Artificial Intelligence in Mental Health: Challenges in the Indian Context

Artificial Intelligence (AI) has emerged as one of the most promising technological aids. It refers to the field of computer science and technology that is expected to perform tasks that normally require human intelligence.^[1] AI has several components, including supervised and unsupervised machine learning, reinforcement learning, deep learning, natural language processing and computer vision, among others. ^[2]It has been incorporated in most of the important industrial sectors all over the world, and mental healthcare is no exception. AI in mental health has made significant progress, where tools utilising voice analysis, facial expression analysis and wearable devices that track real-time data on an individual’s behaviour are used for early detection and predictive models for psychiatric disorders.^[3]

Read More: Psychology and Artificial Intelligence

AI algorithms use patient data, including past treatment history, medical history, and genetic makeup, to suggest personalised treatment plans and reduce trial and error often associated with psychiatric treatment. Besides, virtual therapy chatbots like Woebot and Wyse have emerged as one of the promising psychotherapy delivery tools.^[3] Teletherapy enhancement by AI analyses real-time patient emotions and provide therapist with insights based on the same.^[3]

Read More: AI Chatbots for People with Disabilities: Bridging the Accessibility Gap 

There are several challenges to incorporating AI in the management of psychiatric disorders, like concerns about data security and privacy, transparency of AI algorithms and lack of clarity on the accountability for the services provided by AI. These issues have been raised worldwide.^[1] In this article, some of the challenges specific to developing countries like India are discussed along with possible solutions that can be worked upon. 

Acceptability

Patients still prefer in-person therapy and consultation sessions over virtual or teleconsultations, AI being predominantly associated with the latter. ^[4]Patients may not feel comfortable sharing their personal and emotional problems with an AI assistant or virtual chatbot. In such a case, the treatment gap due to the lesser availability of mental health professionals remains pertinent. Hence, AI should be used in collaboration with mental health professionals on aspects of psychoeducation, drug compliance monitoring and timely follow-ups. This will ensure better patient adherence to treatment, the acceptability of AI and reduce the burden on specialists.

Accessibility

Internet accessibility is still restricted, as a large chunk of the population resides in rural areas. Besides, the lack of technical know-how required for using AI-powered virtual platforms and devices, especially in the older generation, makes their use a distant reality.^[4] Initiative needs to be taken on improving the technical knowledge about the use of AI by government and non-governmental organisations.

Electronic medical records (EMR)

The role of AI in the diagnosis and management of psychiatric disorders relies heavily on the availability of EMR to access treatment history, laboratory reports and genetic reports. In India, the digitalisation of patient health records is in its nascent stage. While in the private sector, some of the big multispecialty hospital chains have digital data, the majority of the public sector is still reliant on physical health records. Furthermore, there is reluctance about data sharing among the hospitals that may slow down the incorporation of AI-powered tools.^[5] There is a need to gradually digitalise the public sector that manages a huge patient burden for AI to be used and improve the services countrywide. The process can be started from all the institutes of national importance, allowing data sharing and incorporation of AI, and later can be followed by individual state governments.

Diversity

India is a land of diverse cultures and languages. Most of the AI algorithms learn from historical data that may reflect social inequalities and result in biases. Besides, most of the AI tools are not available in vernacular languages. While the people residing in urban cities may adapt to the services, the majority of the population would fail to do so due to the cultural variations that affect psychiatric disorders.

The solutions given by AI chatbots that may be acceptable to one population may be counterproductive for another group of people owing to their prevalent cultural norms. The AI algorithms have to be made more culturally sensitive and adjusted accordingly based on feedback from feedback of specialist. Academic institutes from various regions of the country can be trusted with the process. 

Read More: Will AI Change the World for Everybody? An Inclusive Perspective

Stigma

Mental Health is still associated with a lot of stigma in our country. Though the use of AI-powered tools can be beneficial in such circumstances, as people may seek help in privacy without the need to physically visit mental health professionals. It may raise concerns about data security. People may tend to overshare personal information on such platforms, especially the free or low-cost applications, potentially risking their privacy in case of a data breach. Thus, awareness programs on topics like data security, role delineation of AI and human specialists should be undertaken.  

Conclusion

Among healthcare sectors, mental health hasa huge scope for enhancing its services via AI. India presents new sets of challenges owing to its diverse demography and culture, economic disparity and government initiatives. Acknowledging them and finding means to improvise them is the only way forward.