**4. Future directions**

In the final sections, we will develop a theoretical basis for the importance of understanding the role of luminance contrast for precise, visually guided movements of the upper extremities. Following this, we briefly describe our next set of experiments aimed at a deeper understanding of the role of vision for motor performance in virtual environments across the lifespan.

### **4.1 Contrast sensitivity and tuning of neuronal populations**

The neural processing of sensory information is described by the tuning of neuronal populations to specific stimuli (Desimone & Duncan, 1995). Within areas of the visual cortex, groups of neurons fire in response to the presence of afferent information. This firing rate is tuned to specific aspects of the stimulus, thereby increasing the precision by which the system can differentiate visual information. Within neuronal populations, firing rates differ among neurons. Some neurons will fire constantly with the presence of a stimulus, known as tonic firing. Other neurons fire rapidly at the onset of the stimulus, and rapidly decrease activity thereafter; this is known as phasic activity. This phasic activity makes the general sensorimotor system particularly sensitive to changes in stimulation. The visual cortex is no exception. Phasic neurons located within the visual cortex are sensitive to areas within the visual scene that are actively changing. It is simple to understand this in the case of a moving object, however the border of a stationary object also has this effect. Specifically, when the eye is moving and a stationary image passes over the moving retina, the border of the stationary object causes the visual scene to abruptly change, and phasic neurons react accordingly.

Sensitivity to object borders is dependent on the visual contrast between the object and its background. Contrast is described by two characteristics, luminance (brightness) and chromaticity (color), which are processed differently in the dorsal and ventral visual streams. The two visual stream hypothesis put forth by Goodale and Milner has a wide breadth of experimental support in explaining multiple functions for visual processing (Goodale and Milner 1992 as cited in Milner & Goodale, 2008). Briefly, the ventral stream includes structures along the pathway from the visual cortex in the occipital lobe to the inferotemporal lobe. This circuit has been implicated in the use of vision for perception of the surrounding environment, allowing the conscious experience of seeing the world around us. The dorsal stream includes the pathway from the visual cortex to the posterior parietal lobe. This pathway is responsible for the visuomanual transformations that allow visual information to guide our motor system in interacting with the surrounding environment. The neuronal structure of the ventral stream allows for high spatial resolution and sensitivity to chromaticity (Wade et al., 2002). Processing of color information in the ventral stream plays a role in the perception of objects (Kleinholdermann et al., 2009; Morrone, Denti, & Spinelli, 2002). The role of color contrast in visual processing for motor output has not been clearly elucidated, but it appears the strict dichotomous notion of the

was most pronounced in middle age adults, but weakly present in all age groups. While there are many directions to head with future research, the finding regarding the possibility of performance enhancement through the manipulation of luminance contrast is one

In the final sections, we will develop a theoretical basis for the importance of understanding the role of luminance contrast for precise, visually guided movements of the upper extremities. Following this, we briefly describe our next set of experiments aimed at a deeper understanding of the role of vision for motor performance in virtual environments

The neural processing of sensory information is described by the tuning of neuronal populations to specific stimuli (Desimone & Duncan, 1995). Within areas of the visual cortex, groups of neurons fire in response to the presence of afferent information. This firing rate is tuned to specific aspects of the stimulus, thereby increasing the precision by which the system can differentiate visual information. Within neuronal populations, firing rates differ among neurons. Some neurons will fire constantly with the presence of a stimulus, known as tonic firing. Other neurons fire rapidly at the onset of the stimulus, and rapidly decrease activity thereafter; this is known as phasic activity. This phasic activity makes the general sensorimotor system particularly sensitive to changes in stimulation. The visual cortex is no exception. Phasic neurons located within the visual cortex are sensitive to areas within the visual scene that are actively changing. It is simple to understand this in the case of a moving object, however the border of a stationary object also has this effect. Specifically, when the eye is moving and a stationary image passes over the moving retina, the border of the stationary object causes the visual scene to abruptly change, and phasic neurons react accordingly.

Sensitivity to object borders is dependent on the visual contrast between the object and its background. Contrast is described by two characteristics, luminance (brightness) and chromaticity (color), which are processed differently in the dorsal and ventral visual streams. The two visual stream hypothesis put forth by Goodale and Milner has a wide breadth of experimental support in explaining multiple functions for visual processing (Goodale and Milner 1992 as cited in Milner & Goodale, 2008). Briefly, the ventral stream includes structures along the pathway from the visual cortex in the occipital lobe to the inferotemporal lobe. This circuit has been implicated in the use of vision for perception of the surrounding environment, allowing the conscious experience of seeing the world around us. The dorsal stream includes the pathway from the visual cortex to the posterior parietal lobe. This pathway is responsible for the visuomanual transformations that allow visual information to guide our motor system in interacting with the surrounding environment. The neuronal structure of the ventral stream allows for high spatial resolution and sensitivity to chromaticity (Wade et al., 2002). Processing of color information in the ventral stream plays a role in the perception of objects (Kleinholdermann et al., 2009; Morrone, Denti, & Spinelli, 2002). The role of color contrast in visual processing for motor output has not been clearly elucidated, but it appears the strict dichotomous notion of the

**4.1 Contrast sensitivity and tuning of neuronal populations** 

particular area of interest.

