Baselines for three-dimensional perception of combined linear and angular self-motion with changing rotational axis.

The laws of physics explain many human misperceptions of whole-body passive self-motion. One classic misperception occurs in a rotating chair in the dark: If the chair is decelerated to a stop after a period of counterclockwise rotation, then a subject will typically perceive clockwise rotation. The laws of physics show that, indeed, a clockwise rotation would be perceived even by a perfect processor of angular acceleration information, assuming that the processor is initialized (prior to the deceleration) with a typical subject's initial perception - of no rotation in this case. The motion perceived by a perfect acceleration processor serves as a baseline by which to judge human self-motion perception; this baseline makes a rough prediction and also forms a basis for comparison, with uniquely physiological properties of perception showing up as deviations from the baseline. These same principles, using the motion perceived by a perfect acceleration processor as a baseline, are used in the present paper to investigate complex motions that involve simultaneous linear and angular accelerations with a changing axis of rotation. Baselines - motions that would be perceived by a perfect acceleration processor, given the same initial perception (prior to the motion of interest) as that of a typical subject - are computed for the acceleration and deceleration stages of centrifuge runs in which the human carriage tilts along with the vector resultant of the centripetal and gravity vectors. The computations generate a three-dimensional picture of the motion perceived by a perfect acceleration processor, by simultaneously using all six interacting degrees of freedom (three angular and three linear) and taking into account the non-commutativity of rotations in three dimensions. The resulting three-dimensional baselines predict stronger perceptual effects during deceleration than during acceleration, despite the equal magnitudes (with opposite direction) of forces on the subject during acceleration and deceleration. For a centrifuge run with the subject facing tangentially in the direction of motion, the deceleration baseline shows a perception of forward tumble (pitch rotation) beginning with ascent from the earth, while the acceleration baseline does not have analogous pitch and vertical motion. These results give a three-dimensional explanation for certain puzzling acceleration-deceleration perceptual differences observed experimentally by Guedry, Rupert, McGrath, and Oman (Journal of Vestibular Research, 1992.). The present analysis is consistent with, and expands upon, previous analyses of individual components of motion.

Identification of head motions by central vestibular neurons receiving linear and angular input.

Most naturally occurring displacements of the head in space, due to either an external perturbation of the body or a self-generated, volitional head movement, apply both linear and angular forces to the head. The vestibular system detects linear and angular accelerations of the head separately, but the succeeding control of gaze and posture often relies upon the combined processing of linear and angular motion information. Thus, the output of a secondary neuron may reflect the linear, the angular, or both components of the head motion. Although the vestibular system is typically studied in terms of separate responses to linear and angular acceleration of the head, many secondary and higher-order neurons in the vestibular system do, in fact, receive information from both sets of motion sensors. The present paper develops methods to analyze responses of neurons that receive both types of information, and focuses on responses to sinusoidal motions composed of a linear and an angular component. We show that each neuron has a preferred motion, but a single neuron cannot code for a single motion. However, a pair of neurons can code for a motion by the relative phases of firing-rate modulation. In this way, information about motion is enhanced by neurons combining information about linear and angular motion.

Timing of secondary vestibular neuron responses to a range of rotational head movements.

Secondary vestibular neurons exhibit a wide variety of responses to a head movement, with the response of each secondary neuron depending upon the particular primary afferents converging onto it. A single head movement is thereby registered in a distributed manner. This paper focuses on implications of afferent convergence to the relative timing of secondary neuron response modulation during rotational movements about a combination of horizontal axes. In particular, the neurons of interest are those that receive input from afferents innervating the vertical semicircular canals, and the movements of interest are those that have a sinusoidal component about one vertical canal axis and a sinusoidal component about another, approximately orthogonal, vertical canal axis. Under these conditions, the present research shows that it is possible for two or more secondary neurons to have a different relative timing of response (i.e., different relative phase of the periodic modulation in firing rate) for different head movements, and for the neurons to switch their order of response for different movements. For particular head movements, those same neurons will respond in phase. From the point of view of the nervous system, the relative timing of neuron responses may tell which movement is taking place, but with certain restrictions as discussed in the present paper. Shown here is that, among those head movements for which the two components of rotation may be at any phase relative to one another and have any relative amplitude, an in-phase response of just two neurons cannot identify a single motion. Two neurons that respond in phase for one motion must respond in phase for an entire range of motions; all motions in that range are thus response-equivalent, in the sense that the pair of neurons cannot distinguish between the two motions. On the other hand, an in-phase response of three neurons can identify a single motion, for certain patterns of primary afferent convergence.

