58 How trustworthy are clinically trained Small Language Models for extracting genomic information from reports?

Abstract

constraints of assessment can often be problematic.We therefore wanted to design visual resources to explain autism and ADHD, but which could be personalised and tailored to a young person.Method After identifying the need for young-person focused resources, informal feedback was gathered from families attending our multi-disciplinary clinic, alongside scoping of existing resources via literature searching.Following discussion and idea generation from a team experienced in assessing young people with neurodisability, a list was drawn up so that resources could be developed.Results Two sets of card decks were developed which includes 65 descriptors of the different characteristics associated with autism and ADHD.These were based on both diagnostic criteria and well-known strengths and struggles of young people with these diagnoses.Language was neuroaffirmative, simplified to a few words and supplemented with a picture.Each descriptor was made into an individual image, so that all the relevant cards could be selected on a case-by-case basis, for the young person to then have a personalised, descriptive and visual representation of their condition to take-away.Conclusion Following the creation of the resource, initial feedback was collected from professionals regarding the suitability and usability, with positive feedback.The next stage is to collect feedback from the young people, take to the GOSH Young Person Forum and also review the longer term impact on a child's understanding of their diagnosis following use of the cards.

Ähnliche Arbeiten