Automatic Temporal Analysis of Speech: A Quick and Objective Pipeline for the Assessment of Overt Stuttering

Purpose: Developmental stuttering is a communication disorder characterized by the production of stuttering disfluency, such as blocks, repetitions, prolongations. Accurate measurement of overt stuttering behavior can aid in diagnostic evaluation, determination of the optimal course of treatment, and for tracking progress in that treatment. In this study, we propose a novel method for the assessment of speech fluency, based on the automatic detection and quantification of discrete pause and vocal events in continuous speech. Our first hypothesis is that adults who stutter (AWS) will exhibit more pausing or hesitancy in speech compared to adults who does not stutter (AWNS). Our second hypothesis is that the speech of AWS will be more temporally irregular than AWNS, as evidenced by greater variability in the duration of pause and vocal events. Our third hypothesis is that our ATAS metrics will be accurate in predicting status of each participant as either an AWS or AWNS grouping. Methods: We have constructed a novel methodology, Automatic Temporal Analysis of Speech (ATAS), to quantify the temporal regularities of pause and vocal events within the speech stream. Our seven ATAS metrics relevant to this preliminary study include: speech rate, total pause time, pause count, mean pause duration, mean vocal duration, pause duration variability, and vocal duration variability. Oral reading audio samples from a total of 35 English-speaking participants were used: 17 AWS and 18 AWNS.Results: AWS, in general, exhibited more pausing or hesitancy in speech compared to AWNS, as evidenced by slower speech rate, greater total pause time, and longer mean duration of pause events. Numerous pause and vocal metrics acquired from ATAS were correlated with stuttering frequency, which is suggestive that automatically detected temporal metrics of pause and vocal events within continuous speech are highly associated with stuttering.Conclusion: The automatic and instrumental detection and quantification of pause and vocal events during speech may provide an alternative and complementary method that SLPs and other health professionals can use in the assessment of fluency disorders, such as stuttering.