Three-dimensional baselines for perceived self-motion during acceleration and deceleration in a centrifuge.

Three-dimensional motion trajectories were computed, representing the motions that would be perceived by a perfect processor of acceleration information during the acceleration and deceleration stages of a centrifuge run. These motions serve as "baselines" for perceived self-motion in a centrifuge, and depend on the initial perception of orientation and velocity immediately preceding the acceleration and immediately preceding the deceleration. The baselines show that a perfect processor of acceleration information perceives self-motion during centrifuge deceleration significantly differently from self-motion during centrifuge acceleration, despite the fact that the angular accelerations have equal magnitude (with opposite direction). At the same time, the baselines can be compared with subjects' reported perceptions to highlight limitations of the nervous system; limitations and peculiarities of the nervous system are identified as deviations from a baseline. As a result, peculiarities of the nervous system are held responsible for any perception of pitch or roll angular velocity or change in tilt of the body-horizontal plane of motion during the centrifuge run. On the other hand, baselines explain perception of tilt position during deceleration, linear velocity, possible lack of significant linear velocity during deceleration, and yaw angular velocity, including on-axis angular velocity during centrifuge deceleration. The results lead to several experimental questions.

The shape of self-motion perception--I. Equivalence classification for sustained motions.

Two completely different motions of a subject relative to the earth can induce exactly the same stimuli to the vestibular, somatosensory and visual systems. When this happens, the subject may experience disorientation and misperception of self-motion. We have identified large classes of motions that are perceptually equivalent, i.e. indistinguishable by the subject, under three sets of conditions: no vision, with vision and earth-fixed visual surround, and with vision during possible movement of the visual surround. For each of these sets of conditions, we have developed a classification of all sustained motions according to their perceptual equivalences. The result is a complete list of the possible misperceptions of sustained motion due to equivalence of the forces and other direct stimuli to the sensors under the given conditions. This research expands the range of possible experiments by including all components of linear and angular velocity and acceleration. Many of the predictions in this paper can be tested experimentally. In addition, the equivalence classes developed here predict perceptual phenomena in unusual motion environments that are difficult or impossible to investigate in the laboratory.

The shape of self-motion perception--II. framework and principles for simple and complex motion.

There have been numerous experimental studies on human perception and misperception of self-motion and orientation relative to the earth, each focusing on one or a few types of motion. We present a formal framework encompassing many types of motion and including all angular and linear components of velocity and acceleration. Using a mathematically rigorous presentation, the framework defines the space of all possible motions, the map from motion to sensor status, the space containing each possible status of the sensors, and the map from sensor status to perceived motion. The shape of the full perceptual map from actual motion to perceived motion is investigated with the framework, using formal theory and a number of published experimental results. Two principles of simple motion perception and four principles of complex motion perception are presented. The framework also distinguishes the roles of physics and the nervous system in the process of self-motion perception for both simple and complex motions. The present rigorous development of the self-motion perception framework allows the scientist to compare and contrast results from many studies with differing types of motion. The six principles formalized here comprise a foundation with which to explain and predict perceptual phenomena, both those observed in the past and those to be encountered in the future. The framework is especially aimed to expand our capacity to investigate complex motions such as those encountered in everyday life or in unusual motion environments.