You don’t always have a respirator with you, but you usually do. Is that possibly even enough?
According to Recently published research From Stanford University’s Institute of Emergency Medicine, smartphones may soon function as personal inhalers. The aim is to use software to detect how high a person’s blood alcohol level is based on their voice – without the need for a traditional breath alcohol tester.
As the researchers were able to show in their research, the software analyzes the subtlest changes in the voice, such as pitch, frequency and speaking speed. Based on this information, the degree of alcoholism can be determined with considerable accuracy.
As part of the study, 18 people aged 21 to 62 were given vodka and alcohol according to their body weight. Before and after drinking alcohol, subjects had to speak into a smartphone microphone. The software was able to accurately determine a blood alcohol level of 0.8 for almost all people.
A total of 18 participants (72% male, age 21–62 years) read a randomly assigned tongue twister every hour before drinking and up to 7 hours after drinking a weight-dependent amount of alcohol. Vowel segments were cleaned and segmented into 1-s windows. We developed support vector machine models for detecting alcohol intoxication, defined as breath alcohol concentration > 0.08%, comparing the baseline vocal spectrographic signature with each subsequent time point, and examined accuracy with 95% confidence intervals (CIs).
from the study
According to the authors, the accuracy of this new alcohol test is 98 percent, which is in line with previous studies of similar methods. In the future, the software will be further trained so that even small language samples are sufficient for reliable measurements. In addition, text input or a person’s movement can be evaluated for increased accuracy.
However, there are still limitations. So far, smartphone alcohol measurement works only in English and with pure alcohol consumption. Mixed consumption remains a challenge. It must be noted that the group of people tested was still too small to make a reliable statement about the reliability of this method.
follow us on Google News And talk to us SmartDroid Chat on Telegram. * Links marked are regulated affiliate links.
