Portable neurological disease assessment using temporal analysis of speech
Conference
Yadav, N, Poellabauer, C, Daudet, L et al. (2015). Portable neurological disease assessment using temporal analysis of speech
. 77-85. 10.1145/2808719.2808727
Yadav, N, Poellabauer, C, Daudet, L et al. (2015). Portable neurological disease assessment using temporal analysis of speech
. 77-85. 10.1145/2808719.2808727
This paper discusses and evaluates the extraction and analysis of temporal acoustic features from speech using mobile devices to detect patterns which could be indicative of neurodegenarative diseases. Timing information in speech can play a critical role in future portable applications that help detect the presence of speech disorders, which are often indicators of a more serious underlying neurological condition. Speech data collected from a series of mobile administered speech tests at 47 high schools and colleges in the the Midwestern United States are analyzed for the specific case of mild traumatic brain injuries (mTBI). In this paper, we focus on both the system and test design used to collect and extract temporal metrics from the speech data, as well as the statistical analysis of this data to find patterns that are indicative of a concussion. Preliminary results suggest a correlation between certain temporal features and the likelihood of a concussion.
