Imran Musaji, assistant professor, Ph.D., CCC-SLP in the Department Communication Sciences and Disorders (CSD), presented “An Explainability Framework for CSD AI Applications” at The American Speech-Language-Hearing Association (ASHA) 2025 Convention (Centennial year) in Washington, DC Nov. 20, 2025.
Description: This session introduces CLEAR-AI (Clinician-led Evaluation and Assessment of Readiness of AI), a structured framework for evaluating AI applications in communication sciences and disorders (CSD). Key AI concepts—including interpretability, transparency, and bias—will be explained in clinician-friendly terms. Attendees will explore AI evaluation principles from healthcare frameworks, emphasizing fairness, accountability, and ethical AI use. CLEAR-AI provides a systematic approach to assess AI-driven tools based on data sources, training methods, clinical alignment, and oversight. Practical examples will illustrate how to apply the framework to AI tools in speech-language pathology and audiology. This session equips clinicians with strategies to critically appraise and safely implement AI in practice, continuing the discussion initiated in the 2024 ASHA Short Course.