Tuesday 24 December 2013

Detection of Disease Through Sound Analysis

A new generation of voice analysis software can tell if someone is sick or depressed. Certainly useful as a diagnostic tool. But whether this new form of invasion of privacy?


For decades the voice recognition software was able to detect the words and turn them into text , or respond to commands . But now the developers dive deeper : they examine the acoustic quality of sound to determine medical conditions or even emotions of the speaker .

A software is being developed to detect whether a person suffers from a neurological disease or psychiatric problems . There are even apps that can tell if a person is tired , or depressed .Some experts say the technology is likely to be well received - but not everyone agrees that this is a good thing
.diagnosis objectiveMax Little , a mathematician and researcher for the Massachusetts Institute of Technology ( MIT ) , has developed a sound analysis technology that can detect Parkinson 's disease .Little and his team have put together an audio sample of the people , who suffered from Parkinson's disease and not , when you say " aaah . " If a healthy person voice sounded strong and stable , vibration sound heard on patients with Parkinson's


The vocal cords in a number of conditions


Using machine learning algorithms to detect vibrations in the voice, Little has developed a model that can identify the sound quality of a person suffering from Parkinson's - with about 99 percent accuracy.

Parkinson's Voice Initiative is processing the audio sample rate of 17,000 gathered to try and answer the question whether this technology is suitable for mobile phones

Little mention detection of Parkinson's through the phone " is very convenient in terms of technology , " because three quarters of the world's citizens have access to a mobile phone. Yet another advantage of using this kind of software to diagnose the disease is to eliminate human subjectivity .Tests for Parkinson's and other neurological diseases involving clinical expert opinion . This can lead to different answers because based on human judgment .music brainJörg Langner , a mathematician at the Charité Hospital in Berlin , taking inspiration from musicology . He makes an analytical model that reads six features in sound : loudness , articulation , tempo , rhythm , melody and timbre ." Everything that happens in the brain affect speech production , " said Langner . " Therefore we can track what happens to the brain through the analysis of speech sounds . "Langner recent research related to clinical applications focused on the diagnosis of Attention Deficit Hyperactivity Disorder ( ADHD ) in children .Based on the analysis of voice , speech Langner found differences between children with and without ADHD , including fluctuations in loudness of speech and melody .Tim Langner also claimed to detect levels of depression . " It can be useful to prevent suicide , " said Langner .

Rumble Know? 

In this spektogram, color indicates a low frequency sound and vibration of the sound minor depression

Jarek Krajewski, a psychologist at the University of Wuppertal, has been researching how sound analysis can be used to detect a person's emotional state.

"It can be very basic, such as anger, sadness or joy," said Krajewski. "Or it could be something quite abstract, such as self-confidence."

According to Krajewski there are hundreds of potential applications, including business, science, gaming, marketing, health care and the world of dating.

But Krajewski warns that these technologies also have risks.

"No longer may no longer private emotions in this world," he said. Government, insurance companies, and others can monitor the emotions and personality, Krajewski added, "and draw conclusions that may not be beneficial for people who are monitored."
 
Categories:

0 komentar:

Post a Comment