Introduction
Mental Health has long been an overlooked medical subject. Although as the years pass by, we realize how important mental health is to our everyday lives and in turn how it can improve our individual lives and the communities around us when properly taken care of.
As we realized the importance of mental health and taking care of it, more people increasingly started seeking mental health counseling. With more and more people seeking counseling, we also noticed how vastly understaffed the mental health sector is. With that, we started finding other ways in which we can look over our mental health, and with that came artificial intelligence (AI).
Ever since the advent of AI, it has been very helpful in many different fields. It is used for shopping, advertisement, banking, transportation, communication, translation, and so many other human endeavors. With AI being used and being vastly successful in all these fields, it’s only natural that it has been explored in medical field as well. As it is, AI is used in many different medical fields and is ever so helpful in making the work of healthcare professionals easier and the experience for the patients smoother.
Even though there has been a lot of concerned as to whether AI should be used in tackling mental healthcare needs, it could be quite revolutionary in improving our mental health and in turn our daily lives.
In this article, therefore, we will be discussing the advantages and disadvantages of AI in mental healthcare, the steps already taken to include AI and the concerns many have about including AI in our mental healthcare sector.
History
The 1900s was when computers 1st started appearing, and with this appearance increasing over decades, scientists, engineers and others started thinking about creating robots. They thought about making them to understand humans in some sort of way so that they can help us with our tasks. An example of an AI that was developed back in the 1900s is one called ELIZA. ELIZA was programmed with a little bit of psychological understanding, so when it was tested it answered people’s question with rather simple answers but still that was quite a breakthrough and is one of the 1st AI breakthroughs that would help in bringing us to where we are today(Olawade, D. B., Wada, O. Z., Odetayo, A., David-Olawade, A. C., Asaolu, F., Eberhardt, J.).
The modern AIs are now used in mental health as a way to have early detection, continuous monitoring, individualized treatment, virtue therapy and so many others. One popular type of AI used for modern mental healthcare is the chatbot. The chatbot is basically an AI that is programmed to chat and respond to the users in an appropriate manner. Using all the information, it filters out the data and comes out with accurate and proper answers or responses. A couple of popular chatbots used for mental health are:
Woebot – A chatbot that is used in depression and anxiety therapies, it has proved to be quite effective.
Wysa– A chatbot that is used in improving the users overall mental health.
Talkspace – Online therapy that uses AI to match the patients and therapists that would be best suited for each other.
BetterHelp – Just like talkspace it uses AI to match patients and therapists, but betterhelp has a bigger pool of mental health issues that it addresses.
Moodfit – An app that uses AI to track and analyze the mood of the user and uses that information to help the user better stabilize their mood and mental health.
Headspace – An app that uses AI to help identify meditations and mindfulness exercises that best suit the user’s needs.
Calm – An app that offers meditations and stories to help the user get better sleep and peace.
Shine – An app that uses AI to help personalize daily inspiration and content to the specific user.
Cerebral – Uses AI to help therapists personalize the best treatment plans for individuals.
PTSD Coach – An app that uses AI to help provide the appropriate resources for managing PTSD to the user.
There are so many others like these and so many others that help battle many different mental health issues.
Machine Learning
You might be wondering how is it that these AI machines are even taught to understand and process the necessary data in order to give an accurate output. One of the popular ways is through ‘Deep Learning’; Deep learning technique (DL) is a type of machine learning in which computers and AI are taught to learn and process data using artificial neural networks almost similar to a human brain (Andrew, J., Rudra, M., Eunice, J., & Belfin, R. V.). A couple popular DL techniques are:
Artificial Neural Networks (ANN)- ANN is made to work similar to the way neurons work in a human brain, the way they are all interconnected and the way they work.
Convulsion Neural Networks (CNN)- This type is used primarily for image classification, it takes an image as input and the image goes through a process for the computer or AI to then create an output based on what it’s learnt of the image.
Recurrent Neural Networks (RNN)- Used for data with or relating to temporal features or relating to time. It takes data from sequences and time patterns to then help improve accuracy in diagnosis and treatment.
Generative Networks- The DL techniques need data to even work in the 1st place. Generative networks generate synthetic or artificial data, so they are in charge of making sure there is data when a part of an important data is missing. They ensure there is enough and consistent data being given to the DL to properly do its job.
Diagnosis with AI
Diagnosis of a mental health disorder or illness is a hard and tedious process. It takes several months and sometimes we cannot even tell until the illness has progressed beyond a certain limit. Scientists and researchers have been developing AI that can go through vast amounts of data of a patient through their social media usage, their behavior, medical history and so much more. Once the AI has gone through all that it analyzes it and predicts the mental illness of the patient or one that they have a possibility of, later in life, developing. This really helps with figuring out which patients are likely to develop a certain mental illness way before it gets too much and consumes them.