**4. Future directions** 

across the lifespan.

two visual stream hypothesis may be overly rigid. Recent investigation has shown that for simple eye movements and pointing tasks, color information can be used to guide movement (White, Kerzel, & Gegenfurtner, 2006). Pisella, Arzi, and Rossetti (1998) studied the ability of humans to utilize color information to quickly update their movements in a perturbation paradigm. While movement reorganization was possible utilizing only color information, the results showed a distinct slowing of movement reorganization. Brenner and Smeets (2004) also studied a similar paradigm, finding that color could in fact be utilized rather quickly for task reorganization; however, they still showed a minor slowing compared with movement reorganization based on luminance information. Luminance contrast, while also important in perception, may have more direct implications for motor output. Motion sensitivity is dependent on contrast sensitivity and motion sensitivity is a hallmark of the neuronal structure of the dorsal stream (Born & Bradley, 2005). Therefore luminance contrast may be an important source of visual sensory feedback for motor output.

Properties of visual feedback are used both in the planning and online control of movement. The specific role of luminance contrast for such processes has not been clearly identified, and previous study of this topic is sparse. Recently Braun et al. (2008) investigated whether initiation of eye movements differed when tracking two types of targets, one with luminance contrast compared to the background and one isoluminant with the background (i.e. defined by color only). They showed a strong and significant effect of target contrast on speed of eye movement initiation, with tracking of isoluminant targets delayed by 50 ms. They also showed lower eye accelerations to these no-contrast targets. For upper extremity control, studies have shown mixed results. White, Kerzel, and Gegenfurtner (2006) showed that there was no difference in accuracy or response latency when comparing simple rapid aiming movements to targets of high luminance contrast versus isoluminant targets. In a more complex task, Kleinholdermann et al. (2009) looked at the influence of the target object's luminance contrast as subjects performed reach to grasp movements within a desktop augmented (physical object with graphical overlay) environment. Participants were not provided with a head-coupled stereoscopic view, nor were they provided any visual representation of the hand. They were given a view of the environment that included only a virtual image overlaying the actual target disk. The independent variables controlled by the experimenters were the visual properties of chromatic and luminance contrast between the target object and the environment background. The results of this study showed only a minimal effect of luminance contrast on the formation of grasp aperture. They concluded that isoluminant targets were as suitable for the motor planning of grasp as targets defined by a luminance contrast or a luminance plus chromatic contrast. However, because current theories of motor control rest on the premise that object location can be precisely identified in relation to limb location (Wolpert, Miall, & Kawato, 1998) we contend that the lack of visual feedback about the limb likely resulted in a ceiling effect for a number of performance measures used by Kleinholdermann et al. Given that neuronal tuning properties make the visual system particularly sensitive to change, it is logical that some property involving a change in visual stimulus may be especially useful in this quick, precise identification of object and limb spatial location. Luminance contrast is such a property. Future experiments should expand upon the work of Kleinholdermann et al. by examining the role of luminance contrast of both the target object and the effector limb for upper extremity performance. Further, the Kleinholdermann et al. paper focused predominantly

Vision for Motor Performance in Virtual Environments Across the Lifespan 165

Aim 2 is to test the interaction of age with visual contrast between the limb/target and background environment. We will use the same reach to grasp paradigm, but collect data on a group of healthy adults age 18-25, a middle age group 40-50, and a group of healthy adults age 60+. We believe that older adults will only effectively use visual feedback of self in the highest contrast condition. This will allow inferences about the age-related processing of

We anticipate the results of this line of research will have implications in numerous fields. First, the information gained will have direct bearing on computer science for the userspecific design of next generation 3D virtual environments. As the world population continues to age, understanding of how to enhance performance with computer interfaces must take into account the physiologic changes that occur over time. Luminance contrast appears to be an important factor in upper extremity control, and one that is known to play a role in performance changes with age in natural environments. It stands to reason then that performance in a primarily visual environment, such as a 3D VE, will rely heavily on the neural processing of contrast. Secondly, we believe the field of rehabilitation will benefit indirectly through improvements in user-centered design. Currently, 3D VEs are regularly studied as a means to improve upon current practices in rehabilitation of patients poststroke. Unfortunately, one barrier to success continues to be usability and provision of costeffective, age-appropriate sensory feedback. Information on performance changes in older adults related to manipulation of luminance contrast may be of use to both program designers and rehab clinicians. For example, if older adults perform movements in VEs under certain contrast conditions in a manner equivalent to a natural environment, rehab clinicians may want to capitalize on such parameters to improve functional carryover of training to activities of daily living. Lastly, we believe results from our current and future study will contribute to the fields of gerontology and behavioural neuroscience by expanding our knowledge of visual processing and motor behaviour across the lifespan.

