Purpose: This review aims to delve into the question of whether artificial intelligence (AI) could be used to create images that are linguistically and culturally relevant for individuals with acquired language disorders with specific consideration of Turkish. Images are one of highly preferred stimuli among acquired language disorders both in research and clinical contexts.
Method: A narrative review methodology was adopted in the study by synthesizing the findings from previous studies. A literature review was carried out using various databases including Web of Science, Scopus, and EBSCO with search terms including “acquired language disorders”, “aphasia”, “image research”, “artificial intelligence”, “text-to-image generation”, and “prompt engineering”.
Results: Previous studies on image research including neurotypical adults and individuals with aphasia were introduced to draw a comprehensive picture of what ideal image characteristics would involve. The potential uses of AI in acquired language disorders were discussed with particular focus on image generation tools by unraveling the options of image generation, practice of writing prompts, potential benefits and challenges of text-to-image generation tools, and the need to consider cultural and linguistic diversity in the formation of images.
Conclusion: There is a lack of research that embodies AI and image research together, which means that there needs to be a close collaboration between the researchers in the field of AI and those in the aphasia rehabilitation. Regarding languages other than English, AI-generated image tools need to consider the unique linguistic and cultural characteristics of the languages concerned. It is suggested that the use of these tools should be promoted among speech and language therapists in their clinical practice, and their experiences with these tools should be documented so that image generation could be a part of clinical practice.
Keywords
acquired language disorders, aphasia rehabilitation, artificial intelligence, human-computer interaction, image research, text-to-image generation, Turkish
References