**1. Introduction**

16 Will-be-set-by-IN-TECH

150 Virtual Reality and Environments

Tost D., Ferré M., Garcia P., Tormos J., Garcia A., Roig T. & Grau, S. (2009). Previrnec:

*Rehabilitation*, pp. 1–8.

a cognitive telerehabilitation system based on virtual environments, *Virtual*

Current trends in neuroscience research are heavily focused on new technologies to study and interact with the human brain. Specifically, three-dimensional (3D) virtual environment (VE) systems have been identified as technology with good potential to serve in both research and applied settings. For the purpose of this chapter, a virtual environment is defined as a computer with displays and controls configured to immerse the operator in a predominantly graphical environment containing 3D objects in 3D space. The operator can manipulate virtually displayed objects in real time using a variety of motor output channels or input devices. The use of VEs has almost exclusively been limited to experimental processes, utilizing cumbersome equipment well suited for the laboratory, but unrealistic for use in everyday applications. As the evolution of computer technology continues, the possibility of creating an affordable system capable of producing a high-quality 3D virtual experience for home or office applications comes nearer to fruition. However, in order to improve the success and the cost-to-benefit ratio of such a system, more precise information regarding the use of VEs by a broad population of users is needed. The goal of this chapter is to review knowledge relating to the use of visual feedback for human performance in virtual environments, and how this changes across the lifespan. Further, we will discuss future experiments we believe will contribute to this area of research by examining the role of luminance contrast for upper extremity performance in a virtual environment.

## **2. Background**

The following sections identify the well-known physiologic changes that occur in the sensorimotor system as part of the natural human aging process. Further, we discuss some of the limited work that has been done to understand the implications of these changes for the design of VEs.

#### **2.1 Changes to the human sensorimotor system across the lifespan**

The human body is a constantly changing entity throughout the lifespan. Most physiologic processes begin to decline at a rate of 1% per year beginning around age 30, and the sensorimotor system is no exception (Schut, 1998). There is a general indication from the research that both the processing of afferent information and the production of efferent

Vision for Motor Performance in Virtual Environments Across the Lifespan 153

accuracy of movement (Chaput & Proteau, 1996), reaction time (Light, 1990; Sparrow et al., 2006; Yan et al., 1998), strength (Roos et al., 1997; Vandervoort, 2002), hand dexterity (Contreras-Vidal, Teulings, & Stelmach, 1998; Seidler, Alberts, & Stelmach, 2002), and postural control (Jonsson, Henriksson, & Hirschfeld, 2007; Koceja, Allway, & Earles, 1999; Mourey et al., 1998; Romero & Stelmach, 2003). En masse, these changes have the potential to contribute to a spiral of disuse and loss of function that often characterizes the process of aging. Due to the tendency for visual dominance in aged humans (Lemay et al., 2004), and the task specificity of human movement (Proteau, 1992), the fact that visual sensory feedback is much less rich in a virtual than natural environment makes it imperative to study human performance in such surroundings. Research is needed to improve our understanding of sensorimotor changes, and their consequences for performance, for an

**2.2 Three-dimensional virtual environments and human computer interaction (HCI)** 

Today, the users of computers include people from all age groups. Very little information is available on how the performance of individuals in a VE changes throughout the lifespan as a function of the natural aging process. Prior to designing programs for individuals in special subgroups, such as rehabilitation programs designed for patients with neurological lesions, it will be important to understand what age-specific requirements will be beneficial to the user. For instance, because there is a paucity of information on how healthy adults in the older age groups commonly affected by stroke interact in VEs, it is likely that a system designed as an adjunct to standard rehabilitation will struggle to gain success without a foundation of baseline knowledge. This level of information regarding subjects of various age groups will greatly assist in producing successful, cost-effective VEs. Unfortunately, although computers have been commonplace in homes and workenvironments for decades, the literature on interface design as it relates to age is only very recent, and is limited in scope. Early computer interface design relied primarily on the intuition of the designer (Hawthorn, 2001; Hawthorn, 2007). There was a distinct disparity between what designers recognized as necessary interface components and what was truly usable by the lay population. As access to computer technology improved and allowed the spread of computers into the hands of consumers, a necessary change to usercentered design followed. Typically, however, in order to be a feasible process, the representative users must have a basic level of proficiency with computer skills and language. This resulted in a general exclusion of both young and old age groups from the design process. In the late 1990's, interest in age-specific design increased, and there is now a reasonable body of knowledge on the design of standard computer interface

While the bulk of age-specific computer design information relates to ways to improve cognitive performance through specific training or tutorial methods (Hawthorn, 2007), there is some scientific literature which explores the areas of human motor control (Laursen, Jensen, & Ratkevicius, 2001; Smith, Sharit, & Czaja, 1999). Most of this information centers on the input device, specifically mouse usage in older adults. Smith et al. (1999) reported that there are many age-related changes in performance, and in general, it is quite difficult for older individuals to use a mouse. The act of double-clicking seems to consistently be the

aging population interacting in three-dimensional environments.

