No full-text available
To read the full-text of this research,
you can request a copy directly from the authors.
Stepping of the forelegs over obstacles establishes long-lasting memories in cats
Abstract
Although visual input is used heavily during locomotion [1], intermittent visual input is sufficient for most walking tasks. A number of techniques provide evidence that suggests that short-term visual memory is used to fill in the resulting gaps. When stepping over obstacles, for example, humans fixate the obstacle primarily one or two steps before they reach it [2], and removing their vision during the step over the obstacle does not affect their ability to step over it accurately [3]. Walking cats consistently look two or three steps ahead when walking [4], and can continue stepping accurately among obstacles for about four steps when visual input is suddenly removed [5]. This use of short-term memory raises questions pertinent for those interested in the neurobiology of walking as well as those interested in memory in general. Our laboratory has begun to exploit the fact that walking quadrupeds must rely on some form of visual memory to guide their hind legs over obstacles. Our experiments show that stepping over obstacles triggers long-lasting memories in walking cats.
... In addition, in many circumstances, animals step accurately with their hind limbs over obstacles that the forelimbs have avoided even though they can no longer see the obstacle as the hind limbs step [4]. McVea and Pearson [9,10] have investigated this form of obstacle avoidance memory by having cats pause after stepping over an obstacle with the forelimbs. They found that hind limb lifting to avoid the obstacle persisted after delays as long as 10 min, even though the obstacle may have been removed in the interval. ...
... For the experiments, the movement of the first forelimb or the first hind limb to clear the obstacle was the dependent variable and limb amplitude (the highest point of the trajectory) was the main dependent variable. In all experiments, stepping distance was shorter for trials on which horses were paused versus trials on which the horses walked continuously, confirming McVea and Pearson [9,10] that Fig. 1. Illustration of the testing situation. ...
... The findings of the present study are consistent with a number of the results reported by McVea and Pearson [9,10] in the cat. Limb lifting of the forelimbs to clear an obstacle contributes to hind limb elevation and may be essential for the hind limb response. ...
An animal that has stepped over an obstacle with its forelimbs uses a memory of the obstacle to guide the hind limbs so that they also clear the obstacle, even in situations in which long pauses are introduced between forelimb and hind limb stepping. To further clarify the features of hind limb obstacle clearance memory, the present study examined hind limb obstacle clearance in the horse. A rider guided horses over obstacles and paused the horse over obstacles in tests that examined the relationship between forelimb and hind limb stepping, with the following results. First, the horses displayed memory for an obstacle as measured by hind limb lifting over the obstacle for durations lasting as long as 15 min. The response was not dependent upon ongoing visualization of the obstacle, as limb lifting was unaffected by visual occlusion with blinders, a blindfold, or by removing the obstacle during the pause. Second, previous experience of stepping over an obstacle led to pause-related hind limb lifting at the object's previous location even on trials for which there was no obstacle and so no preceding forelimb lifting. Third, whereas a horse would lift its hind limbs to clear two successively presented obstacles, replacing an obstacle before the horse after the forelimbs had cleared the obstacle prevented subsequent hind limb lifting at the obstacle's previous location. Taken together the results show that hind limb obstacle clearance is guided by a place-object memory. The results are discussed in relation to the differential sensory and memonic control of forelimb and hind limb stepping with the suggestion that place-object memory can guide hind stepping as well as overshadow working memory from front leg stepping.
... Indeed, patients with Alzheimer's disease (AD) have a higher frequency of contact with obstacles than healthy older adults 9 , and such contact seems to be more frequent in the trailing limb 10 . This may be because stepping over an obstacle with the trailing limb is a movement that is not visible, and thus, is guided by the working memory of the obstacle height [10][11][12][13] . This concept has also been confirmed in studies using quadrupedal animals, who cannot see their hindlimb movements, wherein AD model mice showed a higher frequency of contact of the hindlimbs during an obstacle avoidance task than non-AD model mice 14 . ...
... A lower clearance of the trailing limb compared to that of the leading limb was observed among older adults whose impaired memory function was assumed to diminish their ability to internally represent an obstacle encountered during walking 11 . This supports the concept that the memory of an obstacle encountered during walking would persist during obstacle avoidance [11][12][13]27 . ...
An association between cognitive impairment and tripping over obstacles during locomotion in older adults has been suggested. However, owing to its memory-guided movement, whether this is more pronounced in the trailing limb is poorly known. We examined age-related changes in stepping over, focusing on trailing limb movements, and their association with cognitive performance. Age-related changes in obstacle avoidance were examined by comparing the foot kinematics of 105 older and 103 younger adults when stepping over an obstacle. The difference in the clearance between the leading and trailing limbs (Δ clearance) was calculated to determine the degree of decrement in the clearance of the trailing limb. A cognitive test battery was used to evaluate cognitive function among older adults to assess their association with Δ clearance. Older adults showed a significantly lower clearance of the trailing limb than young adults, resulting in greater Δ clearance. Significant correlations were observed between greater Δ clearance and scores on the Montreal Cognitive Assessment and immediate recall of the Wechsler Memory Scale-Revised Logical Memory test. Therefore, memory functions may contribute to the control of trailing limb movements, which can secure a safety margin to avoid stumbling over an obstacle during obstacle avoidance locomotion.
... Indeed, patients with Alzheimer's disease (AD) have a higher frequency of contact with obstacles than healthy older adults 9 , and such contact seems to be more frequent in the trailing limb 10 . This may be because stepping over an obstacle with the trailing limb is mostly guided by the working memory of the obstacle height [10][11][12][13] . This concept has been also con rmed in studies using quadrupedal animals, which cannot visually recognize their hindlimbs movements, where AD model mice showed a higher frequency of contact of the hindlimbs during an obstacle avoidance task than non-AD model mice 14 . ...
... For instance, accurate stepping movements of the trailing limb based on obstacle memory can be performed after a delay period of 2 minutes 25 . Low clearance of the trailing limb is observed in older adults whose impaired memory function is therefore assumed to be diminishing their ability to internally represent an obstacle encountered during walking 11 , supporting the concept that the memory of an obstacle encountered during walking would persist during obstacle avoidance [11][12][13]25 . ...
An association between cognitive impairment and tripping over obstacles during locomotion in older adults has been suggested. However, owing to its memory-guided movement, whether this is more pronounced in the trailing limb is poorly known. We examined the age-related changes in stepping-over, focusing on trailing limb movements, and their association with cognitive performance. Age-related change in obstacle avoidance was examined by comparing the foot kinematics of 105 older and 103 younger adults when stepping over an obstacle. The difference in clearance between the leading limb and trailing limb (Δ clearance) was calculated to determine the degree of decrement in the clearance of the trailing limb. A cognitive test battery was used to evaluate cognitive function among older adults for assessing their association with Δ clearance. Older adults showed a significantly lower clearance of the trailing limb than younger adults, resulting in a greater Δ clearance. The significant correlations between greater Δ clearance and scores of Montreal Cognitive Assessment and delayed recall of the Wechsler Memory Scale-Revised Logical Memory. Our results suggest that memory functions may contribute to the control of trailing limb movements, which can secure a safety margin to avoid stumbling on an obstacle, during obstacle avoidance locomotion.
... Triggers are assumed to have a range of roles in HMC, including movement initiation (via the Go signal, for example) and transitions from one phase of locomotion to another [16,17]. However, the main aspect of IC considered here is the use of SMH to approximate the predicted state in open-loop, in order to reduce the computational and communication burden. ...
... The analysis was facilitated by converting the time-delayed continuous system to an equivalent time-delayed discrete-time system using the standard zero-order hold (ZOH), as detailed in Stability of continuous LTI systems. The equivalent discrete-time system is stable as long as all the eigenvalues of A ex , specified in Eq (16), are inside the unit circle. Note that an eigenvalue λ d of a discrete-time system depends on the discretization time Δ, and evolves as l k d at t = kΔ. ...
