Utility of lay and clinical narratives for transparent autism diagnosis using BioBERT deep learning - Takeaways - MDSpire

Leveraging Lay and Clinical Narratives for Enhanced Transparency in Autism Diagnosis with BioBERT Deep Learning Techniques

  • By

  • Gondy Leroy

  • Himanshu Nimbarte

  • Madhuri Sai Kandula

  • Prosanta Barai

  • Sumi Lee

  • Nell Maltman

  • Sydney Rice

  • July 17, 2026

Share

  • 1

    Early autism diagnosis is hindered by reliance on clinical observation and limited specialist availability.

  • 2

    BioBERT was trained to label autism behavioral descriptions using DSM-5 criteria, enhancing transparency in clinical decision-making.

  • 3

    The model achieved higher precision on clinical descriptions (69%) and higher recall on lay descriptions (83%).

  • 4

    Lay behavioral descriptions provided diagnostically valuable information comparable to clinical observations.

  • 5

    Integrating lay information into diagnostic workflows could accelerate autism diagnosis without compromising clinical utility.

Original Source(s)

Utility of lay and clinical narratives for transparent autism diagnosis using BioBERT deep learning

  • by Gondy Leroy, Himanshu Nimbarte, Madhuri Sai Kandula, Prosanta Barai, Sumi Lee, Nell Maltman, Sydney Rice

  • July 17, 2026

Related Content

Hierarchical neighbor integration graph attention network for autism spectrum disorder diagnosis

Exercise dose architecture as a precision-informed developmental framework in autism spectrum disorder: a narrative review of mechanisms, sex- and development-related heterogeneity, and clinical translation

Trends in antipsychotic drug utilisation and indication in the paediatric population: a nationwide cohort study in Sweden