Brave New World of AI in Mental Healthcare

With the start of the new year, I want to start with the hopeful topic of artificial intelligence, and its first cousin generative AI. There has been much news coverage about AI including a recent New York Times article breathlessly covering significant investment in AI startups even as legacy tech company continue with mass layoffs. You might have also listened to the LifeStance podcast episode where a LifeStance clinician interviews Chat GPT, a chatbot that is able to respond to questions in a more natural human way and can also quickly to admit mistakes and reject falsehoods. To me, a mental health chatbot seems to fit right in a sci-fi thriller, but it will be some time before we see them in use in a clinical setting. However, there are already many ways that AI is being used to deliver improved mental healthcare.

The mental health crisis accelerated by the pandemic continues to play a role in rising rates of anxiety and depression . 84% of psychologists who treat anxiety disorders have seen increased demand for anxiety treatment. Researchers are developing AI tools to help therapists provide care for this increase in demand. While no one can predict how widespread AI applications may occur in the mental health space, there are many areas where AI is already being deployed to assist therapists and clinics to provide better and more care. AI can complement mental health practitioners in the areas of personalized treatment plans, development of therapist training and delivery of as-needed mental health care support.

The ultimate quality control companion

Artificial intelligence is part of the progression of technology that started with buzzy words like big data, internet of things and machine learning. In essence, the AI tools in mental healthcare rely on analysis and algorithms of large amounts data. There was a recent study reported in Translational Psychiatry that used machine learning to sift through symptoms and blood marker data to identify PTSD cases in veterans of the Iraq and Afghanistan wars. The process helped identify military service-related PTSD so that appropriate mental health care could be provided to those most at risk.

Another area where AI has been applied most quickly in mental health care is in creation of treatment plans and even in analyzing therapy sessions. A company called Lyssn AI works with clinics and universities to monitor and assess every single text, call, or video so therapists can more easily find clues and patterns to apply the right treatment plan to a patient. Lyssn AI is constantly improving and can even help therapists identify the words and phrases that have the?most impact with patients.

Greater Access to Therapist Training

The Trevor Project, an organization that prevents suicides in a suicide prevention and crisis intervention organization for lesbian, gay, bisexual, transgender, queer & questioning (LGBTQ) young, has also thought about how to best use AI in support of their mission. They concluded that the best way to integrate AI into their organization is to address the urgent need for more counselors by using AI to train them more quickly. The Trevor Project partnered with Google.org to launch The Crisis Contact Simulator , a first-of-its kind counselor training tool powered by AI. I like how The Trevor Project has integrated AI in a way that also acknowledges the importance of person-to-person connections between their crisis counselors and at-risk LGBTQ youth.

As-Needed Health Care Support

To me, one of the great promises of robots is the ability to interact with us in meaningful ways and come to our rescue when we need it. Certainly, Star Wars shares this point-of-view as their anthropomorphic droids express human emotions and motivations while saving the day. While we are unfortunately, far away from having our own BB-8 or B2EMO, AI is being used to help ease loneliness and address depression.

One study included the use of a chat AI called Woebot that used cognitive behavioral therapy to provide on-demand content to treat substance abuse in patients. The app is designed to address cravings and help patients understand their own patterns of craving, loneliness and anxiety. It also contains an empathy component to simulate the human relationship to encourage its use. At the end of the two week trial, Woebot reduced depression symptoms.

Other uses of AI in mental health treatment are in actual robots. New York City purchased over 800 robots called ElliQ to help combat loneliness in the elderly. It regularly conducts wellness checks using a Siri/Alexa-like method. While ElliQ looks like a tech gadget other AI-enhanced robots have humanoid or animal forms. Pepper is a humanoid robot that recognizes faces and basic human emotions. Pepper is optimized for human interaction through conversations and his touch screen and can remind patients to take medication, encourage exercise, and provide companionship.

Perhaps the best known care robot is the baby seal Paro . Numerous studies and articles have documented its ability to alleviate depression and anxiety. It is also the only robot approved by the FDA as an insurance cost. Paro responds to its name and, when petted or stroked will coo and repeat behaviors to attract more affection. Licensed nursing homes can even get these robo-seals for free. Beyond the embedded AI, Paro is so cute and it could be a worthy replacement for an actual dog.

While there is great potential in the use of AI in mental healthcare, its efficacy is still uncertain.

The use of AI in therapy requires safeguards and more testing as trust is the most important factor in the road to recovery in mental healthcare. The relationship between the therapist and the patient is at the center of mental healthcare, so even if AI is able to overcome some major challenges, the technology is unlikely to be patient-facing in mental health