**across the lifespan** 

systems for various age groups.

signals steadily change as a function of age. Multiple authors demonstrate physical changes in brain tissues (Andersen, Gundersen, & Pakkenberg, 2003; Kuo & Lipsitz, 2004; Raz & Rodrigue, 2006), changes in excitability of the corticospinal tract and anterior horn cells (Rossini, Desiato, & Caramia, 1992), and changes in neurotransmitter systems. There is a general loss of neural substrate, including grey and white matter. This has been demonstrated in both the cerebral cortex (Raz & Rodrigue, 2006) and the cerebellum (Andersen et al., 2003). These tissue changes then result in a myriad of functional changes within the central nervous system (CNS). There is a general deterioration of motor planning capabilities (Sterr & Dean, 2008;Yan, Thomas, & Stelmach, 1998) and feed-forward anticipatory control (Hwang et al., 2008) with aging. Along with this decrease in planning ability, there also appears to be slowing of central processing (Chaput & Proteau, 1996; Inui, 1997; Light, 1990; Shields et al., 2005). This change in processing is partially due to neurophysiologic changes within the CNS resulting in a decrease in the available resources of the processing pool (Craik & McDowd, 1987; Schut, 1998). Loss of attentional resources also contributes to this slowing of central processing (Goble et al., 2008; Kluger et al., 1997; Sparrow, Begg, & Parker, 2006). This in itself results from a multifactorial process relating to neurophysiologic changes in the CNS and degradation of afferent information arriving from compromised peripheral receptors (Chaput & Proteau, 1996; Goble et al., 2008). The result of these attentional and processing changes is a decline in the ability to integrate multiple sensory modalities causing a relative decrease in the use of proprioceptive feedback and an increased use of vision (Adamo, Martin, & Brown, 2007; Chaput & Proteau, 1996; Goble et al., 2008; Lemay, Bertram, & Stelmach, 2004). This shift to the use of visual resources is due to the tendency of the CNS to re-weight sensory information when one source of feedback is compromised (Horak & Hlavacka, 2001), as well as a general systems neuroplasticity effect (Heuninckx, Wenderoth, & Swinnen, 2008; Romero et al., 2003). These compensatory neuroplastic changes are the end manifestation of the normal aging process within the CNS.

The peripheral nervous system (PNS) undergoes concordant neurophysiologic changes as well (Chaput & Proteau, 1996; Goble et al., 2008; Roos, Rice, & Vandervoort, 1997). These changes occur in both the afferent and efferent pathways. Studies have shown both a decrease in number and density of proprioceptors (Goble et al., 2008), as well as a slowing of sensory receptors in general (Light, 1990). In the efferent systems, research demonstrates a loss of motor units and a decrease in firing rate and increased discharge variability of intact motor units (Roos et al., 1997). The available literature also demonstrates a loss of larger motor neurons resulting in a net decrease of alpha motor neurons, a slowing in the conduction velocity of remaining motor neurons, and changes in the excitability of alpha motor neurons (Leonard et al., 1997; Roos et al., 1997).

The changes in the CNS and PNS with age are accompanied by changes in the muscular system as well. In the aging adult, research shows a loss of muscle fibers and a decrease in size of remaining fibers resulting in a net loss of muscle mass (Roos et al., 1997). Changes in motor units in the PNS result in fiber type changes, causing a loss of fast-twitch fibers and a proportional increase of slow-twitch fibers.

Transformations in the sensorimotor system have a resultant detrimental effect on motor performance in daily life. This decrease has a physiologic basis in aging and is amplified by disuse and dysfunction. In general, aging adults demonstrate decreases in movement speed (Light, 1990; Mankovsky, Mints, & Lisenyuk, 1982; Poston et al., 2008; Yan et al., 1998),

