Abstract
Diagnosing language disorders associated with autism is a complex and nuancedchallenge, often hindered by the subjective nature and variability oftraditional assessment methods. Traditional diagnostic methods not only requireintensive human effort but also often result in delayed interventions due totheir lack of speed and specificity. In this study, we explored the applicationof ChatGPT, a state of the art large language model, to overcome theseobstacles by enhancing diagnostic accuracy and profiling specific linguisticfeatures indicative of autism. Leveraging ChatGPT advanced natural languageprocessing capabilities, this research aims to streamline and refine thediagnostic process. Specifically, we compared ChatGPT's performance with thatof conventional supervised learning models, including BERT, a model acclaimedfor its effectiveness in various natural language processing tasks. We showedthat ChatGPT substantially outperformed these models, achieving over 13%improvement in both accuracy and F1 score in a zero shot learningconfiguration. This marked enhancement highlights the model potential as asuperior tool for neurological diagnostics. Additionally, we identified tendistinct features of autism associated language disorders that varysignificantly across different experimental scenarios. These features, whichincluded echolalia, pronoun reversal, and atypical language usage, were crucialfor accurately diagnosing ASD and customizing treatment plans. Together, ourfindings advocate for adopting sophisticated AI tools like ChatGPT in clinicalsettings to assess and diagnose developmental disorders. Our approach not onlypromises greater diagnostic precision but also aligns with the goals ofpersonalized medicine, potentially transforming the evaluation landscape forautism and similar neurological conditions.