Detecting stress and depression in adults with aphasia through speech analysis
Stephanie Gillespie, Elliot Moore II, Jacqueline Laures-Gore, Matthew Farina, Scott Russell, Yash-Yee Logan
Abstract
Aphasia is an acquired communication disorder often resulting from stroke that can impact quality of life and may lead to high levels of stress and depression. Depression diagnosis in this population is often completed through subjective caregiver questionnaires. Stress diagnostic tests have not been modified for language difficulties. This work proposes to use speech analysis as an objective measure of stress and depression in patients with aphasia. Preliminary analysis used linear support vector regression models to predict depression scores and stress scores for a total of 19 and 18 participants respectively. Teager Energy Operator-Amplitude Modulation features performed the best in predicting the Perceived Stress Scale score based on various measures. The complications of speech in people with aphasia are examined and indicate the need for future work on this understudied population.
BibTeX
@inproceedings{icassp2017_detectingstressa,
title = {Detecting stress and depression in adults with aphasia through speech analysis},
author = {Stephanie Gillespie and Elliot Moore II and Jacqueline Laures-Gore and Matthew Farina and Scott Russell and Yash-Yee Logan},
booktitle = {ICASSP 2017},
year = {2017}
}