What Makes AI in Mental Health So Appealing?

“How are you doing today?” “What’s going on in your world right now?” “How do you feel?” These might seem like simple questions a caring friend would ask. However, in the present day of mental health care, they can also be the start of a conversation with your virtual therapist.

Innovative technology is offering new opportunities to millions of Americans affected by different mental health conditions.

Advancements in artificial intelligence (AI) are bringing psychotherapy to more people who need it. Nonetheless, the benefits of these methods need to be carefully balanced against their limitations. The long-term efficacy of the AI approach regarding mental health is yet to be tested, but the initial results are promising.

Mental Disorders Are the Costliest Condition in the U.S.

According to the National Institute of Mental Health (NIMH), one in five adults in the United States (17.9 percent) experiences some type of a mental health disorder. Mental illness not only reduces an individual’s quality of life, but it also links with increased health spending.

Charles Roehrig, founding director of the Center for Sustainable Health Spending at Altarum Institute in Ann Arbor, Michigan, notes that mental disorders, including dementia, now top the list of medical conditions with the highest estimated spending.

Mental health is now the most expensive part of our health care system, overtaking heart conditions, which used to be the costliest.

Approximately $201 billion is spent on mental health annually. As more people reach old age, increasing prevalence of certain health conditions, such as dementia, is expected to increase this figure higher, with accompanying calls for new management strategies.

Due to the costs associated with treatment, many individuals who experience mental health problems do not receive timely professional input. Cost is not the only contributing factor; other reasons include a shortage of therapists and the stigma associated with mental illness.

A Friendly Robot for Personalized CBT

Clinical research psychologist Dr. Alison Darcy created Woebot, a Facebook-integrated computer program that aims to replicate conversations a patient might have with his or her therapist. Woebot is a chatbot that resembles an instant messaging service. The digital health technology asks about your mood and thoughts, “listens” to how you are feeling, learns about you, and offers evidence-based cognitive behavior therapy (CBT) tools. Interactions with Woebot aim to emulate a real-life face-to-face meeting, and the interaction is tailored to the individual’s situation.

However, Darcy is careful to point out that Woebot is just a robot and cannot replace human connection. Also, some people might require different types of therapeutic engagement and treatment than a virtual session can provide. Nonetheless, many experts agree that options like Woebot make CBT more accessible to a modern generation that chronically lacks time and is accustomed to 24/7 connectivity.

This carefully designed software offers private sessions that do not need to be pre-booked, and are affordable.

Woebot is not the first attempt to treat people by placing them in front of an avatar. Other attempts have been made to improve people’s mental health using chatbots. Some of the early chatbots were designed in the 1960s at the MIT Artificial Intelligence Laboratory. Their program ELIZA was able to simulate a short conversation between a therapist and a patient and is considered the grandparent of current systems being used today.

Advances in natural language processing and the popularity of smartphones have made chatbots the new starlet of AI in mental health care.

Chatbots are constantly improving to become more human-like and natural. They also offer different language options. For example, Emma speaks Dutch and is a bot designed to help with mild anxiety, while Karim speaks Arabic and has been assisting Syrian refugees struggling to cope after fleeing the atrocities of war.

Both programs were designed by Silicon Valley startup X2AI. Currently, the company is promoting its latest psychological AI product—Tess. Tess can perform CBT, as well as purportedly improve the burnout associated with caregiving.

What Makes AI in Mental Health So Appealing?

When evaluating the use of chatbots in health care, the International Committee of the Red Cross notes in its 2017 report that initial reviews of the messaging-app bots have been mixed. While it has been recognized that they are not expensive and are easy to deploy, some limitations have also been described, such as technical glitches. Furthermore, robots do not have a mind of their own; they follow a pre-defined script. Therefore, they are not always able to understand the user and his or her intent. Therefore, some experts suggest that this medium probably should be used in conjunction with a human therapist to ensure nothing gets missed.

Nonetheless, some initial studies on the efficacy of chatbots for mental health have been promising.

The first randomized control trial with Woebot showed that after just two weeks, participants experienced a significant reduction in depression. Furthermore, a high level of engagement was observed, with individuals using the bot nearly every day.

A virtual therapist named Ellie has also been launched and trialed by the University of Southern California’s Institute for Creative Technologies (ICT). Initially, Ellie was designed to treat veterans experiencing depression and post-traumatic stress syndrome.

What is so special about the technology is that Ellie can detect not only words, but also nonverbal cues (e.g. facial expression, gestures, posture). Nonverbal signs are very important in therapy, yet can be subtle and difficult to pick up. The ICT team led by Louis-Philippe Morency and Albert “Skip” Rizzo developed their virtual therapist so it can gather and analyze multisensory information and help assess a user. Ellie’s creators argue that this virtual human can advance mental health and improve diagnostic precision.