signals steadily change as a function of age. Multiple authors demonstrate physical changes in brain tissues (Andersen, Gundersen, & Pakkenberg, 2003; Kuo & Lipsitz, 2004; Raz & Rodrigue, 2006), changes in excitability of the corticospinal tract and anterior horn cells (Rossini, Desiato, & Caramia, 1992), and changes in neurotransmitter systems. There is a general loss of neural substrate, including grey and white matter. This has been demonstrated in both the cerebral cortex (Raz & Rodrigue, 2006) and the cerebellum (Andersen et al., 2003). These tissue changes then result in a myriad of functional changes within the central nervous system (CNS). There is a general deterioration of motor planning capabilities (Sterr & Dean, 2008;Yan, Thomas, & Stelmach, 1998) and feed-forward anticipatory control (Hwang et al., 2008) with aging. Along with this decrease in planning ability, there also appears to be slowing of central processing (Chaput & Proteau, 1996; Inui, 1997; Light, 1990; Shields et al., 2005). This change in processing is partially due to neurophysiologic changes within the CNS resulting in a decrease in the available resources of the processing pool (Craik & McDowd, 1987; Schut, 1998). Loss of attentional resources also contributes to this slowing of central processing (Goble et al., 2008; Kluger et al., 1997; Sparrow, Begg, & Parker, 2006). This in itself results from a multifactorial process relating to neurophysiologic changes in the CNS and degradation of afferent information arriving from compromised peripheral receptors (Chaput & Proteau, 1996; Goble et al., 2008). The result of these attentional and processing changes is a decline in the ability to integrate multiple sensory modalities causing a relative decrease in the use of proprioceptive feedback and an increased use of vision (Adamo, Martin, & Brown, 2007; Chaput & Proteau, 1996; Goble et al., 2008; Lemay, Bertram, & Stelmach, 2004). This shift to the use of visual resources is due to the tendency of the CNS to re-weight sensory information when one source of feedback is compromised (Horak & Hlavacka, 2001), as well as a general systems neuroplasticity effect (Heuninckx, Wenderoth, & Swinnen, 2008; Romero et al., 2003). These compensatory neuroplastic changes are the end manifestation of the normal aging process within the CNS. The peripheral nervous system (PNS) undergoes concordant neurophysiologic changes as well (Chaput & Proteau, 1996; Goble et al., 2008; Roos, Rice, & Vandervoort, 1997). These changes occur in both the afferent and efferent pathways. Studies have shown both a decrease in number and density of proprioceptors (Goble et al., 2008), as well as a slowing of sensory receptors in general (Light, 1990). In the efferent systems, research demonstrates a loss of motor units and a decrease in firing rate and increased discharge variability of intact motor units (Roos et al., 1997). The available literature also demonstrates a loss of larger motor neurons resulting in a net decrease of alpha motor neurons, a slowing in the conduction velocity of remaining motor neurons, and changes in the excitability of alpha

motor neurons (Leonard et al., 1997; Roos et al., 1997).

proportional increase of slow-twitch fibers.

The changes in the CNS and PNS with age are accompanied by changes in the muscular system as well. In the aging adult, research shows a loss of muscle fibers and a decrease in size of remaining fibers resulting in a net loss of muscle mass (Roos et al., 1997). Changes in motor units in the PNS result in fiber type changes, causing a loss of fast-twitch fibers and a

Transformations in the sensorimotor system have a resultant detrimental effect on motor performance in daily life. This decrease has a physiologic basis in aging and is amplified by disuse and dysfunction. In general, aging adults demonstrate decreases in movement speed (Light, 1990; Mankovsky, Mints, & Lisenyuk, 1982; Poston et al., 2008; Yan et al., 1998), accuracy of movement (Chaput & Proteau, 1996), reaction time (Light, 1990; Sparrow et al., 2006; Yan et al., 1998), strength (Roos et al., 1997; Vandervoort, 2002), hand dexterity (Contreras-Vidal, Teulings, & Stelmach, 1998; Seidler, Alberts, & Stelmach, 2002), and postural control (Jonsson, Henriksson, & Hirschfeld, 2007; Koceja, Allway, & Earles, 1999; Mourey et al., 1998; Romero & Stelmach, 2003). En masse, these changes have the potential to contribute to a spiral of disuse and loss of function that often characterizes the process of aging. Due to the tendency for visual dominance in aged humans (Lemay et al., 2004), and the task specificity of human movement (Proteau, 1992), the fact that visual sensory feedback is much less rich in a virtual than natural environment makes it imperative to study human performance in such surroundings. Research is needed to improve our understanding of sensorimotor changes, and their consequences for performance, for an aging population interacting in three-dimensional environments.