Background and objectives
Human motor control (HMC) has been hypothesized to involve state estimation, prediction and feedback control to overcome noise, delays and disturbances. However, the nature of communication between these processes, and, in particular, whether it is continuous or intermittent, is still an open issue. Depending on the nature of communication, the resulting control is referred to as continuous control (CC) or intermittent control (IC). While standard HMC theories are based on CC, IC has been argued to be more viable since it reduces computational and communication burden and agrees better with some experimental results. However, to be a feasible model for HMC, IC has to cope well with inaccurately modeled plants, which are common in daily life, as when lifting lighter than expected loads. While IC may involve event-driven triggering, it is generally assumed that refractory mechanisms in HMC set a lower limit on the interval between triggers. Hence, we focus on periodic IC, which addresses this lower limit and also facilitates analysis.
Theoretical methods and results
Theoretical stability criteria are derived for CC and IC of inaccurately modeled linear time-invariant systems with and without delays. Considering a simple muscle-actuated hand model with inaccurately modeled load, both CC and IC remain stable over most of the investigated range, and may become unstable only when the actual load is much smaller than expected, usually smaller than the minimum set by the actual mass of the forearm and hand. Neither CC nor IC is consistently superior to the other in terms of the range of loads over which the system remains stable.
Numerical methods and results
Numerical simulations of time-delayed reaching movements are presented and analyzed to evaluate the effects of model inaccuracies when the control and observer gains are time-dependent, as is assumed to occur in HMC. Both IC and CC agree qualitatively with previously published experimental results with inaccurately modeled plants. Thus, our study suggests that IC copes well with inaccurately modeled plants and is indeed a viable model for HMC.
... In quadrupedal animals, obstacle WM is especially important for ensuring hindlimb clearance, as the animal can no longer see the obstacle once it has passed under the body. Instead, an internal representation of the obstacle maintained in WM is used to guide hindleg stepping [4][5][6]. ...
... To examine WM-guided obstacle avoidance, each trial was composed of three phases: the approach; delay; and continuation. As foreleg obstacle clearance is essential for establishing a robust WM of the obstacle [5], the initial approach phase of obstacle-present (OP) trials consisted of the animals stepping over the obstacle with only their forelegs ( Figure 1A). Hindleg clearance was then delayed. ...
In complex environments, information about surrounding obstacles is stored in working memory (WM) and used to coordinate appropriate movements for avoidance. In quadrupeds, this WM system is particularly important for guiding hindleg stepping, as an animal can no longer see the obstacle underneath the body following foreleg clearance. Such obstacle WM involves the posterior parietal cortex (PPC), as deactivation of area 5 incurs WM deficits, precluding successful avoidance. However, the neural underpinnings of this involvement remain undefined. To reveal the neural substrates of this behavior, microelectrode arrays were implanted to record neuronal activity in area 5 during an obstacle WM task in cats. Early in the WM delay, neurons were modulated generally by obstacle presence or more specifically in relation to foreleg step height. Thus, information about the obstacle or about foreleg clearance can be retained in WM. In a separate set of neurons, this information was recalled later in the delay in order to plan subsequent hindleg stepping. Such early and late delay period signals were temporally bridged by neurons exhibiting obstacle-modulated activity sustained throughout the delay. These neurons represented a specialized subset of all recorded neurons, which maintained stable information coding across the WM delay. Ultimately, these various patterns of task-related modulation enable stable representations of obstacle-related information within the PPC to support successful WM-guided obstacle negotiation in the cat.
... Furthermore, this long-lasting working memory of obstacle size and location is only established if the forelegs have already stepped over the obstacle. If a cat is stopped just before stepping over an obstacle, the memory of its size and location decays over seconds (McVea and Pearson, 2007). We have suggested that an efference copy of foreleg motor commands likely combines with visual information about an obstacle's size and location to establish a neural representation of obstacle location relative to the body that persists while the animal is straddling the obstacle. ...
... Although our current study suggests that spatial memories of objects encountered while walking might be stored in area 5 for long delays but elsewhere for shorter delays, we must also consider that the duration of memories of obstacles depends crucially on whether or not the forelegs have stepped over the obstacle. Straddled obstacles are remembered for very long durations (up to at least 10 min) (McVea and Pearson, 2006), whereas obstacles that are not straddled are remembered for much shorter durations (McVea and Pearson, 2007). In fact, the time course of the memory loss of an obstacle perceived only visually (i.e., in the absence of the forelegs stepping over the obstacle) in intact cats is very similar to the time course of loss we observed in our animalsFigure 7. Anatomy of lesions in area 5 of the left hemisphere of the three cats receiving simultaneous bilateral lesions (cats 1–3). ...
Walking animals rely on working memory to avoid obstacles. One example is the stepping of the hindlegs of quadrupeds over an obstacle. In this case, the obstacle is not visible at the time of hindleg stepping, because of its position between the fore and hindlegs, and working memory must be used to avoid it. We have previously shown that this memory is very precise and surprisingly long-lasting and that it depends on the stepping of the forelegs over the obstacle for its initiation. In this study, we test the hypothesis that area 5 in the posterior parietal cortex of cats is necessary for the maintenance of this long-lasting working memory. We report that small bilateral lesions to area 5 do not affect the amplitude of normal stepping of the hindlegs over obstacles, but they profoundly reduce the long-lasting working memory of obstacles. We propose that inputs to area 5 associated with foreleg stepping initiate long-lasting activity that maintains the memory of obstacle height in another brain region to guide the hindlegs over obstacles.
... These inter-study discrepancies could be explained by the differences in obstacle height between the studies (3-7 cm in the previous study; 15 cm in our study), which may have resulted in differences in the appearance of visual illusions. Furthermore, considering that stepping over an obstacle with the trailing limb is guided mostly by memory (McVea & Pearson, 2007;Sakurai et al., 2021), there may have been an optimal approach distance and obstacle height that would best appeal to memory to create an illusion effect on the trailing foot. ...
An adequate foot clearance height while stepping over an obstacle is important for safety in daily life. In the present study, we examined whether visual illusions affect foot clearance during a stepping-over action, and whether this is further influenced by gaze behavior. Twelve participants stepped over an obstacle placed four meters away under conditions of three different obstacle characteristics: white, horizontal, or vertical lines. We measured the participants’ foot clearances during the step-over action and their gaze behavior during the approaching phase. Participants stepped significantly higher over the obstacles in the vertical lines (illusion) condition. The duration of gaze fixation on the obstacle positively correlated with increased foot clearance in the vertical condition, suggesting that the effect of the visual illusion on foot clearance was enhanced by prolonged gaze fixation. Conversely, prolonged fixation negatively correlated with foot clearance in the white (control) condition, implying that a cautious perception of an obstacle may contribute to efficient stepping-over action.
... Cutaneous afferents, particularly from the superficial peroneal nerve, trigger the SCR in cats and humans (Prochazka et al., 1978;Forssberg, 1979;Wand et al., 1980;Buford and Smith, 1993;Schillings et al., 1996Schillings et al., , 2000Schillings et al., , 2005van Wezel et al., 1997;Lam et al., 2003;Quevedo et al., 2005;Potocanac et al., 2016;Merlet et al., 2022;McVea and Pearson, 2007b). In our study, cats showed SCR in the ipsilesional right hindlimb in the first 2 weeks after hemisection, as shown by Doperalski et al. (2011). ...
Following incomplete spinal cord injury in animals, including humans, substantial locomotor recovery can occur. However, functional aspects of locomotion, such as negotiating obstacles remains challenging. We collected kinematic and electromyography data in ten adult cats (5 males, 5 females) before and at weeks 1-2 and 7-8 after a lateral mid-thoracic hemisection on the right side of the cord while they negotiated obstacles of three different heights. Intact cats always cleared obstacles without contact. At weeks 1-2 after hemisection, the ipsilesional right hindlimb contacted obstacles in ∼50% of trials, triggering a stumbling corrective reaction or absent responses, which we termed Other. When complete clearance occurred, we observed exaggerated ipsilesional hindlimb flexion when crossing the obstacle with contralesional left limbs leading. At weeks 7-8 after hemisection, the proportion of complete clearance increased, Other responses decreased, and stumbling corrective reactions remained relatively unchanged. We found redistribution of weight support after hemisection, with reduced diagonal supports and increased homolateral supports, particularly on the left contralesional side. The main neural strategy for complete clearance in intact cats consisted of increased knee flexor activation. After hemisection, ipsilesional knee flexor activation remained but it was insufficient or more variable as the limb approached the obstacle. Intact cats also increased their speed when stepping over an obstacle, an increase that disappeared after hemisection. The increase in complete clearance over time after hemisection paralleled the recovery of muscle activation patterns or new strategies. Our results suggest partial recovery of anticipatory control through neuroplastic changes in the locomotor control system.
