To find out more about a friend, client, or even the next crush, exploring social media is a common path. If likes, selfies and comments give us so many clues about us, how much can technology say about our mental health? This is what a new current of science investigates.
Analyzing messages on Facebook, coloring photos on Instagram and even measuring the time between clicks are on the radar. The hypothesis is that data collected by smartphones can be used to identify patterns of behavior and social interactions. Without replacing psychologists and psychiatrists, but to assist face-to-face consultations. The model grows, as does the ethical debate (More information on the side).
In such a survey, a group of teenagers answer cell phone questionnaires about how they feel. They can be audios and even emojis to narrate emotions. On a daily basis, an application on their cell phones captures fragments of ambient sounds and measures the movement of the devices. Everything is analyzed to know the risk of depression – initial results come out this year.
“The big challenge is not capturing and processing data. The question is how to make sense of them”, says Christian Kieling, professor of Child and Adolescent Psychiatry at the Federal University of Rio Grande do Sul (UFRGS), who is in charge of the project, which monitors 150 teenagers via smartphones. Among the volunteers, there are young people already diagnosed with depression, at high risk of having the disorder and at low risk, according to a scientifically validated scale.
In the audios, they evaluate the content and the form. The app captures, every 15 seconds, samples of ambient sound. And there’s the confidentiality pact: scientists don’t hear the conversation, but they know the number of voices, to measure social interaction.
The app collects geolocation data and activity and rest patterns – it is allowed to turn off at any time. Therapies against depression encourage connection and physical activity. Information about interactions and spatial movement can facilitate personalized interventions. The group should also have consultations with psychiatrists, blood tests and resonance.
MAP.
Another study, linked to the Federal University of São Carlos (UFSCar), foresees the technology to help in the early identification of possible depressive profiles. The work started in 2021, after the suicide of a student. A computational model will analyze student texts on Facebook.
The tool, created at UFSCar in partnership with the Federal do Triângulo Mineiro (UFTM) and the George Mason University (USA), tries to “read” words and expressions that indicate a possible depressive profile. The robot is smart, but when deciphering the writing, intonation and irony escape it, for example. “It is not because it has processing power that artificial intelligence is better than us”, says Helena Caseli, professor of Computing at UFSCar.
For more robust analysis, physiological signals (heartbeat and sleep patterns) will be collected through smart watches. The results can serve as an “epidemiological map” – and institutional strategies for student well-being – as well as individualized analyses. One of the strengths is to compare data from a patient today with previous information about him and see any changes.
EMOTIONS.
For Felipe Giuntini, a researcher at Sidia, a center for innovation in digital solutions, it is possible to see a pattern of emotions in the data processing of networks. In his doctorate at the University of São Paulo (USP), he collected publications on Reddit, a popular social network in the United States, for ten years.
Posts – including emojis – were selected from a support group for people with depression. The analysis mapped words like “sadness”, “shame” and “enthusiastic” to see patterns and learn from the network itself. For Giuntini, the algorithm helps to understand patients’ mood swings.
On another front, the idea is to take those who are still far away to the office. “The person goes to the cardiologist and discovers an arrhythmia at the check-up. This does not happen in mental health”, says Alexandre Loch, from the USP Institute of Psychiatry. The average delay from the first signs to the diagnosis of Obsessive Compulsive Disorder (OCD), for example, is 11 years.
Loch tests software to assess images of the face and speech of volunteers, aged 18 to 35, in face-to-face interviews. Computer analyzes track pauses in speech, eye movements, gesticulation, and disconnected speech – aspects that would be noticed by the psychiatrist at the consultation. But who leaves the factory understands divans? A study with artificial intelligence to detect skin cancer at Stanford University (USA) shows that the algorithm discerned lesions like a dermatologist. In Psychiatry, each one exposes anger or sadness in a way. “As it is more subjective and symbolic, it is difficult for the machine to learn”, says Loch.
The information is from the newspaper. The State of São Paulo.
know more
+ SP: Man dies standing, leaning against car, and scene scares residents on the coast
+ One twin became vegan, the other ate meat. Check the result
+ Reincarnation in history: an age-old belief
+ Andressa Urach asks for money on the internet: ‘Help me pay my card bill’
+ Horoscope: check today’s forecast for your sign
+ CNH: see what you need to know for the application and renewal
+ See which were the most stolen cars in SP in 2021
+ Expedition identifies giant squid responsible for ship sinking in 2011
+ Everything you need to know before buying a crockpot
+ US Agency warns: never wash raw chicken meat
+ What is known about fluorone?
+ Trick to squeeze lemons becomes a craze on social media
+ IPVA 2022 SP: see how to consult and pay the tax
#Therapist #cell #phone #Scientists #test #technology #mental #disorders