Clinical Scorecard: Utilizing Large Language Models to Enhance Diagnosis of Language Disorders Linked to Autism and Recognize Unique Characteristics
At a Glance
Category
Detail
Condition
Language disorders associated with Autism Spectrum Disorder (ASD)
Key Mechanisms
Use of large language models (LLMs) to identify autism-related linguistic features such as echolalia, pronoun reversal, and atypical language usage, improving diagnostic sensitivity and precision
Target Population
Individuals with ASD exhibiting language and communication difficulties, including adults
Care Setting
Clinical diagnostic settings involving ASD assessment and intervention planning
Key Highlights
LLMs demonstrated over 10% improvement in sensitivity and positive predictive value for detecting ASD-related language patterns compared to baseline models.
Ten key linguistic features of autism-associated language disorders were identified, including echolalia and pronoun reversal.
LLMs can serve as supplementary diagnostic tools to enhance clinical judgment and enable personalized therapeutic strategies.
Guideline-Based Recommendations
Diagnosis
Use standardized tools like ADOS-2 for ASD diagnosis, supplemented by LLM analysis to identify subtle language disorder features.
Incorporate LLMs to improve detection of specific linguistic markers such as echolalia and atypical pragmatic language use.
Management
Develop tailored treatment plans based on identified language disorder features to address individual patient needs.
Utilize LLM insights to guide personalized therapeutic strategies targeting communication difficulties in ASD.
Monitoring & Follow-up
Regularly assess language and communication patterns to track progress and adjust interventions accordingly.
Leverage automated language analysis tools to monitor changes in linguistic features over time.
Risks
Recognize that LLMs are supplementary tools and not standalone diagnostic instruments; clinical expertise remains essential.
Be aware of potential variability in language disorder presentation across individuals and contexts, requiring cautious interpretation of automated analyses.
Patient & Prescribing Data
Individuals with ASD exhibiting language and communication impairments, including adults
LLM-assisted identification of specific language disorder features supports personalized intervention planning and may improve therapeutic outcomes.
Clinical Best Practices
Combine traditional clinical assessments with LLM-based language analysis to enhance diagnostic accuracy.
Focus on identifying key linguistic markers such as echolalia, pronoun reversal, and pragmatic language impairments for targeted interventions.
Use LLM tools to complement clinical judgment, especially in low-resource settings or when labeled clinical data are limited.