Significance statement:
Most spinal cord injuries (SCI) are incomplete and people can recover some walking functions. However, the main challenge for people with SCI that do recover a high level of function is to produce a gait that can adjust to everyday occurrences, such as turning, stepping over an obstacle, etc. Here, we use the cat model to answer two basic questions: How does an animal negotiate an obstacle after an incomplete SCI and why does it fail to safely clear it? We show that the inability to clear an obstacle is because of improper activation of muscles that flex the knee. Animals recover a certain amount of function thanks to new strategies and changes within the nervous system.
... During real locomotion, when the foot dorsum hits an obstacle during the swing phase, cutaneous inputs trigger a functional reflex response called the stumbling corrective reaction to rapidly move the leg away from and over the stimulus or obstacle. The stumbling corrective reaction has been observed in many mammals, including intact and spinal-transected cats (Forssberg et al., 1977;Prochazka et al., 1978;Forssberg, 1979;Wand et al., 1980;Buford and Smith, 1993;Quevedo et al., 2005a;McVea and Pearson, 2007) and mice (Mayer and Akay, 2018), as well as in human infants, adults and the elderly (Schillings et al., 1996(Schillings et al., , 2000(Schillings et al., , 2005Van Wezel et al., 1997;Zehr et al., 1997;Haridas and Zehr, 2003;Lam et al., 2003;Potocanac et al., 2016). ...
Coordinating the four limbs is an important feature of terrestrial mammalian locomotion. When the foot dorsum contacts an obstacle, cutaneous mechanoreceptors send afferent signals to the spinal cord to elicit coordinated reflex responses in the four limbs to ensure dynamic balance and forward progression. To determine how the locomotor pattern of all four limbs changes in response to a sensory perturbation evoked by activating cutaneous afferents from one hindlimb, we electrically stimulated the superficial peroneal (SP) nerve with a relatively long train at four different phases (mid-stance, stance-to-swing transition, mid-swing, and swing-to-stance transition) of the hindlimb cycle in seven adult cats. The largest functional effects of the stimulation were found at mid-swing and at the stance-to-swing transition with several changes in the ipsilateral hindlimb, such as increased activity in muscles that flex the knee and hip joints, increased joint flexion and toe height, increased stride/step lengths and increased swing duration. We also observed several changes in support periods to shift support from the stimulated hindlimb to the other three limbs. The same stimulation applied at mid-stance and the swing-to-stance transition produced more subtle changes in the pattern. We observed no changes in stride and step lengths in the ipsilateral hindlimb with stimulation in these phases. We did observe some slightly greater flexions at the knee and ankle joints with stimulation at mid-stance and a reduction in double support periods and increase in triple support. Our results show that correcting or preventing stumbling involves functional contributions from all four limbs.
... Limb-specific memories might be stored for the lead and trail legs but can be affected by the sensorimotor information in the other limb. Indeed, proprioceptive feedback and an efferent copy signal provided when stepping over an obstacle with the lead limb enhanced memory of the obstacle height that was recalled in the trail limb movement compared with the case in which only visual information was available (McVea and Pearson, 2007;McVea et al., 2009;Shinya et al., 2012). The present study suggested that neural resources of limbspecific motor memories for obstacle crossing movements in lead and trail legs were shared based on visual input regarding the interaction between obstacle properties and limb movements. ...
When walking around a room or outside, we often need to negotiate external physical objects, such as walking up stairs or stepping over an obstacle. In previous studies on obstacle avoidance, lead and trail legs in humans have been considered to be controlled independently on the basis of visual input regarding obstacle properties. However, this perspective has not been sufficient because the influence of visuomotor transformation in the lead leg on the trail leg has not been fully elucidated due to technical limitations in the experimental tasks of stepping over physical obstacles. In this study, we investigated how visuomotor transformation in the lead leg affected movement trajectories in the trail leg using a visually guided task of crossing over a virtual obstacle. Trials for stepping over a physical obstacle were established followed by visually guided tasks in which cursors corresponding to the subject’s lead and trail limb toe positions were displayed on a head-mounted display apparatus. Subjects were instructed to manipulate the cursors so that they precisely crossover a virtual obstacle. In the middle of the trials, the vertical displacement of the cursor only in the lead leg was reduced relative to the actual toe movement during one or two consecutive trials. This visuomotor perturbation resulted in higher elevation not only in the lead limb toe position but also in the trail limb toe trajectories, and then the toe heights returned to the baseline in washout trials, indicating that the visuomotor transformation for obstacle avoidance in the lead leg affects the trail leg trajectory. Taken together, neural resources of limb-specific motor memories for obstacle crossing movements in the lead and trail legs can be shared based on visual input regarding obstacle properties.
... The ability to traverse large obstacles is crucial for both animals and robots. Many insects [1][2][3][4][5], reptiles [6][7][8][9], small birds [10], and small mammals [11] encounter large bump-like obstacles in their natural habitats, such as branch litter and roots on the rainforest floor and rock beds near river or in desert environments (figure 1), which they need to traverse in order to survive [12,13]. Similarly, the speed at which search-and-rescue robots [14] traverse terrains such as building rubble and landslides where large bumps are abundant (figure 1) could determine the success of failure of a critical mission [15]. ...
Small animals and robots must often rapidly traverse large bump-like obstacles when moving through complex 3-D terrains, during which, in addition to leg-ground contact, their body inevitably comes into physical contact with the obstacles. However, we know little about the performance limits of large bump traversal and how body-terrain interaction affects traversal. To address these, we challenged the discoid cockroach and an open-loop six-legged robot to dynamically run into a large bump of varying height to discover the maximal traversal performance, and studied how locomotor modes and traversal performance are affected by body-terrain interaction. Remarkably, during rapid running, both the animal and the robot were capable of dynamically traversing a bump much higher than its hip height (up to 4 times the hip height for the animal and 3 times for the robot, respectively) at traversal speeds typical of running, with decreasing traversal probability with increasing bump height. A stability analysis using a novel locomotion energy landscape model explained why traversal was more likely when the animal or robot approached the bump with a low initial body yaw and a high initial body pitch, and why deflection was more likely otherwise. Inspired by these principles, we demonstrated a novel control strategy of active body pitching that increased the robot maximal traversable bump height by 75%. Our study is a major step in establishing the framework of locomotion energy landscapes to understand locomotion in complex 3-D terrains.
... The primary aim of this study was to test the hypothesis that there is flexibility in the time that information about obstacle size can be sampled. Studies of both humans [1] and non-human animals [19,20] have demonstrated that walkers are able to retain knowledge of obstacle characteristics such as height for an extended period of time. Although memories of obstacle dimensions must be formed during the current approach to be useful [21], this could enable walkers to sample information about obstacle size in advance and use that information to properly elevate the feet during obstacle crossing. ...
The present study investigated differences in the pickup of information about the size and location of an obstacle in the path of locomotion. The main hypothesis was that information about obstacle location is most useful when it is sampled at a specific time during the approach phase, whereas information about obstacle size can be sampled at any point during the last few steps. Subjects approached and stepped over obstacles in a virtual environment viewed through a head-mounted display. In Experiment 1, a horizontal line on the ground indicating obstacle location was visible throughout the trial while information about obstacle height and depth was available only while the subject was passing through a viewing window located at one of four locations along the subject’s path. Subjects exhibited more cautious behavior when the obstacle did not become visible until they were within one step length, but walking behavior was at most weakly affected in the other viewing window conditions. In Experiment 2, the horizontal line indicating obstacle location was removed, such that no information about the obstacle (size or location) was available outside of the viewing window. Subjects adopted a more cautious strategy compared to Experiment 1 and differences between the viewing window conditions and the full vision control condition were observed across several measures. The differences in walking behavior and performance across the two experiments support the hypothesis that walkers have greater flexibility in when they can sample information about obstacle size compared to location. Such flexibility may impact gaze and locomotor control strategies, especially in more complex environments with multiple objects and obstacles.
