The COVID-19 pandemic worsened the mental health situation in the Region and in the world, increasing new cases of mental health conditions and worsening pre-existing ones. It also produced significant disruptions in services for mental, neurological, and substance use disorders. Populations that historically faced an increased burden of mental health conditions and reduced access to treatment are now disproportionately affected by the mental health impacts of COVID-19.
In the Region of the Americas, mental, neurological, and substance use (NSM) disorders and suicide account for more than a third (34%) of the total years lived with disability, with depressive disorders being the leading cause of disability. Almost 100,000 people die by suicide each year in the Region.
Now more than ever, as the COVID-19 pandemic highlights and deepens longstanding inequities in mental health in the Region, it is essential that new solutions be worked on to make mental health care a reality. for all people.
However, one of the many problems that hinders society's ability to cope with these disorders is that the diagnosis of mental health problems requires specialists, whose availability varies dramatically around the world.
It is in this sense that the development of a machine learning methodology to facilitate the assessment of mental health could, for example, provide an additional means to help detect, prevent and treat these health problems.
How can AI help Mental Health?
Experts at the University of Texas at Austin are investigating how artificial intelligence (AI) can help young people dealing with mental health issues. “From text messages to social media posts, there are algorithms that can process that language and detect behavior patterns, including emotions and feelings,” explains Professor S. Craig Watkins, founder of the Institute for Media Innovation at Moody College of Communication.
Watkins teamed up with a team of graduate students from the Information School (iSchool) to investigate the power of what they call “values-driven AI.” Her project explores how AI-powered technology can remove or reduce barriers for adolescents or young adults seeking help related to their mental health.
Barriers can include a lack of knowledge about available resources, cost, and accessible sites, as well as the stigma and shame often associated with mental health. The objective of Good Systems, interdisciplinary university research, is to improve AI to use it for the benefit of society.
“AI can analyze the content people create, the conversations they participate in, the communities they are connected to, and what information they search for on the web. All of these things could serve to identify the onset of a mental health issue or someone currently dealing with issues," Watkins says, adding that while data privacy issues still need to be addressed, the potential for AI to improve access to psychological care and increase early detection of mental illness justifies the effort to find the right balance on the issue.
A matter of data
In addition, new research by an interdisciplinary team led by Denis Engemann of the National Institute for Research in Informatics and Automatics (INRIA) shows that machine learning from large population cohorts it can provide "rough measures" of brain-related health problems without the need for specialist evaluation.
Researchers tapped into the UK Biobank, one of the world's largest and most comprehensive biomedical databases, containing detailed and reliable data related to the health of the British population, to develop health-sensitive AI models mental.
The Biobank not only stores biological and medical data, but also data from questionnaires on personal circumstances and habits, such as age, education, tobacco and alcohol consumption, sleep duration and physical exercise, among other things . Specifically for this study, these questionnaires also include behavioral and sociodemographic data, such as the individuals' moods and feelings, and biological data include Magnetic Resonance Imaging (MRI) from 10,000 participants' brain scans.
Scientists combined these two data sources to build models that approximate measures of brain age and scientifically defined traits of intelligence and neuroticism. These are "surrogate measures" that correlate with specific diseases or outcomes that cannot be directly measured.
Development of approximations of this model have been used successfully in the past to predict 'brain age' from MR imaging. This body of previous neuroclinical work served as a starting point for Denis Engemann and his team.
“In this paper, we have generalized this methodology in two ways. First, we show that, beyond biological aging, the same framework of indirect measures is applicable to constructs more directly related to mental health. Second, we show that useful surrogate measures can be derived from data other than brain imaging, such as sociodemographic and behavioral data,” Engemann explains.
The researchers validated their proxy measures by demonstrating the same results on a separate subset of data from the UK Biobank. The results of this work offer a glimpse of a future in which psychologists and machine learning models could work hand-in-hand to produce increasingly accurate and personalized mental assessments.
For example, in the future customers or patients could grant a machine learning model secure access to their social media accounts or mobile phone data, then return indirect measures that are useful to both the customer as for the mental health or education expert.
Vocational training
For psychology students, finding psychometrics and data analysis subjects, among others, was a surprise when they expected to learn about people. However, we are now in a phase in which the information that an instrument/questionnaire can give us is nothing compared to what companies like Google, Twitter, Facebook or internet providers obtain from people's activity.
We share information by walking the smartphone, communicating with words or emoticons. When we are bad, our behavior changes and the decisions we make are reflected in our behavior. Having an email and sharing information, contribute as a whole knowledge about how we are. However, the data needs to be understood, interpreted culturally, socially and psychologically, that is where there is an intersection between the potential of artificial intelligence in structuring data, and the social sciences.
Professional-to-patient care is conditioned by numbers. People who can be served, and budget available to cover the service. Using artificial intelligence, attention can be maintained while we have more information on developments in mental health. In this way, we adopt promotion and prevention measures that are more appropriate to the behavior of the individual / social group. And if they cannot attend sessions, they still continue to generate and receive digital information.
Any field of mental health needs information. Regardless of the approach, information is needed on the what, precedents, and consequences of care. A person's ability to analyze all the interfering variables throughout life is limited the more patients they see.
Given this reality, the possibility presented by the current moment together with artificial intelligence requires digital literacy. While new interdisciplinary training is making its way, for the most part we continue training in degrees without any connection to other fields of knowledge. With COVID-19 impacting behavior globally, more digital literacy needs to be embedded from the start. Online training resources are increasing, the important thing is to learn to discern which ones add value.
Sources:
https://www.paho.org/es/campanas/world-mental-health-day-2021
Cyberhealth.es
unamglobal
https://www.consalud.es/tecnologia/inteligencia-artificial-evaluaciones-salud-mental_103865_102.html
Please follow and like us: