DESIGN AND QUALITATIVE EVALUATION OF TACTILE DEVICES FOR STROKE REHABILITATION G.V. Merrett* 1 , C.D. Metcalf* † , D. Zheng*, S. Cunningham † , S. Barrow*, S.H. Demain † * Electronics and Computer Science, University of Southampton, Hampshire, UK, SO17 1BJ † Faculty of Health Sciences, University of Southampton, Hampshire, UK, SO17 1BJ 1 E-mail: gvm@ecs.soton.ac.uk Keywords: tactile, haptic, stroke, shape memory alloy Abstract Rehabilitation environments combining virtual reality with everyday motor tasks can promote recovery from neurological illness, such as stroke. Tactile devices, providing physical stimulation to the skin, may improve motor retraining. While many tactile devices have been reported, there is a distinct paucity of studies evaluating how they are perceived. This multidisciplinary research has investigated three tactile devices (vibration motors, a motor-driven ‘squeezer’, and shape memory alloys) for providing a realistic sensation of static interaction with virtual objects. These devices have been iteratively redesigned and qualitatively evaluated with healthy human participants. This paper presents the devices, their evaluation, and iterative redesign. 1 Introduction Rehabilitation environments that combine virtual reality with everyday motor tasks can promote recovery from neurological illness, such as stroke [1]. Intact sensation is essential for a multitude of activities including fine motor control [2], injury prevention [3] and a normal quality of life. Haptic devices are used to provide artificial stimulation to the skin to generate a ‘sense of touch’, and may offer more realistic and immersive virtual environments, subsequently leading to improved motor retraining. Therefore, haptic devices are increasingly incorporated into therapy to retrain sensory abilities [4] to convey properties of virtual objects (enhancing the realism of the sensorial experience) [5]. Haptic sensations can be broadly classified as coming from two distinct inputs, namely kinaesthetic and cutaneous stimuli [6]. Kinaesthetic inputs are generated by stimulating receptors in the muscles, joints and tendons, and represent movement, position and posture. Cutaneous (or, as referred to in this paper, tactile) inputs are generated by stimulating mechanoreceptors in the skin, and detect skin contact with objects and perception of surface properties. Clearly, the perception of touching an object relies on both of these inputs; consider picking up a box, the skin on the finger and hand is stimulated by the deformation of the fingertip (tactile), and the finger joints, muscles and tendons detect the normal force exerted through gripping (kinaesthetic). However, in this research we are concerned only with generating realistic tactile sensations, as this does not require intrusive hardware such as mechanical exoskeletons [7]. Whilst the ultimate aim of this research is an integrated sensory and motor rehabilitation tool, we first needed to develop and evaluate the tactile devices to ensure that realistic sensations are generated; this is the focus of this paper. Many tactile devices have been reported in the literature but, despite considerable advancements, there is a distinct paucity of studies evaluating the perceptual experience of their use. This research was undertaken by a multidisciplinary team, with a collaborative, user-led approach to development. The team consisted of engineers, psychologists and clinicians who equally contributed opinions to designing the tactile devices. Three tactile devices were developed (e.g. Fig. 1), capable of giving rise to realistic sensations of static interaction with virtual objects. The devices were qualitatively evaluated with human participants and iteratively redeveloped based on the feedback. Seven non-impaired participants took part (4 male, 3 female; age 23-59yrs, mean=36.4, SD=13.5). Non-impaired participants ensured we did not give people with impaired sensation noxious stimuli, and established a normative baseline before evaluating on people with stroke. Ethical approval was granted by the Faculty of Health Science’s Ethics Committee, University of Southampton (ref 2010-025). Figure 1: ‘Vibration’ tactile device (inset-top: devices mounted on the thumb and index finger; inset-bottom: CAD drawing of the casing). This paper is structured as follows: Section 2 gives an overview of the existing research on tactile devices, followed by details of the vibration (Section 3), motor-driven ‘squeezer’ (Section 4), and SMA (Section 5) devices that have been developed in this research. For clarity, the results