... The ability to traverse large obstacles is crucial for both animals and robots. Many insects [1][2][3][4][5], reptiles [6][7][8][9], small birds [10], and small mammals [11] encounter large bump-like obstacles in their natural habitats, such as branch litter and roots on the rainforest floor and rock beds near river or in desert environments (figure 1), which they need to traverse in order to survive [12,13]. Similarly, the speed at which search-and-rescue robots [14] traverse terrains such as building rubble and landslides where large bumps are abundant (figure 1) could determine the success of failure of a critical mission [15]. ...
Small animals and robots must often rapidly traverse large bump-like obstacles when moving through complex 3D terrains, during which, in addition to leg-ground contact, their body inevitably comes into physical contact with the obstacles. However, we know little about the performance limits of large bump traversal and how body-terrain interaction affects traversal. To address these, we challenged the discoid cockroach and an open-loop six-legged robot to dynamically run into a large bump of varying height to discover the maximal traversal performance, and studied how locomotor modes and traversal performance are affected by body-terrain interaction. Remarkably, during rapid running, both the animal and the robot were capable of dynamically traversing a bump much higher than its hip height (up to 4 times the hip height for the animal and 3 times for the robot, respectively) at traversal speeds typical of running, with decreasing traversal probability with increasing bump height. A stability analysis using a novel locomotion energy landscape model explained why traversal was more likely when the animal or robot approached the bump with a low initial body yaw and a high initial body pitch, and why deflection was more likely otherwise. Inspired by these principles, we demonstrated a novel control strategy of active body pitching that increased the robot's maximal traversable bump height by 75%. Our study is a major step in establishing the framework of locomotion energy landscapes to understand locomotion in complex 3D terrains.
... Given the intricate coordination of the lead and trail limbs, it is plausible that lead limb foot clearance over the obstacle influenced trail limb foot clearance. Research in both cats (McVea and Pearson 2007) and humans (Lajoie et al. 2012) has shown that movement of the lead/forelimbs over an obstacle establishes motor memories that can be maintained for minutes (tested up to 2 min for humans and 10 min in cats). These motor memories of the lead/forelimbs crossing the obstacle are subsequently used to guide trail/hindlimb foot clearance. ...
... 2a,b). These findings are consistent with those of previous studies of cats 12,13,24 . We therefore believe that the delayed obstacle avoidance task can measure the capability of working memory to guide the leading hindlimb of mice stepping over an obstacle. ...
Memory function deficits induced by Alzheimer's disease (AD) are believed to be one of the causes of an increased risk of tripping in patients. Working memory contributes to accurate stepping over obstacles during locomotion, and AD-induced deficits of this memory function may lead to an increased risk of contact with obstacles. We used the triple transgenic (3xTg) mice to examine the effects of memory deficits in terms of tripping and contact with obstacles. We found that the frequency of contact of the hindlimbs during an obstacle avoidance task increased significantly in 10-13 month-old 3xTg (Old-3xTg) mice compared with control mice. However, no changes in limb kinematics during unobstructed locomotion or successful obstacle avoidance locomotion were observed in the Old-3xTg mice. Furthermore, we found that memory-based movements in stepping over an obstacle were impaired in these mice. Our findings suggest that working memory deficits as a result of AD are associated with an increased risk of tripping during locomotion.
... This information becomes particularly relevant when navigating through an unknown or irregular terrain. For cats (McVea and Pearson, 2007;McVea et al., 2009;Wilkinson and Sherk, 2005) and humans (Mohagheghi et al., 2004;Patla and Vickers, 2003) it is known that targeting of leg movements is primarily mediated by visual information that is captured on average two steps ahead. Likewise Niven and colleagues showed that locusts visually target their front legs towards the position of a ladder rung and information about the position of the rung is acquired before leg swing is initiated (Niven et al., 2010). ...
In its natural habitat, Carausius morosus climbs on the branches of bushes and trees. Previous work suggested that stick insects perform targeting movements with their hind legs to find support more easily. It has been assumed that the animals use position information from the anterior legs to control the touchdown position of the ipsilateral posterior legs. Here we address the questions if not only the hind but also the middle leg performs targeting, and if targeting is still present in a walking animal when influences of mechanical coupling through the ground are removed. If this were the case, it would emphasize the role of underlying neuronal mechanisms. We studied whether targeting occurred in both legs, when the rostral neighboring leg, i.e. either middle- or front leg, was placed at defined positions relative to the body, and analyzed targeting precision for dependency on the targeted position. Under these conditions, the touchdown positions of the hind legs show correlation to the position of the middle leg parallel and perpendicular to the body axis while only weak correlation exists between the middle and front legs, and only in parallel to the body axis. In continuously walking tethered animals targeting accuracy of hind and middle legs parallel to the body axis was barely different. However, targeting became significantly more accurate perpendicular to the body axis. Our results suggest that a neural mechanism exists for controlling the touchdown position of the posterior leg but that the strength of this mechanism is segment-specific and dependent on the behavioral context in which it is used.
... In contrast, McVea and Pearson (2007b) have argued that in cats, the obstacle memory is formed based on feedback and/or an efference copy signal from the passage of the forelimbs over the obstacle and that visual information regarding the obstacle details is not necessary. Preventing cats from stepping over an initially visible obstacle with the forelimbs before stepping Fig. 6. ...
Stepping over obstacles requires vision to guide the leading leg, but direct visual information is not available to guide the trailing leg. The neural mechanisms for establishing a stored obstacle representation and thus facilitating the trail leg trajectory in humans are unknown. Twenty-four subjects participated in one of three experiments, which were designed to investigate the contribution of visual, proprioceptive, and efference copy signals. Subjects stepped over an obstacle with their lead leg, stopped, and straddled the obstacle for a delay period before stepping over it with their trail leg while toe elevation was recorded. Subsequently, we calculated maximum toe elevation and toe clearance. First, we found that subjects could accurately scale trail leg toe elevation and clearance, despite straddling an obstacle for up to 2 min, similar to quadrupeds. Second, we found that when the lead leg was passively moved over an obstacle (eliminating an efference copy signal and altering proprioception) without vision, trail leg toe elevation and clearance were reduced, and variability increased compared with when subjects actively moved their lead leg. Trail leg toe measures returned to normal when vision was provided during the passive manipulation. Finally, we found that altering lead leg proprioceptive feedback by adding mass to the ankle had no effect on trail leg toe measures. Taken together, our results suggest that humans can store a neural representation of obstacle properties for extended periods of time and that vision appears to be sufficient in this process to guide trail leg trajectory.
... Rubrospinal neurons modulate their activity during obstructed locomotion in the cat (Lavoie and Drew 2002). There is a corticorubral projection from posterior parietal cortex (Fanardjian and Papoyan 1997), where neurons also modulate their activity during adaptive locomotion (McVea and Pearson 2007;Pearson and Gramlich 2010), that would be spared by the M1 lesion. Our findings point to principal subcortical control of the trajectory and shared, parallel, cortical, and subcortical control of response decision making and scaling (at least in the WT mouse). ...
