Abstract
Background: Two previous single-case studies have reported that voice recognition software (VRS) can be a powerful tool for circumventing impaired writing in aphasia. However, these studies report mixed results regarding transfer of skills to functional tasks, such as emailing.
Method: A single-case therapy study was conducted with “Stephen”, a 63-year-old man with fluent aphasia and severe acquired dysgraphia and dyslexia limiting his social participation and ability to return to work. Treatment consisted of 16 one-hour sessions. Stephen was trained to use Dragon NaturallySpeakingRTM VRS to assist writing and Read+WriteGoldRTM text-to-speech software to assist reading, and to develop computer skills required to use email. Outcome measures evaluated writing efficiency and communicative effectiveness, the functional impact of the intervention, and changes in participation.
Results: Training produced significant gains in the efficiency and communicative effectiveness of Stephen’s writing, despite his underlying writing impairment remaining unchanged. Gains generalised to everyday functional communication, leading to increased social participation with Stephen undertaking a wider range of social activities and increasing his social network following treatment. Gains were maintained at follow-up assessment.
Discussion: Results indicate that a relatively short training period with assistive technologies achieved extensive generalisation to independent, functional communicative writing. Indeed, for this case, VRS training may have exceeded the degree of improvement in functional text writing that could have been achieved through impairment therapy, since gains were not limited to treated vocabulary. Some challenges were encountered in training Stephen to use VRS but, through adaptations to the training process, were largely overcome. Importantly, regaining independent writing skills resulted in profound and life-changing improvements to social participation. This may have resulted in Stephen reconnecting with important aspects of his pre-stroke identity, and improving his self-esteem.
Conclusion: This case adds to a small evidence base indicating that training in the use of VRS, in combination with text-to-speech software, may be an effective way to address writing impairments in chronic aphasia for individuals with relatively well-preserved spoken output. Not only can these technologies improve the efficiency and communicative effectiveness of writing, they can also lead to significant gains in functional communication and social participation. Further research is needed trialling this approach with a larger group of people with aphasia.
Acknowledgements
We thank Stephen for participating in the study, Jane Marshall for commenting on an earlier draft, Carolyn Bruce for sharing experiences of VRS training via personal communication, City University London for support during data collection, and Barts Charity for their support during the writing of this paper.
Notes
1. It should be noted that VRS accuracy is imperfect even for unimpaired speakers, with reported word recognition error rates of 21–34% (Hakkani-Tur, Vergyri, & Tur, Citation2010), although accuracy is likely to be higher with more recent software versions.
2. In recent versions of Dragon (e.g., Dragon 13 Home) the training process has been improved and is more accessible for people with aphasia—the user needs to talk into the microphone using connected speech but no longer needs to read aloud a set passage. The software is also available as an app for iPad and iPhone, which does not require any speech training.