A New Method for Diagnosing Speech and Language Disorders
Approximately 8-9 percent of children in the United States have some sort of speech or language disorder. In other words, they have trouble communicating with their voice. About 5 percent of these cases reveal themselves by first grade but the others go undiagnosed for many more years. Catching these disorders early is the most important step in treating them and researchers have found a promising method for detecting the symptoms early. In a preliminary study, a program developed by researchers at the computer science and artificial intelligence lab at MIT was able to help with identifying, and even diagnosing speech and language disorders in children.
The program
works by analyzing recordings of children speaking. The children were given a
series of pictures with a story and were asked to retell the story in their own
words. Their recordings are analyzed by the program for signals such as pauses
in speech, and then compared to a data set of signals consistently present in
children with these types of disorders. The system can then conclude if the
child has a language disorder or not, which a physician can then use to decide
treatment. The fact that this is a computer program is particularly exciting
because it can potentially be extended for use as an at home treatment. John
Guttag, the author of the paper about this study believes it has the potential
to be enormously beneficial to society. He is quoted saying, “You could imagine the storytelling task being
totally done with a tablet or a phone. I think this opens up the possibility of
low-cost screening for large numbers of children, and I think that if we could
do that, it would be a great boon to society.”
----------------------------------------------------------------------------------------------------------------------
Sources:
http://news.mit.edu/2016/automated-screening-childhood-communication-disorders-0922
https://www.nidcd.nih.gov/health/statistics/statistics-voice-speech-and-language
No comments:
Post a Comment