In voluntary control, supraspinal motor systems select the appropriate response and plan movement mechanics to match task constraints. Spinal circuits translate supraspinal drive into action. We studied the interplay between motor cortex (M1) and spinal circuits during voluntary movements in wild-type (WT) mice and mice lacking the α2-chimaerin gene (Chn1(-/-)), necessary for ephrinB3-EphA4 signaling. Chn1(-/-) mice have aberrant bilateral corticospinal systems, aberrant bilateral-projecting spinal interneurons, and disordered voluntary control because they express a hopping gait, which may be akin to mirror movements. We addressed three issues. First, we determined the role of the corticospinal system in adaptive control. We trained mice to step over obstacles during treadmill locomotion. We compared performance before and after bilateral M1 ablation. WT mice adaptively modified their trajectory to step over obstacles, and M1 ablation increased substantially the incidence of errant steps over the obstacle. Chn1(-/-) mice randomly stepped or hopped during unobstructed locomotion but hopped over the obstacle. Bilateral M1 ablation eliminated this obstacle-dependent hop selection and increased forelimb obstacle contact errors. Second, we characterized the laterality of corticospinal action in Chn1(-/-) mice using pseudorabies virus retrograde transneuronal transport and intracortical microstimulation. We showed bilateral connections between M1 and forelimb muscles in Chn1(-/-) and unilateral connections in WT mice. Third, in Chn1(-/-) mice, we studied adaptive responses before and after unilateral M1 ablation. We identified a more important role for contralateral than ipsilateral M1 in hopping over the obstacle. Our findings suggest an important role for M1 in the mouse in moment-to-moment adaptive control, and further, using Chn1(-/-) mice, a role in mediating task-dependent selection of mirror-like hopping movements over the obstacle. Our findings also stress the importance of subcortical control during adaptive locomotion because key features of the trajectory remained largely intact after M1 ablation.
... In invertebrates, mechanosensory feedback is usually integrated into motor control at a local level (Büschges 2005), although clearly this information about the animals surroundings can also be used by the brain to alter behavior (Watson et al. 2002a, b;Ridgel and Ritzmann 2005;Ritzmann et al. 2005;Ridgel et al. 2007). Detection of obstacles beyond the range of mechanosensors requires remote sensing, such as vision (for example in cats McVea and Pearson 2007) or active sensing systems, such as echolocation and electroreception (Nelson and MacIver 2006). The integration of such information into a three-dimensional framework relative to the animals own body is computationally more challenging than local sensing and it is generally supported by the development of specialized brain structures (e.g., optic lobes). ...
Animal locomotion is produced by co-coordinated patterns of motor activity that are generally organized by central pattern generators and modified by sensory feedback. Animals with remote sensing can anticipate obstacles and make adjustments in their gait to accommodate them. It is largely unknown how animals that rely on touch might use such information to adjust their gait. One possibility is immediate (reflexive) change in motor activity. Elongated animals, however, might modulate movements by passing information from anterior to posterior segments. Using the caterpillar Manduca sexta we examined the movements of the most anterior abdominal prolegs as they approached an obstacle. The first pair of prolegs anticipated the obstacle by lifting more quickly in the earliest part of the swing phase: the caterpillar had information about the obstacle at proleg lift-off. Sometimes the prolegs corrected their trajectory mid-step. Removal of sensory hairs on the stepping leg did not affect the early anticipatory movements, but did change the distance at which the mid-step corrections occurred. We conclude that anterior sensory information can be passed backwards and used to modulate an ongoing crawl. The local sensory hairs on each body segment can then fine-tune movements of the prolegs as they approach an obstacle.
... At its core lies a psychological phenomenon which was described only a few years earlier: Active engagement of the brain provides learning capabilities which are difficult or impossible to achieve by passive observation alone 2,3 . This phenomenon is today known as the generation effect 4 and can also be observed in animals such as monkeys 5 , cats 6 or fruit flies 7 . Despite the impact learning-by-doing has on society and the ubiquity of the generation effect, the mechanism by which activity enhances passive learning is unknown. ...
At the heart of learning-by-doing lies a well-known psychological phenomenon: information will be remembered better if it is actively generated rather than passively read or heard. First described in humans, this generation effect can also be observed in various animal models. However, the neurobiological mechanisms underlying the generation effect are unknown. Here we show that two reciprocal interactions between its active and passive components contribute to the generation effect in flies. One interaction consists of the active (skill-learning) component facilitating the passive (fact-learning) component. Fact-learning, on the other hand, inhibits skill-learning. Experiments with adenylyl cyclase I deficient _rutabaga_ mutant flies revealed that the fact- but not the skill-learning component requires this evolutionarily conserved learning gene. Using mushroom-body deficient transgenic flies we observed that the mushroom-bodies mediate the inhibition of skill-learning. This inhibition also enables generalization and prevents premature habit formation. Extended training in wildtype flies produced a phenocopy of mushroom-body impaired flies, such that generalization was abolished and goal-directed actions were transformed into habitual responses. Thus, our results identify various neural processes underlying learning-by-doing, delineate some of their synergisms and provide a framework for further dissecting them in a genetically tractable model system.
... Most importantly, the present test provides a challenge to ongoing locomotion because the spacing of the rungs is varied. The variation can be used to challenge ongoing walking and/or memory of the stepping patterns (McVea and Pearson,2007). ...
Progress in the development of animal models for/stroke, spinal cord injury, and other neurodegenerative disease requires tests of high sensitivity to elaborate distinct aspects of motor function and to determine even subtle loss of movement capacity. To enhance efficacy and resolution of testing, tests should permit qualitative and quantitative measures of motor function and be sensitive to changes in performance during recovery periods. The present study describes a new task to assess skilled walking in the rat to measure both forelimb and hindlimb function at the same time. Animals are required to walk along a horizontal ladder on which the spacing of the rungs is variable and is periodically changed. Changes in rung spacing prevent animals from learning the absolute and relative location of the rungs and so minimize the ability of the animals to compensate for impairments through learning. In addition, changing the spacing between the rungs allows the test to be used repeatedly in long-term studies. Methods are described for both quantitative and qualitative description of both fore- and hindlimb performance, including limb placing, stepping, co-ordination. Furthermore, use of compensatory strategies is indicated by missteps or compensatory steps in response to another limb's misplacement.
Every day walking consists of frequent voluntary modifications in the gait pattern to negotiate obstacles. After spinal cord injury, stepping over an obstacle becomes challenging. Stepping over an obstacle requires sensorimotor transformations in several structures of the brain, including the parietal cortex, premotor cortex, and motor cortex. Sensory information and planning are transformed into motor commands, which are sent from the motor cortex to spinal neuronal circuits to alter limb trajectory, coordinate the limbs, and maintain balance. After spinal cord injury, bidirectional communication between the brain and spinal cord is disrupted and animals, including humans, fail to voluntarily modify limb trajectory to step over an obstacle. Therefore, in this review, we discuss the neuromechanical control of stepping over an obstacle, why it fails after spinal cord injury, and how it recovers to a certain extent.
Following incomplete spinal cord injury in animals, including humans, substantial locomotor recovery can occur. However, functional aspects of locomotion, such as negotiating an obstacle remains challenging. We collected kinematic and electromyography data in ten adult cats before and at weeks 1-2 and 7-8 after a lateral mid-thoracic hemisection while they negotiated obstacles of three different heights. Intact cats always cleared obstacles without contact. At weeks 1-2 after hemisection, the ipsilesional hindlimb contacted obstacles in ~50% of trials, triggering a stumbling corrective reaction or lack of response. When complete clearance occurred, we observed exaggerated ipsilesional hindlimb flexion when it crossed the obstacle with the contralesional limbs leading. At weeks 7-8 after hemisection, complete clearance increased in favor of absent responses while the proportion of stumbling corrective reactions remained relatively the same. We found redistribution of weight support after hemisection, with reduced diagonal supports and increased homolateral supports, particularly on the contralesional side. The main neural strategy for complete clearance in intact cats consisted of increased activation of muscles that flex the knee. After hemisection, knee flexor activation remained but it was insufficient or more variable as the limb approached the obstacle. Intact cats also increased their speed when stepping over an obstacle, an increase that disappeared after hemisection. The increase in complete clearance over time after hemisection paralleled the recovery of muscle activation patterns or new strategies. Our results suggest partial recovery of anticipatory control through neuroplastic changes in the locomotor control system.
