Rocking or rolling--perception of ambiguous motion after returning from space

Abstract

The central nervous system must resolve the ambiguity of inertial motion sensory cues in order to derive an accurate representation of spatial orientation. Adaptive changes during spaceflight in how the brain integrates vestibular cues with other sensory information can lead to impaired movement coordination, vertigo, spatial disorientation, and perceptual illusions after return to Earth. The purpose of this study was to compare tilt and translation motion perception in astronauts before and after returning from spaceflight. We hypothesized that these stimuli would be the most ambiguous in the low-frequency range (i.e., at about 0.3 Hz) where the linear acceleration can be interpreted either as a translation or as a tilt relative to gravity. Verbal reports were obtained in eleven astronauts tested using a motion-based tilt-translation device and a variable radius centrifuge before and after flying for two weeks on board the Space Shuttle. Consistent with previous studies, roll tilt perception was overestimated shortly after spaceflight and then recovered with 1-2 days. During dynamic linear acceleration (0.15-0.6 Hz, ±1.7 m/s2) perception of translation was also overestimated immediately after flight. Recovery to baseline was observed after 2 days for lateral translation and 8 days for fore-aft translation. These results suggest that there was a shift in the frequency dynamic of tilt-translation motion perception after adaptation to weightlessness. These results have implications for manual control during landing of a space vehicle after exposure to microgravity, as it will be the case for human asteroid and Mars missions.

Figure 1. Gain perception during static tilt in roll induced by centrifugation (A) or static tilt in pitch induced by actual tilt relative to gravity (B) before (L-120, L-90, L-60 days) and after (R+0 to R+8 days) spaceflight.

In A, gain was calculated as the ratio of subjective reported tilt versus tilt of the gravitoinertial acceleration (GIA) vector. In B, gain was calculated as the ratio of subjective reported tilt versus actual chair tilt. Dashed lines correspond to a gain of unity. Mean ± SE of 10 subjects. *p<0.05 relative to the last pre-flight session.

Figure 2. Perception gains during lateral oscillations at 0.15, 0.3, and 0.6 Hz before and after spaceflight.

A. The reported magnitude of self-displacement sideways was divided by the actual magnitude of the translation stage (translation was actually ±11.4, ±9.7, and ±6.1 cm at 0.15, 0.3, and 0.16 Hz, respectively). B. The reported angle of roll tilt was divided by the tilt of the GIA when the translation stage reached maximum eccentricity (GIA tilt was theoretically ±10° at each frequency). Dashed lines correspond to a gain of unity. Mean ± SE of 10 subjects. *p<0.05 relative to the last pre-flight session.

Figure 3. Perception gains during fore-aft oscillations at 0.15, 0.3, and 0.6 Hz before and after spaceflight.

A. The reported magnitude of fore-aft self-displacement was divided by the actual magnitude of sled translation along the track (translation was actually ±195, ±49, and ±12 cm at 0.15, 0.3, and 0.16 Hz, respectively). B. The reported angle of tilt in pitch was divided by the maximum tilt of the chair (tilt varied between ±10° at each frequency). Dashed lines correspond to a gain of unity. Mean ± SE of 10 subjects. *p<0.05 relative to the last pre-flight session.

Figure 4. Differences between the translation gain and the tilt gain during dynamic lateral oscillations and roll-induced tilt (A) and during fore-aft oscillations and pitch tilt (B).

Data points represent the mean differential gains obtained during translation or tilt at 0.15, 0.3, and 0.6 Hz. Curves are quadratic fits to the data using the least-square regression method (R² are 0.96 or higher). The intercept of each curve fit with the dashed line is the crossover frequency between the perception of tilt or translation: a negative difference (below zero) corresponds to a perceived tilt > perceived translation; a positive difference (above zero) corresponds to a perceived translation > perceived tilt.

Figure 5. The phase of the subjective fore-aft direction of head motion relative to tilt motion depends from whether the axis of rotation during body tilt in pitch is perceived to be above or below the subject’s head.