Speech Reference Intervals: An Assessment of Feasibility in Depression Symptom Severity Prediction

White, Lauren; Carr, Ewan; Dineley, Judith; Botelho, Catarina; Conde, Pauline; Matcham, Faith; Oetzmann, Carolin; Folarin, Amos; Fairs, George; Norbury, Agnes; Goria, Stefano; Vairavan, Srinivasan; Wykes, Til; Dobson, Richard; Naraya, Vaibhav; Hotopf, Matthew; Abad, Alberto; Trancoso, Isabel; Cummins, Nicholas

doi:10.21437/Interspeech.2025-1438

Speech Reference Intervals: An Assessment of Feasibility in Depression Symptom Severity Prediction

Lauren White, Ewan Carr, Judith Dineley, Catarina Botelho, Pauline Conde, Faith Matcham, Carolin Oetzmann, Amos Folarin, George Fairs, Agnes Norbury, Stefano Goria, Srinivasan Vairavan, Til Wykes, Richard Dobson, Vaibhav Naraya, Matthew Hotopf, Alberto Abad, Isabel Trancoso, Nicholas Cummins

Major Depressive Disorder (MDD) is a prevalent mental disorder. Combining speech features and machine learning has promise for predicting MDD, but interpretability is crucial for clinical applications. Reference intervals (RIs) represent a typical range for a speech feature in a population. RIs could increase interpretability and help clinicians identify deviations from norms. They could also replace conventional speech features in machine learning models. However, no work has yet assessed the feasibility of speech RIs in MDD. We generated and compared RIs from three reference datasets varying in size, elicitation prompt, and health information. We then calculated deviations from each RI set for people with MDD to compare performance on a depression symptom severity prediction task. Our RI-based models trained with demographic data performed similarly to each other and equivalent models using conventional features or demographics only, demonstrating the value of RI-derived features.