Obstacle crossing requires visuospatial working memory to guide the trailing leg trajectory when vision in unavailable. Visuospatial working memory, as assessed with neuropsychological tests, declines with age, however, this remains to be investigated functionally in obstacle crossing. There is also evidence that visuospatial encoding during a secondary task interferes with balance control during stepping and walking in older people. Here, we studied the interaction effects of age by delay (study 1) and age by secondary visuospatial task (study 2) conditions on obstacle clearance in a visuospatial working memory -guided obstacle crossing task. Healthy young adults aged 19 to 36 years (n = 20 in study 1 and n = 17 in study 2) and healthy older adults aged 66 to 83 years (n = 29 in study 1 and n = 21 in study 2) were instructed to step over an obstacle with their leading leg and straddle it for a delay period before completing the crossing with their trailing leg. In study 1, two obstacle height conditions (12 cm, 18 cm) and two delay durations (20 s, 60 s) were presented in random order. In study 2, participants were required to attend to either no secondary task (control), a visuospatial secondary (star movement) task, or a nonspatial secondary (arithmetic) task, while straddling the obstacle for a delay duration of 20 s, at obstacle heights of 12 cm and 18 cm, randomly presented. Trailing leg kinematics (mean and variability of maximum toe clearance over the obstacle) were determined via motion capture. There were no statistically significant age by delay or age by secondary task interactions. In study 1, toe clearance variability was significantly greater in young adults and increased with increasing delay duration in both groups. In study 2, compared with the control condition, toe clearance variability was significantly greater in the non-spatial secondary task condition but not in the visuospatial condition. Contrary to our hypotheses, these findings suggest that young and older adults alike can store an obstacle representation via visuospatial working memory for durations of at least 60 s and use this information to safely scale their trailing leg over an obstacle. However, the increase in trailing leg toe clearance variability with delay duration suggests that obstacle representation starts to deteriorate even within the first 20 s regardless of age. The finding that undertaking a concurrent arithmetic task impaired visuospatial working memory-guided obstacle clearance suggests a potential increased risk of tripping during obstacle crossing while dual-tasking in both young and older people.
For older adults especially, to perform everyday activities safely, adaptive locomotion that adjusts basic locomotion pattern according to the environmental features is critical. It is unknown, however, whether their locomotor patterns can be modified when there are subtle environmental changes. We examined adaptive limb movements, focusing on obstacle avoidance and age-related changes during such situations. Younger (102, with a mean age of 27.5 years) and older (101, with a mean age of 78.3 years) participants walked across one obstacle (150 mm height) four different times. The obstacles were then covertly raised or lowered by 10% of the baseline obstacle height (i.e., 165 mm for ascending and 135 mm for descending conditions), and participants were asked to repeat the activity. We measured leading and trailing foot clearances, the vertical distances between toe tips and the upper edge of the obstacle. In the ascending condition, both groups adjusted and raised their limb clearance according to the obstacle height change. Alternatively, foot clearance of the leading limb for the lowered obstacle did not change among the older adults, whereas it changed in the young adults (lowered their clearance). No changes were observed in the trailing foot clearance for the descending conditions in either age group. Our results suggest that when facing environmental changes that compromise safe mobility, individuals can adapt leading limb movement based on subtle environmental changes, irrespective of age. In case of other changes (i.e., in low-risk situations), however, the ability of adaptive locomotion may be affected by aging.
Once an animal steps over an obstacle with its forelimbs, the obstacle is no longer visible. The posterior parietal cortex appears to maintain an obstacle representation in working memory through phasic and sustained activity to allow for appropriate hindlimb elevation.
A working memory of obstacles is essential for navigating complex, cluttered terrain. In quadrupeds, it has been proposed that parietal cortical areas related to movement planning and working memory may be important for guiding the hindlegs over an obstacle previously cleared by the forelegs. To test this hypothesis, parietal areas 5 and 7 were reversibly deactivated in walking cats. The working memory of an obstacle was assessed in both a visually dependent and tactilely dependent paradigm. Reversible bilateral deactivation of area 5, but not area 7, altered hindleg stepping in a manner indicating that the animals did not recall the obstacle over which their forelegs had stepped. Similar deficits were observed when area 5 deactivation was restricted to the delay during which obstacle memory must be maintained. Furthermore, partial memory recovery observed when area 5 function was deactivated and restored within this maintenance period suggests that the deactivation may suppress, but not eliminate, the working memory of an obstacle. As area 5 deactivations incurred similar memory deficits in both visual and tactile obstacle working memory paradigms, parietal area 5 is critical for maintaining the working memory of an obstacle acquired via vision or touch that is used to modify stepping for avoidance.
In complex environments, tripping over an unexpected obstacle evokes the stumbling corrective reaction, eliciting rapid limb hyperflexion to lift the leg over the obstruction. While stumbling correction has been characterized within a single limb in the cat, this response must extend to both forelegs and hindlegs for successful avoidance in naturalistic settings. Furthermore, the ability to remember an obstacle over which the forelegs have tripped is necessary for hindleg clearance if locomotion is delayed. Therefore, memory-guided stumbling correction was studied in walking cats after the forelegs tripped over an unexpected obstacle. Tactile input to only one foreleg was often sufficient in modulating stepping of all four legs when locomotion was continuous, or when hindleg clearance was delayed. When obstacle height was varied, animals appropriately scaled step height to obstacle height. As tactile input without foreleg clearance was insufficient in reliably modulating stepping, efference, or proprioceptive information about modulated foreleg stepping may be important for producing a robust, long-lasting memory. Finally, cooling-induced deactivation of parietal area 5 altered hindleg stepping in a manner indicating that animals no longer recalled the obstacle over which they had tripped. Altogether, these results demonstrate the integral role area 5 plays in memory-guided stumbling correction.
Little is known about the ability of blind people to cross obstacles after they have explored haptically their size and position. Long-term absence of vision may affect spatial cognition in the blind while their extensive experience with the use of haptic information for guidance may lead to compensation strategies. Seven blind and 7 sighted participants (with vision available and blindfolded) walked along a flat pathway and crossed an obstacle after a haptic exploration. Blind and blindfolded subjects used different strategies to cross the obstacle. After the first 20 trials the blindfolded subjects reduced the distance between the foot and the obstacle at the toe-off instant, while the blind behaved as the subjects with full vision. Blind and blindfolded participants showed larger foot clearance than participants with vision. At foot landing the hip was more behind the foot in the blindfolded condition, while there were no differences between the blind and the vision conditions. For several parameters of the obstacle crossing task, blind people were more similar to subjects with full vision indicating that the blind subjects were able to compensate for the lack of vision.
When quadrupeds stop walking after stepping over a barrier with their forelegs, the memory of barrier height and location is retained for many minutes. This memory is subsequently used to guide hind leg movements over the barrier when walking is resumed. The upslope of the initial trajectory of hind leg paw movements is strongly dependent on the initial location of the paw relative to the barrier. In this study, we have attempted to determine whether mechanical factors contribute significantly in establishing the slope of the paw trajectories by creating a four-link biomechanical model of a cat hind leg and driving this model with a variety of joint-torque profiles, including average torques for a range of initial paw positions relative to the barrier. Torque profiles for individual steps were determined by an inverse dynamic analysis of leg movements in three normal cats. Our study demonstrates that limb mechanics can contribute to establishing the dependency of trajectory slope on the initial position of the paw relative to the barrier. However, an additional contribution of neuronal motor commands was indicated by the fact that the simulated slopes of paw trajectories were significantly less than the observed slopes. A neuronal contribution to the modification of paw trajectories was also revealed by our observations that both the magnitudes of knee flexor muscle EMG bursts and the initial knee flexion torques depended on initial paw position. Previous studies have shown that a shift in paw position prior to stepping over a barrier changes the paw trajectory to be appropriate for the new paw position. Our data indicate that both mechanical and neuronal factors contribute to this updating process, and that any shift in leg position during the delay period modifies the working memory of barrier location.
Walking is the most common terrestrial form of locomotion in animals. Its great versatility and flexibility has led to many attempts at building walking machines with similar capabilities. The control of walking is an active research area both in neurobiology and robotics, with a large and growing body of work. This paper gives an overview of the current knowledge on the control of legged locomotion in animals and machines and attempts to give walking control researchers from biology and robotics an overview of the current knowledge in both fields. We try to summarize the knowledge on the neurobiological basis of walking control in animals, emphasizing common principles seen in different species. In a section on walking robots, we review common approaches to walking controller design with a slight emphasis on biped walking control. We show where parallels between robotic and neurobiological walking controllers exist and how robotics and biology may benefit from each other. Finally, we discuss where research in the two fields diverges and suggest ways to bridge these gaps.