#### **2.2 Three-dimensional virtual environments and human computer interaction (HCI) across the lifespan**

Today, the users of computers include people from all age groups. Very little information is available on how the performance of individuals in a VE changes throughout the lifespan as a function of the natural aging process. Prior to designing programs for individuals in special subgroups, such as rehabilitation programs designed for patients with neurological lesions, it will be important to understand what age-specific requirements will be beneficial to the user. For instance, because there is a paucity of information on how healthy adults in the older age groups commonly affected by stroke interact in VEs, it is likely that a system designed as an adjunct to standard rehabilitation will struggle to gain success without a foundation of baseline knowledge. This level of information regarding subjects of various age groups will greatly assist in producing successful, cost-effective VEs. Unfortunately, although computers have been commonplace in homes and workenvironments for decades, the literature on interface design as it relates to age is only very recent, and is limited in scope. Early computer interface design relied primarily on the intuition of the designer (Hawthorn, 2001; Hawthorn, 2007). There was a distinct disparity between what designers recognized as necessary interface components and what was truly usable by the lay population. As access to computer technology improved and allowed the spread of computers into the hands of consumers, a necessary change to usercentered design followed. Typically, however, in order to be a feasible process, the representative users must have a basic level of proficiency with computer skills and language. This resulted in a general exclusion of both young and old age groups from the design process. In the late 1990's, interest in age-specific design increased, and there is now a reasonable body of knowledge on the design of standard computer interface systems for various age groups.

While the bulk of age-specific computer design information relates to ways to improve cognitive performance through specific training or tutorial methods (Hawthorn, 2007), there is some scientific literature which explores the areas of human motor control (Laursen, Jensen, & Ratkevicius, 2001; Smith, Sharit, & Czaja, 1999). Most of this information centers on the input device, specifically mouse usage in older adults. Smith et al. (1999) reported that there are many age-related changes in performance, and in general, it is quite difficult for older individuals to use a mouse. The act of double-clicking seems to consistently be the

Vision for Motor Performance in Virtual Environments Across the Lifespan 155

Fig. 1. Wisconsin virtual environment (WiscVE). Panel A shows the apparatus with

downward facing monitor projecting to the mirror. Images are then reflected up to the user wearing stereoscopic LCD shutter goggles, and thus the images appear at the level of the actual work surface below. Panels B and C demonstrate a reach to grasp task commonly utilized in this environment. The hand and physical cube are instrumented with light emitting diodes (LEDs) that are tracked by the VisualEyez (PTI Phoenix, Inc) system, not shown.

This type of interface gives investigators complete control over the three-dimensional visual scene (important in generalizeability to natural environments), and makes for maximal use of the naturalness, dexterity and adaptability of the human hand for the control of computer mediated tasks (Sturman & Zeltzer, 1993). The use of such a tangible user interface removes many of the implicit difficulties encountered with standard computer input devices due to natural aging processes (Smith et al., 1999). The exploitation of these abilities in computergenerated environments is believed to lead to better overall performance and increased richness of interaction for a variety of applications (Hendrix & Barfield, 1996; Ishii & Ullmer, 1997; Slater, Usoh, & Steed, 1995). Furthermore, this type of direct-manipulation environment capitalizes on the user's pre-existing abilities and expectations, as the human hand provides the most familiar means of interacting with one's environment (Schmidt & Lee, 1999; Schneiderman, 1983). Such an environment is suitable for applications in simulation, gaming/entertainment, training, visualization of complex data structures, rehabilitation and learning (measurement and presentation of data regarding movement

disorders). This allows for ease of translation of our data to marketable applications.

The VE provides a head-coupled, stereoscopic experience to a single user, allowing the user to grasp and manipulate augmented objects. The system is configured as follows (Figure 1): 3-D motion information (e.g. movement of the subject's hand, head and physical objects within the environment) is monitored by a VisualEyez (PTI Phoenix, Inc.) motion

most problematic. Difficulty with cursor control is also named as a top complaint among older individuals (Hawthorn, 2001; Hawthorn, 2007). It has also been shown that performance within a standard computer interface is slower and results in a greater number of errors with increased age of the operator. These specific limitations point to the need to develop new interfaces that capitalize on natural manipulation, thereby eliminating difficulties with the functional abstraction of input devices.

Contrary to standard computer interface systems, little is known about the age-related variance of HCI within three-dimensional virtual environments. The literature in this subject area is nearly non-existent. There is some evidence of age-related differences in performance between children and adults, as well as young adults and older participants. This research indicates relevant disparities in reactions to environmental immersion, usage of various input devices, size estimation ability, navigational skills and completion time for gross motor tasks (Allen et al., 2000). According to these authors, "these results highlight the importance of considering age differences when designing for the population at large." Currently, the International Encyclopedia of Ergonomics and Human Factors (Karwowski, 2006) leaves the explanation of age-related differences in virtual environments to a short, two sentence description recommending that equipment be tailored to physically fit the smaller frames of children, and for designers to take into consideration the changes in sensory and motor functions of the elderly. Other than these works, very little specific knowledge regarding age and motor control in virtual environments has been elicited through research, especially as it relates to precision movements with the upper extremities. This fact has led us to begin a series of experiments investigating the use of vision for precise sensorimotor control of the upper extremity in virtual environments, and how that usage changes as a function of age.
