: Ronald L. Small).

The second aim of our research was to measure visual and vestibular thresholds for detecting and discriminating spatial orientation stimuli. We conducted three such experiments using a motion simulator with attached stereo visual display. We used a two-alternative-forced-choice procedure because this method minimizes potential sources of response noise and bias. Also, because all data were collected using a common method and apparatus, it is possible to make interesting comparisons across conditions.

The first experiment investigated perception of heading, which is the direction of self-motion. Heading perception is fundamental to effective navigation and vehicle guidance, so it is important to understand the factors that influence the precision of visual and vestibular heading estimates. We measured discrimination thresholds for heading azimuth and elevation in visual-only and vestibular-only conditions with observers oriented upright and side-down relative to gravity. Visual thresholds were significantly lower than vestibular, and upright thresholds were generally lower than side-down. Vestibular results revealed that observers are better at discriminating head-centric azimuth than elevation, regardless of body orientation. In other words, sensitivity to heading depends more upon the direction of self-motion relative to the head, than relative to gravity. We conducted two follow-up experiments using the same subjects. We found that performance on the three tasks was correlated and determined that the asymmetric vestibular sensitivity to heading stems from an underlying difference in sensitivity to linear acceleration along the interaural and dorsoventral axes of the head.

In a second experiment we investigated the question of whether or not vestibular cues to linear and angular self-motion interact during rotation about an earth-vertical axis. Linear and angular self-motion is signaled by the otoliths and canals, respectively. During rotation about an earth-horizontal axis (i.e. tilt), these signals must interact so that the brain correctly interprets the change in the otolith signal as due to a change in the direction of gravity. In contrast, we found no evidence of canal-otolith interaction during rotation about an earth-vertical axis. Thresholds for detecting linear acceleration were not influenced by the speed of angular motion.

In a third experiment, we investigated whether or not vestibular cues to linear self-motion facilitate the detection of visual object motion during self-motion. Linear self-motion gives rise to a distinctive globally consistent pattern of motion on the retina known as optic flow. Independently moving objects in the scene will generate visual motion signals that deviate from the global pattern. Thus, observers must parse the optic flow in order to judge object motion during self-motion. We found that thresholds for discriminating object motion in optic flow patterns were reduced when vestibular signals consistent with the linear self-motion depicted by the optic flow were presented simultaneously. In other words, vestibular signals facilitate optic flow parsing.

The research described above contributes to a better understanding of spatial orientation in general. The modeling review provides unique insights that will be valuable for developing and evaluating models to predict and investigate situations where spatial disorientation is likely to occur. The results of the experiments provide a better understanding of visual and vestibular perception of spatial orientation stimuli which may be applied to the development of countermeasures like better cockpit display technology, and improved motion simulations. The psychophysical methods employed could be used to develop novel techniques for crew training and evaluation.

The psychophysical measurements we are making also contribute to a better general understanding of spatial orientation perception. The methodologies we have developed for these experiments could be used to assess performance of spatial orientation tasks by astronauts. They could also be used to assess the performance of aircraft pilots or patients with visual or vestibular deficits.

We used two-interval-forced-choice experiments to measure visual and vestibular sensitivity to spatial orientation stimuli. In particular, we measured thresholds for discriminating heading, which is the direction of linear self-motion, and investigated the dependency of discrimination performance on movement direction in head-coordinates and world-coordinates. We related vestibular heading thresholds to baseline measures of sensitivity to linear acceleration in head and world coordinates. In another experiment we determined that canal and otolith signals do not interact in subjects asked to detect linear acceleration during angular movement. These threshold measures may be used to determine the necessary parameters of Kalman or particle filter models described above, which depend on the reliability or noise present in vestibular sensory signals. In addition, the experimental methods we developed could be used to assess sensitivity of crew members to spatial orientation stimuli more precisely than is possible with standard medical diagnostics.

A final experiment investigated vestibular contributions to discriminating object motion during self-motion. This task requires observes to parse the visual optic flow pattern in order to estimate object motion and self-motion separately. Thresholds were lower when congruent vestibular stimulation was presented. This is one more demonstration of how visual and vestibular self-motion signals are intimately linked and should be considered in tandem as different parts of the same problem.