During adaptive locomotion, vision is used to guide the lead limb; however, the individual must rely on knowledge of obstacle height and position, termed obstacle memory, to guide the trail limb. Previous research has demonstrated that visual sampling of the obstacle during approach was adequate to provide obstacle height information, but online visual update of distance to the obstacle was required to plan and implement appropriate foot placement. Our purpose was to determine whether obstacle height memory, coupled with a visible obstacle position cue, could successfully guide the foot during obstacle crossing. Subjects first stepped over an obstacle for 25 trials; then, the obstacle was removed, but its position was marked with high-contrast tape; subjects were instructed to step over the obstacle as if it was still there (termed "virtual obstacle") for 25 trials. No changes in foot placement were observed; therefore, the position cue provided salient online information to guide foot placement. Average failure rates (subject would have contacted the virtual obstacle if it was present) were 9 and 47 % (lead and trail limb, respectively). Therefore, action was impaired for both limbs when guided by obstacle height memory, but action was impaired to a greater extent for the trail limb. Therefore, viewing the obstacle during approach appears to facilitate the memory needed to guide obstacle crossing, particularly for the trail limb. This is likely because the lead limb is visible in the peripheral visual field during crossing, but the trail limb is not.
A concept emerging from recent studies on obstacle avoidance in quadrupeds is that working memory of the height of an obstacle established by visual information is enhanced by motor interactions with the obstacle. In this investigation, we found that this concept is valid in adult humans when viewing and walking up stairs. The main finding was that the memory of the height of stairs was enhanced when information about stair height was gained by walking up a short flight of stairs compared to when information about stair height was gained by vision alone. By measuring the maximum toe clearance when subjects step onto a stair, we observed that maximum toe clearance increased after diverting vision from the stair for a few seconds prior to stepping. Most of this increase occurred within a 2-s period between diverting vision from the stair and initiating the step. By contrast, this increase in maximum toe clearance after diverting vision from a stair was significantly reduced after subjects walked up two stairs prior to stepping onto a stair without vision. This reduction persisted for delays as long as 10 s between diverting vision from the stair and initiating the step. In four of twelve subjects, the maximum toe clearance after these long periods without vision of the stair was close to the value when steps were made with full vision of the stairs.
In quadrupeds, a unique form of memory is used to guide the hind legs over barriers that have already been stepped over by the forelegs. This memory is very long-lasting (many minutes), incorporates precise information about the size and position of the barrier relative to the hind legs, and is updated as the animal steps sequentially across a barrier. Recent findings from electrophysiological and lesion studies have revealed that neuronal systems in the parietal cortex are necessary for establishing the long-lasting feature of the memory and may be involved in representing the current position of the barrier relative to the moving body. We hypothesize that the latter involves the modulation of activity in neuronal systems in the posterior parietal cortex by efference copy signals of motor commands for stepping and by sensory signals from muscle proprioceptors. We propose that motor pattern generation for walking occurs within a framework of a body schema that constantly informs pattern generating networks about the geometry of the body and the location of near objects relative to the body.
Many animals rely on vision for navigating through complex environments and for avoiding specific obstacles during locomotion. Navigation and obstacle avoidance are tasks that depend on gathering information about the environment by vision and using this information at later times to guide limb and body movements. Here we review studies demonstrating the use of short-term visual memory during walking in humans and cats. Our own investigations have demonstrated that cats have the ability to retain a memory of an obstacle they have stepped over with the forelegs for many minutes and to use this memory to guide stepping of the hindlegs to avoid the remembered obstacle. A brain region that may be critically involved in the retention of memories of the location of obstacles is the posterior parietal cortex. Recordings from neurons in area 5 in the posterior parietal cortex in freely walking cats have revealed the existence of neurons whose activity is strongly correlated with the location of an obstacle relative to the body. How these neurons might be used to regulate motor commands remains to be established. We believe that studies on obstacle avoidance in walking cats have the potential to significantly advance our understanding of visuo-motor transformations. Current knowledge about the brain regions and pathways underlying visuo-motor transformations during walking are reviewed.
A number of internal model concepts are now widespread in neuroscience and cognitive science. These concepts are supported by behavioral, neurophysiological, and imaging data; furthermore, these models have had their structures and functions revealed by such data. In particular, a specific theory on inverse dynamics model learning is directly supported by unit recordings from cerebellar Purkinje cells. Multiple paired forward inverse models describing how diverse objects and environments can be controlled and learned separately have recently been proposed. The 'minimum variance model' is another major recent advance in the computational theory of motor control. This model integrates two furiously disputed approaches on trajectory planning, strongly suggesting that both kinematic and dynamic internal models are utilized in movement planning and control
Spatio-temporal gaze behaviour patterns were analysed as normal participants wearing a mobile eye tracker approached and stepped over obstacles of varying height in the travel path. We examined the frequency and duration of three types of gaze fixation with respect to the participants' stepping patterns: obstacle fixation (ObsFix); travel fixation (TravFix) (when the gaze is stable and travelling at the speed of whole body) and fixation in the 4-6m region (Fix4-6). During the approach phase to the obstacle, participants fixated on the obstacle for approximately 20% of the travel time. Only Fix4-6 duration was modulated as a function of obstacle height by regulating the frequency and reflected the increased time needed for detection of the small low contrast obstacle in the travel path. Frequency of ObsFix increased significantly as a function of obstacle height and reflected visuo-motor transformation needed for limb elevation control. Participants did not fixate on the obstacle as they were stepping over, but did the planning in the steps before. TravFix duration and frequency was constant while Fix4-6 duration was higher in the step before and step over the obstacle reflecting visual search of the landing area for the lead limb following obstacle avoidance. These results clearly show that obstacle information provided by vision is used in a feed-forward rather than on-line control mode to regulate locomotion. Information about self-motion acquired from optic flow during TravFix can be used to control velocity of locomotion.
Skilled motor behavior relies on the brain learning both to control the body and predict the consequences of this control. Prediction turns motor commands into expected sensory consequences, whereas control turns desired consequences into motor commands. To capture this symmetry, the neural processes underlying prediction and control are termed the forward and inverse internal models, respectively. Here, we investigate how these two fundamental processes are related during motor learning. We used an object manipulation task in which subjects learned to move a hand-held object with novel dynamic properties along a prescribed path. We independently and simultaneously measured subjects' ability to control their actions and to predict their consequences. We found different time courses for predictor and controller learning, with prediction being learned far more rapidly than control. In early stages of manipulating the object, subjects could predict the consequences of their actions, as measured by the grip force they used to grasp the object, but could not generate appropriate actions for control, as measured by their hand trajectory. As predicted by several recent theoretical models of sensorimotor control, our results indicate that people can learn to predict the consequences of their actions before they can learn to control their actions.
This study explored the duration of cats' working memory for hidden objects. Twenty-four cats were equally divided into four groups, which differed according to the type of visual cues displayed on and/or around the hiding boxes. During eight sessions, the four groups of cats were trained to locate a desirable object hidden behind one of the four boxes placed in front of them. Then, the cats were tested with retention intervals of 0, 10, 30 and 60 s. Results revealed no significant differences between the groups during training or testing. In testing, the cats' accuracy to locate the hidden object rapidly declined between 0 and 30 s but remained higher than chance with delays of up to 60 s. The analysis of errors also indicated that the cats searched as a function of the proximity of the target box and were not subjected to intertrial proactive interference. This experiment reveals that the duration of cats' working memory for disappearing objects is limited and the visual cues displayed on and/or around the boxes do not help the cats to memorize a hiding position. In discussion, we explore why the duration of cats' working memory for disappearing objects rapidly declined and compare these finding with those from domestic dogs. The irrelevance of visual cues displayed on and around the hiding boxes on cats' retention capacity is also discussed.
We examined the ways in which memories of previously seen obstacles can alter the stepping of walking cats. Cats were paused after the forelegs, but not the hindlegs, had stepped over an obstacle. Near the beginning of a variable delay period, the obstacle was lowered. On the subsequent step, the path of the hindlegs allowed us to make inferences about whether the memory of the obstacle was influencing leg movements. We present two main findings. First, the memory of the obstacle persisted for the duration that the animal straddled the original location of the obstacle. In one instance, this interval was 10 min. Second, this memory includes information regarding the size and position of the obstacle relative to the animal. This information is used to plan foot placement and to redirect the step in mid-swing to avoid the previous position of the obstacle.
