Share:
Can depression be recognized on Instagram pictures?
Depression is a serious disorder of the psyche. For sufferers, such a disease can have terrible consequences. In the worst case, depression leads to suicide. But how can we tell if we have depression? Researchers have now developed a new algorithm that can examine photos on Instagram for signs of depression.
Can photos show a computer program if we are depressed? Scientists from the world-renowned Harvard University, in collaboration with the University of Vermont, have developed an algorithm that can detect depression by evaluating photos on Instagram. The experts published the results of their study in the Cornell University Library.
Many people in the world suffer from depression. Physicians and scientists have been looking for new ways for faster diagnosis for years. Researchers have now developed an algorithm that can detect depression by analyzing photos. (Image: peshkova / fotolia.com) Analysis of images can detect depression
What do the photos published by you on Instagram say about their mental state? Apparently a lot. It is even possible to detect an onset of depression by analyzing the images, the researchers say in their study. When people mostly post dark and gray images, they are more likely to suffer from depression.
New algorithm checks images for signs of mental illness
Researchers at Harvard University and the University of Vermont have developed a new algorithm. This can check photos on Instagram for signs of depression and other mental illness, explain the doctors. Thus, these dangerous diseases can be detected at an early stage. According to the researchers, there is a connection between the use of color and the mental state of a person.
Study examines 170 subjects with Instagram account
For their study, the researchers examined over 170 employees from the so-called Amazon Mechanical Turk service. All participants had an Instagram account. Instagram is a free online sharing service for photos and videos. The subjects had to complete a questionnaire, which is also used in clinics for the detection of depression.
Over 100 images per participant analyzed and evaluated
More than 100 images were selected by the researchers from each participant. The pictures were then submitted to people for evaluation. These should rank the images on a scale of zero to five. For example, they should indicate whether the pictures were interesting, sad or happy. In addition, the photos were categorized based on their saturation, hue and number of faces, the experts add.
Saturation, brightness and filters can say a lot about depression
The researchers found in their study that both decreased saturation and brightness, as well as increased hue, can predict depression. In addition, depressed and non-depressed people use different filters, explain the scientists. Depressed people generally used a filter less often.
Algorithm detects depression with a probability of 70 percent
Early detection of depression allows for rapid treatment. For example, sufferers can be treated more effectively with simple but effective therapies for depression and anxiety disorders, the doctors say. The new algorithm made it possible to detect depression with a 70 percent probability.
Nowadays, too often antidepressants are prescribed for no reason
There are several different ways to treat depression successfully. It is not always necessary to have tons of medication. In the present time, even often unnecessary medications are prescribed. For example, antidepressants are often given without depression.
Sports and yoga can help with depression
In addition to the medicines, other methods for the treatment of depression can be used. For example, yoga training works very well in mental illness. In general, it can be stated that sport also helps against depression. This has already been clearly demonstrated by previous studies. (As)