AI is also helpful in determining what mental illness a person may develop and help find ways to prevent it or minimize the impact.
It is shown that a lot of medical health professionals have an internal subconscious bias just like many other professionals from many different fields. A medical health professional may subconsciously believe or assume something in relation to a patient’s race, gender, cultural background, etc. With the use of AI in diagnosis, we can help eliminate a lot of that subconscious bias, and so get the most accurate diagnosis and treatment plan for all patients.
Treatment with AI
It is a known fact that in many things in life one size does not fit all, this also applies to mental health. There are many factors that contribute to every person’s mental health journey making it different from others. In traditional medicine there are many protocols to be followed which do not allow the healthcare professional to accurately diagnose and find an appropriate treatment for a patient based on their own unique experiences. AI though, will analyze each and every factor and experience of an individual to help better identify, diagnose and treat the specific mental health struggles that a person is going through.
There was a study published in the ‘American Journal of Psychiatry’ where it showed how with the use of AI to analyze brain imaging data. It could very accurately predict a patient’s body’s reaction to antidepressant medication. This could prove to be very helpful in the future of personalized medicine including that of mental health.
AI could also be used to analyze an individual’s data from social media, assessments and many other places. These analyses could help the AI understand a person’s routine, when their moods fluctuate, what in particular causes those mood fluctuations as well as things that cause an individual’s happiness. All this data collected would allow the AI to intervene in times that the individual seems to be feeling down the most. It allows the AI to help recommend things to better improve an individual’s mental health on a day to day basis and so helps with prevention and treatment.
Limitations & Concerns
One major limitation we have with AI in mental health care is the size and quality of the data. The quality and size of any data given to the computer is what its going to use to determine the output, so basically the input determines the output. A concern with this is that some people may be able to manipulate the output with what they input into the computer rendering the results inaccurate and unreliable.
Another concern is the absence of human compassion and empathy. AI is not able to understand and connect to humans in a way that another human being can and a lot of times this understanding between us is how professionals are able to figure out and comfort a lot of individuals’ mental health concerns. With a machine there isn’t such a thing and who knows if there ever will be, so that is a huge limitation on the part of AI.
One of the biggest concerns with using AI for mental healthcare is the lack of privacy. AI just gets fed a lot of an individual’s personal information for it to understand the individual better. This information may be accessible to some companies and people. If there is ever a glitch in the software there are no laws currently in place to hold the software developer accountable, and so all this information that had been gathered could be released and spread causing alarm to anyone whose data had been collected.
Conclusion
We have come a long way, from not taking mental health seriously, to having professionals specifically for mental health, to now AI that can diagnose and give treatment recommendations for multiple different mental health problems. With more time, research and effort put into this field of medicine we will be able to move our world to a much healthier and happier future.
One way in which we are thinking of improving our future is through AI wisdom. AI wisdom is a concept that is starting to be widely talked about. It is the idea of us giving AI the wisdom to understand human principles and morals. As for now, AI cannot reflect on its actions and decisions, it does not think about the diversity in human thought, perspective and morality. All these are very important things to take into account when diagnosing and treating any kind of mental illness and AI does not have the capability to do that and may never fully have that. Though we can always try, and with time we might be able to teach AI to understand these important things even if it is only a fraction of it. It would be worth it if this would mean that mental health concerns will be handled in a manner that will leave the majority of the world with easy access and wonderful treatment for the many mental health issues that ravage our societies.
Bibliography
Graham, S., Depp, C., Lee, E. E., Nebeker, C., Tu, X., Kim, H., & Jeste, D. V. (2019, November 7). Artificial Intelligence for Mental Health and Mental Illnesses: an Overview.
Olawade, D. B., Wada, O. Z., Odetayo, A., David-Olawade, A. C., Asaolu, F., Eberhardt, J. (2024, August). Enhancing mental health with Artificial Intelligence: Current trends and future prospects.
Garg, P. (2018, October 22). AI’s Potential to Diagnose and Treat Mental Illness.
Gupta, P. (2024, April 24). AI in Mental Health: Revolutionizing Diagnosis and Treatment.
Andrew, J., Rudra, M., Eunice, J., & Belfin, R. V. (2023, March 31). Artificial intelligence in adolescents mental health disorder diagnosis, prognosis, and treatment.
Ray, A., Bhardwaj, A., Malik, Y. K., Singh, S., & Gupta, R. (2022, February 12). Artificial intelligence and Psychiatry: An overview.