This review focuses on advances in our understanding of the roles played by vision in the control of human locomotion. Vision is unique in its ability to provide information about near and far environment almost instantaneously: this information is used to regulate locomotion on a local level (step by step basis) and a global level (route planning). Basic anatomy and neurophysiology of the sensory apparatus, the neural substrate involved in processing this visual input, descending pathways involved in effecting control and mechanisms for controlling gaze are discussed. Characteristics of visual perception subserving control of locomotion include the following: (a) intermittent visual sampling is adequate for safe travel over various terrains: (b) information about body posture and movement from the visual system is given higher priority over information from the other two sensory modalities; (c) exteroceptive information about the environment is used primarily in a feedforward sampled control mode rather than on-line control mode; (d) knowledge acquired through past experience influences the interpretation of the exteroceptive information; (e) exproprioceptive information about limb position and movement is used on-line to fine tune the swing limb trajectory; (f) exproprioceptive information about self-motion acquired through optic flow is used on-line in a sampled controlled mode. Characteristics of locomotor adaptive strategies are: (a) most adaptive strategies can be implemented successfully in one step cycle provided the attention is biased towards the visual cues: only steering has to be planned in the previous step; (b) stability requirements constrain the selection of specific avoidance strategies; (c) response is not localized to a joint or limb: it is global, complex and task specific; (d) response characteristics are dependent upon available response time; (e) effector system dynamics are exploited by the control system to simplify and effectively control swing limb trajectory. Effects of various visual deficits on adaptive control are briefly discussed.
In his seminal article, Gibson (1958/this issue) laid the foundation for understanding the visuomotor transformations necessary for adaptive locomotor behavior. In this article, I review, the work on visual control of locomotion done in our lab in Waterloo and discuss some new experiments. The major findings are put forward as three new postulates. Postulate 1: Visual exteroceptive information about the environment is used in a sampled, feed-forward mode to control locomotion. Postulate 2: Visual exproprioceptive information about the posture and movement of the lower limb is used in a sampled, online mode to control the swing phase trajectory. Postulate 3: Visual exproprioceptive information about self-motion is used in a sampled, online mode to control locomotion. Our results in most cases provide empirical support for the many insightful postulates of Gibson, in some cases amplify his statements, and in a few cases add to our understanding of how we get about by vision.
Based on theoretical and computational studies it has been suggested that the central nervous system (CNS) internally simulates the behaviour of the motor system in planning, control and learning. Such an internal “forward” model is a representation of the motor system that uses the current state of the motor system and motor command to predict the next state. We will outline the uses of such internal models for solving several fundamental computational problems in motor control and then review the evidence for their existence and use by the CNS. Finally we speculate how the location of an internal model within the CNS may be identified. Copyright © 1996 Elsevier Science Ltd.
A number of internal model concepts are now widespread in neuroscience and cognitive science. These concepts are supported by behavioral, neurophysiological, and imaging data; furthermore, these models have had their structures and functions revealed by such data. In particular, a specific theory on inverse dynamics model learning is directly supported by unit recordings from cerebellar Purkinje cells. Multiple paired forward inverse models describing how diverse objects and environments can be controlled and learned separately have recently been proposed. The 'minimum variance model' is another major recent advance in the computational theory of motor control. This model integrates two furiously disputed approaches on trajectory planning, strongly suggesting that both kinematic and dynamic internal models are utilized in movement planning and control.
One important aspect of locomotor control is the ability of an animal to make anticipatory gait modifications to avoid obstacles, by stepping either around them or over them. This paper reviews some of the evidence that suggests that the motor cortex is one of the principal structures involved in the control of such anticipatory gait modifications in cats, in particular when they are triggered by a visual signal. Evidence for this statement is provided both from experiments in which the motor cortex has been lesioned or inactivated and from studies in which the activity of motor cortical neurones has been recorded during locomotor tasks in which visual information is required to ensure the correct positioning of the paw or an appropriate modification of the limb trajectory. Inactivation of small regions of the motor cortex with the GABA agonist muscimol results in changes in the limb trajectory so that cats hit an obstacle instead of stepping over it as they do normally. A similar disruption of the hindlimb trajectory is seen following lesions of the spinal cord at T13 that interrupt the corticospinal tract. The results from cell recording studies are complementary in that they show that the activity of many identified pyramidal tract neurones increases when the cat is required to modify the forelimb or hindlimb trajectory to step over obstacles. We suggest that the major function of this increased discharge frequency is to regulate the amplitude, duration, and temporal pattern of muscle activity during the gait modification to ensure an appropriate modification of limb trajectory. We further suggest that different groups of pyramidal tract neurones are involved in regulating the activity of groups of synergistic muscles active at different times in the gait modification. For example, some groups of pyramidal tract neurones would be involved in ensuring the appropriate and sequential activation of the muscle groups involved in the initial flexion of the elbow, while others would be active prior to the repositioning of the paw on the support surface. We discuss the possibility that the motor cortical activity seen during locomotion is the sum result of a feedforward signal, which provides visuospatial information about the environment, and feedback activity, which signals, in part, the state of the interneuronal pattern generating networks in the spinal cord. The way in which the resulting descending command may interact with the basic locomotor rhythm to produce the gait modifications is discussed.
Unifying principles of movement have emerged from the computational study of motor control. We review several of these principles and show how they apply to processes such as motor planning, control, estimation, prediction and learning. Our goal is to demonstrate how specific models emerging from the computational approach provide a theoretical framework for movement neuroscience.
Muscle, cutaneous and joint afferents continuously signal information about the position and movement of individual joints. How does the nervous system extract more global information, for example about the position of the foot in space? To study this question we used microelectrode arrays to record impulses simultaneously from up to 100 discriminable nerve cells in the L6 and L7 dorsal root ganglia (DRG) of the anaesthetized cat. When the hindlimb was displaced passively with a random trajectory, the firing rate of the neurones could be predicted from a linear sum of positions and velocities in Cartesian (x, y), polar or joint angular coordinates. The process could also be reversed to predict the kinematics of the limb from the firing rates of the neurones with an accuracy of 1-2 cm. Predictions of position and velocity could be combined to give an improved fit to limb position. Decoders trained using random movements successfully predicted cyclic movements and movements in which the limb was displaced from a central point to various positions in the periphery. A small number of highly informative neurones (6-8) could account for over 80% of the variance in position and a similar result was obtained in a realistic limb model. In conclusion, this work illustrates how populations of sensory receptors may encode a sense of limb position and how the firing of even a small number of neurones can be used to decode the position of the limb in space.
Visual guidance is often critical during locomotion. To understand how the visual system performs this function it is necessary to know what pattern of retinal image motion neurons experience. If a locomoting observer maintains an angle of gaze that is constant relative to his body, retinal image motion will resemble Gibson's (The Perception of the Visual World (1950)) well-known optic flow field. However, if a moving observer fixates and tracks a stationary feature of the environment, or shifts his gaze, retinal motion will be quite different. We have investigated gaze in cats during visually-guided locomotion. Because cats generally maintain their eyes centered in the orbits, their gaze can be monitored with reasonable accuracy by monitoring head position. Using a digital videocamera, we recorded head position in cats as they walked down a cluttered alley. For much of the time, cats maintained a downward angle of gaze that was constant relative to their body coordinates; these episodes averaged 240 ms in duration and occupied 48-71% of the total trial time. Constant gaze episodes were separated by gaze shifts, which often coincided with blinks. Only rarely did we observe instances when cats appeared to fixate and track stationary features of the alley. We hypothesize that walking cats acquire visual information primarily during episodes of constant gaze, when retinal image motion resembles Gibson's conventional optic flow field.
When an observer walks across irregular terrain, he uses vision to plan his steps. How far in advance of each step does he acquire the critical information? We trained cats to walk accurately down a cluttered alley, and then turned out the light in mid-trial. Cats usually continued to walk without error for one to four steps, indicating that they had acquired the information to guide each step well before foot contact.