Researchers at the Massachusetts Institute of Technology Computer Science and Artificial Intelligence Laboratory (CSAIL) and Institute for Medical Engineering and Science have devised a neural network model that can analyze raw text and audio data from clinical interviews to detect speech patterns signaling depression.
The team says this method could be used to develop diagnostic assistants for clinicians that can identify signs of depression in natural conversation.
The technique spots depression-indicative patterns and maps them to individuals, with no additional data required.
CSAIL's Tuka Alhanai explains the process is context-free, "because you're not putting any constraints into the types of questions you're looking for and the type of responses to those questions."
Through a sequence modeling technique, the researchers entered sequences of text and audio data from questions and answers from both depressed and non-depressed people. As the sequences accrued, the model extracted speech patterns that emerged for depressed and non-depressed individuals.
From MIT News
View Full Article
Abstracts Copyright © 2018 Information Inc., Bethesda, Maryland, USA