User-centered design of virtual environments continues to be an under-studied area with regard to both old and young users. Knowledge of human performance, and the nature of the sensory feedback that guides it, will be imperative in the successful, cost-effective design of tangible user interfaces intended for use by these populations. Recent work has shown that young adults can utilize visual information provided in virtual environments differently than both older adults and young children, and therefore more specific agegroup studies are needed. Future studies will focus on specific parameters of visual feedback, such as luminance contrast, and how the provision of such properties in virtual

We would like to thank Drew Rutherford, Alexandra Skogen, Brandon Bernardin, and Stephanie Ehle for their assistance with data collection and analysis. This work was funded

luminance contrast as a visual feedback parameter for motor performance.

**4.3.1 Application of results** 

**5. Conclusion** 

**6. Acknowledgements** 

environments impacts the performance of the user.

by the National Science Foundation grant No. 0916119.

on motor planning. Future studies should examine kinematic findings at a deeper level to understand the use of vision in both open and closed loop functions. Additionally, since the treatment of visual contrast information by the CNS changes with time, as discussed in the next section, new studies must focus on delineating the role of such information for populations differing by age.

#### **4.2 Aging and luminance contrast**

While there are changes at the ocular level with age, the predominant cause of functional decline is due to a slowing of central processing in the brain (Chaput & Proteau, 1996; Inui, 1997; Light, 1990; Shields et al., 2005). The slowing of temporal processing has been specifically implicated in the decline of luminance contrast sensitivity in adults over 60. Motion sensitivity, which is dependent on contrast sensitivity, also declines with age (Spering et al., 2005; Trick & Silverman, 1991). Motion sensitivity is also known to be directly linked to function of dopaminergic circuitry, a system known to play a major role in the aging process (Wild-Wall et al., 2008). Despite these declines, older adults become more dependent on vision over time, resulting from the relative sparing of visual resources when compared to other sources of sensory feedback (Adamo et al., 2007; Chaput & Proteau, 1996; Goble et al., 2008; Lemay et al., 2004). The important concept to note is that this sparing of neurons in the visual systems results in a greater amount of substrate available for positive neuroplastic changes relating to motor output. Indeed, such positive changes have been documented in older adults when trained via the visual system to improve speed of processing (Ball, Edwards, & Ross, 2007; Edwards et al., 2005; Long & Rourke, 1989; Zhou et al., 2006). The key question to consider is how might this potential for plastic changes be manipulated and optimized? Given that the processing of luminance contrast information is linked in multiple ways with speed of processing, and speed of processing is a central theme in aging related functional decline, this visual property may be a useful means to answer the plasticity question. We believe a number of attributes of 3D VEs make them an ideal tool to aid in investigating this question, and believe design of VEs will directly benefit from the information gained. Therefore, we intend to investigate changes in sensorimotor processing of luminance contrast in older adults compared to younger adults. The information gained from this study will be directly applicable to development of technologies to rehabilitate and enhance function in aging and neurologically compromised adults.

#### **4.3 Future research aims**

Aim 1 is to test the hypothesis that luminance contrasts of target and limb have an effect on upper extremity kinematics in a virtual environment. This will be investigated using the methodology described previously with a reach to grasp paradigm. We will test a population of adults age 18-25 without history of visual or upper extremity sensorimotor dysfunction. We intend to study five contrast levels ranging from very low to very high. Based on previous studies of visual feedback, we believe that low levels of luminance contrast will negatively affect kinematic markers of upper extremity performance, for example slowed movement time, when compared to moderate and high levels. We also believe that high levels of contrast will not have a significant effect on performance measures when compared to the moderate level for this group of participants.

Aim 2 is to test the interaction of age with visual contrast between the limb/target and background environment. We will use the same reach to grasp paradigm, but collect data on a group of healthy adults age 18-25, a middle age group 40-50, and a group of healthy adults age 60+. We believe that older adults will only effectively use visual feedback of self in the highest contrast condition. This will allow inferences about the age-related processing of luminance contrast as a visual feedback parameter for motor performance.