Program No. 18.5. Society for Neuroscience, Online, 2008. http://www.abstractsonline.com/Plan/ViewAbstract.aspx?sKey=36c5eea2-17b9-4a8f-877c-539dd9729ce3&cKey=cd1e01f7-0dc7-42bc-9483-71410ec388d0 , Nov-2008

J Vision. 2008;8(6):834a. http://journalofvision.org/8/6/834/ , May-2008

Program No. 367.16. Society for Neuroscience, Online, 2008. http://www.abstractsonline.com/Plan/ViewAbstract.aspx?sKey=a4102a70-1b51-403d-97fd-1736b6374cac&cKey=1bfd1d10-de47-4d8c-9e49-969981f3d92b , Nov-2008

Program No. 461.3. Society for Neuroscience, Online, 2008. http://www.abstractsonline.com/Plan/ViewAbstract.aspx?sKey=77ff31a5-583c-425f-8a6c-8667bebe75c4&cKey=e8c7f3fc-01b4-4654-a6df-38808da91fb5 , Nov-2008

Spatial orientation is achieved by integrating sensory information from different modalities in order to estimate both body orientation and self-motion. Multiple sources of relevant sensory information are often available, so the question becomes, how should the nervous system combine them? If the goal is to achieve the single most probable combined estimate, then the rule is simple; each information source should be weighted based on its relative reliability. This kind of statistically optimal combination rule is known as Bayesian estimation, and it has been shown to accurately characterize perception in a variety of situations.

We aim to develop a comprehensive Bayesian model for spatial orientation perception. This probabilistic model will take visual and vestibular signals as input, and it will generate combined estimates of orientation and linear and angular self-motion as output. The model will make specific, testable predictions about how visual and vestibular information should be combined. It will also make predictions about the time course of perceptual estimates in response to time-varying visual and vestibular inputs.

In order to develop the best possible model, we have conducted a thorough review of existing spatial orientation models. In particular, we have focused on statistically optimal models of spatial orientation that are dynamic, meaning that the model input and output are continuous and vary over time. The two approaches that appear most promising are Kalman and Particle filtering techniques. A Kalman filter is a control system architecture that compares predicted and observed output at each time step and uses the result to drive the dynamics of the system. It is statistically optimal because the gain applied to the feedback signal is influenced by the reliability of the input signals. A Particle filter is a probabilistic modeling technique that relies on simulating the full probability distributions of all inputs and outputs at each time step. In order to evaluate these two alternative approaches we have simulated them. The results of these simulations are guiding development of a revised dynamic Bayesian model.

We are also conducting a number of psychophysical experiments to measure discrimination performance for a variety of spatial orientation stimuli. These experiments use a motion simulator and measure perceptual estimates of orientation and linear and angular self-motion in visual-only, vestibular-only, and combined visual-vestibular conditions. These reliability measures are important parameters of statistically optimal models. We use a two-alternative-forced-choice procedure because this method minimizes potential sources of response noise and bias. Also, because all data is collected using a common method and apparatus, it is possible to make interesting comparisons across conditions.

Recent experiments have investigated perception of heading, which is the direction of self-motion. Heading perception is fundamental to effective navigation and vehicle guidance, so it is important to understand the factors that influence the precision of visual and vestibular heading estimates. We measured discrimination thresholds for heading azimuth and elevation in visual-only and vestibular-only conditions with observers oriented upright and side-down relative to gravity. Visual thresholds were significantly lower than vestibular, and upright thresholds were generally lower than side-down. Both visual and vestibular results revealed that observers are better at discriminating head-centric azimuth than elevation, regardless of body orientation. In other words, sensitivity to heading depends more upon the direction of self-motion relative to the head, than relative to gravity.

Another ongoing experiment is investigating how the threshold for vestibular detection of linear self-motion depends on the direction of translation in head coordinates. We plan to conduct a similar series of experiments to investigate how vestibular detection of angular self-motion depends on the axis of rotation. Linear and angular self-motion thresholds are measurements of the reliabilities of vestibular sensory estimates which are necessary parameters of statistically optimal models.

We are also conducting an experiment to investigate the interaction of vestibular rotation and translation signals for the perception of self-motion on a curved path. Results of this experiment will be compared with predictions of the model, once it is developed. In future we plan to conduct a similar experiment to investigate the perception of tilt versus translation.