Some studies show that we react to avatars as if they were real humans.

Mel Slater of University College London, UK, and her colleagues observed this behavior when they conducted experiments where people were aware that they were interacting with robots, yet they related to them as if they were real.

Some psychologists also argue that we find it easier to share potentially embarrassing information with a virtual therapist. In human-to-human interaction, there is often a degree of self-restraint. Shame can prevent people from sharing openly with another person. However, when sitting with a virtual therapist, subjects were found to be more willing to express themselves, which could have an important therapeutic advantage. When patients talk to a psychotherapy bot, they report not feeling judged. Ellie, Karim and Woebot can make them feel at ease. In addition, robots are always available and can offer a much higher frequency of therapeutic interactions compared to a human therapist.

Preventing Social Isolation Among Young People Using AI

Social networking is very important for young people dealing with mental illness. Extreme social isolation and difficulties building close relationships are often a feature of their lives. Therefore, social networks on the Internet can foster a sense of belonging and encourage positive communication. Although the benefits of online health communities have already been widely recognized, scientists are now tapping into the potential AI can play in making people feel socially more connected.

Simon D’Alfonso of The National Center of Excellence in Youth Mental Health in Melbourne, Australia, and his colleagues have been working on the Moderate Online Social Therapy (MOST) project. The MOST model is being used with young people recovering from psychosis and depression. The technology helps create a therapeutic environment where young people learn and interact, as well as practice therapeutic techniques.

The MOST system has several parts, including The Café section where users can share experiences and gain support and validation from other members. Users can also nominate a problem in the Talk It Out section where problems get solved in a group. Or, they can engage in a behavioral task that uses mindfulness and self-compassion in a Do It! section of the site.

MOST has been used in a series of research trials and was evaluated as a viable mental health tool. Currently, the program is facilitated by human moderators. However, designers of the system plan to eventually replace humans with innovative AI solutions. User content is being analyzed so in the future, an individualized therapy might be offered.

D’Alfonso’s team is also looking to connect with other systems and provide appropriate mobile notifications. For example, if an anxiety attack is detected by the user’s wrist sensor, MOST could immediately offer therapy input on an individual basis.

Virtual Counselor to Reduce Student Stress

Another AI mental health innovation, this one aimed at young people, has been developed by a multidisciplinary group of scientists from Australia and China. They have been pilot testing a novel virtual advisor for university students.

Manolya Kavakli, associate professor at the Macquarie University in Sydney, is leading this project that aims to help students develop better coping techniques, particularly in connection with exam stress. Exams often put tremendous pressure on young people, which can have negative health implications such as depression, insomnia, and suicide. When exposed to excessive stress, timely counseling can be imperative to maintaining health.

Kavakli and colleagues proposed a virtual companion that can be readily available to provide support. Based on preliminary tests, the group believes that the embodied conversational agent they developed could be very useful during busy exam periods.

The virtual counselor mimics a psychologist and offers advice and support with stress management.

During their pilot studies, researchers also wanted to establish how to design a virtual therapist so it was better accepted by users. They found, for example, that the voices of male virtual advisors were perceived as more credible and pleasant. Female voices, on the other hand, were assessed as clearer, more competent and more dynamic. This could have interesting implications on the way virtual intelligent systems and chatbots are designed in the future—developing different personas to maximize the effect of the treatment on the end-user.

Adopted from original article 

Sources:D’Alfonso S., Santesteban-Echarri O., Rice S., et al. Artificial Intelligence-Assisted Online Social Therapy for Youth Mental Health. Frontiers in Psychology, 2017; 8:796.Fitzpatrick KK, Darcy A, Vierhile M. Delivering Cognitive Behavior Therapy to Young Adults With Symptoms of Depression and Anxiety Using a Fully Automated Conversational Agent (Woebot): A Randomized Controlled Trial. JMIR Mental Health 2017;4(2):e19Kavakli M, Li M, Rudra T. Towards the Development of A Virtual Counselor to Tackle Students’ Exam Stress. Journal of Integrated Design & Process Science, 2012;16(1):5Roehrig C. Mental disorders top the list of the most costly conditions in the United States: $201 Billion. Health Affairs, 2016; 35(6):1130-1135.Slater M, Antley A, Sanchez-Vives M, et al. A Virtual Reprise of the Stanley Milgram Obedience Experiments.Plos One, 2006;1(1).


What do you think?

234 Points
Upvote Downvote


Leave a Reply

Your email address will not be published. Required fields are marked *



Nikola Tesla on How Technology Unleashes Women’s True Potential

How to Boost Testosterone — Naturally