Aphasia is a communication disorder affecting approximately 100,000 people in North America, including 6,600 people in Toronto. It is most commonly acquired as the result of a stroke, and affects approximately one third of all stroke survivors [1]. It results in an impairment of language, that is, to the production and/or comprehension of spoken and/or written language. Rehabilitation can reduce the level of impairment, but a significant number of individuals are left with a life-long chronic disability that influences a wide range of activities. Augmentative and alternative communication (AAC) devices leverage the retained abilities of individuals with aphasia to recognize image-based representations of objects. In these systems, words and concepts are represented in a library of images, through which a user can search for words. Once the user has located the relevant image, its letter and sound forms are available for use in communicating with others. However, it is now recognized that electronic speech aids for communication disorders are in a state of crisis. After decades of development, industry leader Dynavox [2] has estimated that just 7 percent of people who would benefit from such an aid actually have one. While this gap is well known, its causes are less well understood.

The first goal of this research is to better understand why so few people with aphasia use such aids. By examining current devices on HCI factors such as interface complexity, organizational model, feature sets, and the need for special training, the effectiveness of current devices will be evaluated. Additionally, this work will strive to understand use of these devices in their common contexts, by evaluating how they are used throughout venues such as the home, rehab centre, or community group. To put these factors into full perspective, current device effectiveness will be weighed against their cost and availability.

The second goal of this research is to examine the potential for everyday devices such as tablets and smart phones to improve AAC design and address current barriers to adoption. Recent advances in computer technology have made smart mobile devices and tablets increasingly ubiquitous and powerful. For example, modern devices are generally equipped with location sensing capabilities such as GPS and Wi-Fi triangulation. Thus, it seems feasible today to build AAC devices that use this information to organize vocabulary based on context, and thereby improve communication. We seek to gather preliminary input into these and other possibilities.