The research described above will contribute to a better understanding of spatial orientation in general. The model may be used to predict and investigate situations where spatial disorientation is likely to occur. The results of the experiments may be applied to the development of countermeasures including better cockpit display technology, improved motion simulations, novel pilot training techniques, and crew screening procedures.

The psychophysical measurements we are making also contribute to a better general understanding of spatial orientation perception. The methodologies we have developed for these experiments could be used to assess performance of spatial orientation tasks by astronauts. They could also be used to assess the performance of aircraft pilots or patients with visual or vestibular deficits.

Psychophysics progress: We have completed a series of experiments investigating human ability to perceive the direction of self-motion based on visual and vestibular cues. The visual cue to heading is the location of the focus of expansion in the optic flow field. The vestibular cue to heading is the direction of inertial acceleration signaled by the otoliths. Prior research has focused on visual discrimination of heading azimuth (heading in the horizontal plane). There have been few studies of visual discrimination of elevation, and fewer comparable studies of non-visual heading discrimination. To investigate human ability to estimate heading under more general conditions, we measured heading discrimination thresholds for azimuth and elevation in visual-only and vestibular-only conditions with observers oriented upright and side-down relative to gravity. Experiments were conducted in the Human Motion Lab at Washington University, a motion simulator consisting of a 6DOF Moog motion base and 90 deg X 90 deg stereo projection screen. Subjects were seated on the motion platform in a padded racing seat, and held in place with a 5-point harness. The head was secured to a cushioned head mount by a form fitted plastic mesh mask. Noise was played through noise-cancellation headphones and responses were collected using a custom made button box. Subjects were asked to discriminate heading azimuth or elevation relative to straight-ahead in a two-interval-forced-choice task. Visual thresholds were significantly lower than vestibular, and upright thresholds were generally lower than side-down. Both visual and vestibular results revealed that observers are better at discriminating head-centric azimuth than elevation, regardless of body orientation. In other words, sensitivity to heading depends more upon the direction of self-motion relative to the head, than relative to gravity. Heading perception is fundamental to effective navigation and vehicle guidance, so it is important to understand how the precision of visual and vestibular heading estimates depend on the direction of self-motion in head-coordinates and world coordinates. Furthermore, it is useful to characterize the reliabilities of visual and vestibular heading estimates because these are necessary parameters of statistically optimal models like the one we aim to develop.

Perception 2007;36 Suppl:86. , Aug-2007

Journal of Vision 2007;7(9):101a. http://journalofvision.org/7/9/101/ , May-2007

Spatial orientation is achieved by integrating sensory information from different modalities in order to estimate both body orientation and self-motion. Multiple sources of relevant sensory information are often available, so the question becomes, how should the nervous system combine them? If the goal is to achieve the single most probable combined estimate, then the rule is simple: each information source should be weighted based on its relative reliability. This kind of probabilistic combination rule is known as Bayesian estimation, and it has been shown to accurately characterize perception in a variety of situations.

My project is to develop a comprehensive Bayesian model for spatial orientation perception. This model will share many desirable features with previously proposed models. It will take visual and vestibular signals as input, it will be dynamic, and it will generate combined estimates of orientation, linear velocity and angular velocity as output. However, it has a distinct advantage over previous models in that it makes specific, test-able predictions about how visual and vestibular information should be combined. Estimates based on multiple cues should be more reliable (less variable) than those based on any of the contributing cues in isolation. Also, cues should be weighted based on their reliability with more reliable cues being weighted more highly.

We will test these predictions in a series of psychophysical experiments. We will measure discrimination performance for each individual in a variety of conditions. The estimates of interest are the outputs of the model, namely orientation, linear velocity and angular velocity. For each of these, we will run visual-only, vestibular-only, and combined visual-vestibular conditions. We will use a two-alternative-forced-choice procedure because this method minimizes potential sources of response noise and bias. It will be possible to make interesting comparisons across conditions because all data will be collected using a common apparatus and method.

This research will contribute to a better understanding of spatial orientation. The model may be used to predict and investigate situations where spatial disorientation is likely to occur. The results of the experiments may be applied to the development of countermeasures including better cockpit display technology, improved motion simulations, novel pilot training techniques and crew-screening procedures.