Artificial Intelligence (AI) is expected to transform the work context profoundly. Currently, multiple AI systems are being studied and applied in the mental healthcare field, challenging traditional ways of performing tasks by professionals. Chatbots are an example of an AI application being explored in mental health. Implementing AI in this field offers a potential solution to some problems associated with the availability, attractiveness, and accessibility of mental healthcare services.

In 2019, a study was conducted to understand different chatbots employed for mental healthcare. To do this, the authors used seven bibliographic databases to explore existing chatbots and their key features. The other systems were classified based on their purpose, the platforms they were built in, how they responded, and the initiative regarding their dialogues. Further, it examined how they generated responses, types of input/output, the disorders they dealt with, and whether the chatbots were embodied. The authors observed that most chatbots only allowed the user to interact by writing and that the most common way of providing feedback was a mix of text, audio, and imagery.

"AI has increasingly become relevant in today's mental health scenario. AI tools provide a wide array of information and serve as resource banks on mental health to the lay public," said Dr. Alok Kulkarni, Senior Consultant and Interventional Psychiatrist at Manas Institute of Mental Health and Neurosciences, in a conversation with INDIAai.

Speaking about AI in the mental health domain, Anju Bhandari Gandhi, a Professor at Panipat Institute of Engineering and Technology (PIET), remarks that Professionals in India uses AI tools rarely as they are confident in their analyses. However, the tend to seek aid from AI tools to derive final diagnosis.

AI and mental health

According to Alok Kulkarni, AI-powered chatbots and virtual assistants provide remote care for individuals in distress, although these are still at a nascent stage and are emerging tools that need longer-term data on efficacy and safety. Researchers increasingly use machine learning algorithms to analyze large datasets to predict neuroscience and mental health patterns. Such analyses are hoped to lead to more effective diagnostic and therapeutic interventions. 

"With the increasing popularity of AI, several web and mobile apps are used by people of different age groups. The users are trying to overcome their mental health struggles with the features available in these user-friendly apps", says Anju Gandhi.

As part of her research in Mental Health, Anju Gandhi has developed an app that can diagnose and predict depression. "The system suggests appropriate medications and solutions to overcome different conditions," she said.

Overcoming the challenges

Anju Gandhi believes that "competition, desires, and expectations create immense stress and make life more challenging for every person of all ages." AI can make diagnoses faster than any manual approach, like in-person counselling.

Alok Kulkarni identified the shortcomings of using AI systems. A person's mood state is an essential determining factor while assessing cognition. "Algorithmic approach may improve the turn-around times for cognitive assessments. However, these may come at the cost of validity of these assessments," he said.

Alok explained that AI algorithms cannot tease out the nuances of a person's complex cognitive and emotional landscape. A depressed patient will perform poorly on cognitive assessments. There's even a term to describe this cognitive state, and it's called 'pseudodementia'.

Challenges to job security

"Mental health, of all the fields, requires a human touch," says Alok Kulkarni. AI tools may make prediction errors while analyzing key results or data or erroneously interpret complex clinical scenarios. AI models should be subject to scientific scrutiny. Long-term data are needed about the safety and efficacy of AI tools.

Integrating cultural aspects into mental healthcare delivery is desirable, but AI-powered tools may not be efficient in integrating these local issues into care delivery. They are prone to perpetuating racial and gendered stereotypes and algorithmic biases. Therefore, complex clinical conundrums ultimately require the competence of a clinician. 

According to Anju Gandhi, AI cannot replace professionals in any field. AI is just a mechanism that can predict results based on previous studies. Patient history is stored in the system, and it is trained to give results based on new observations. Trained professionals are superior to all kinds of machines. AI exist just to add ease and accuracy to their profession. 

A long way to go

Though more people are willing to talk about mental health today, we have a long way to go. AI could be the key to tackling the hurdles in reaching the remote corners of a country like India. In every home of our nation, we expect one individual to be educated enough to operate systems, phones, etc. The available apps are very user-friendly and can be easily accessed by them. Professionals can run drives for mental health awareness and dissipate knowledge using such apps. AI can never replace the competence and empathy of a clinician. However, it can be key to solving numerous challenges the field faces today.