|Year : 2021 | Volume
| Issue : 1 | Page : 45-47
Post-COVID-19 mental health service delivery in India: Potential role of artificial intelligence
Seshadri Sekhar Chatterjee1, Abhijit Dasgupta2, Abir Mukherjee3, Kaustav Chakraborty4
1 Department of Psychiatry, Diamond Harbour Medical College and Hospital, Kalyani, West Bengal, India
2 Department of Data Science, University of Kalyani, Kalyani, West Bengal, India
3 Department of Psychiatry, Medical Superspeciality Hospital, Kalyani, West Bengal, India
4 Department of Psychiatry, JNM Medical College, Kalyani, West Bengal, India
|Date of Submission||01-Aug-2020|
|Date of Decision||01-Sep-2020|
|Date of Acceptance||23-Sep-2020|
|Date of Web Publication||31-Mar-2021|
Dr. Seshadri Sekhar Chatterjee
Diamond Harbour Medical College, West Bengal
Source of Support: None, Conflict of Interest: None
|How to cite this article:|
Chatterjee SS, Dasgupta A, Mukherjee A, Chakraborty K. Post-COVID-19 mental health service delivery in India: Potential role of artificial intelligence. Indian J Soc Psychiatry 2021;37:45-7
|How to cite this URL:|
Chatterjee SS, Dasgupta A, Mukherjee A, Chakraborty K. Post-COVID-19 mental health service delivery in India: Potential role of artificial intelligence. Indian J Soc Psychiatry [serial online] 2021 [cited 2021 Oct 28];37:45-7. Available from: https://www.indjsp.org/text.asp?2021/37/1/45/312873
The COVID-19 pandemic has engulfed the world, exhausting the health resources of almost all the countries, causing seldom-seen-before health and socioeconomic consequences. The post-COVID-19 mental health crisis is inevitable and will be more intriguing and difficult to tackle, particularly in a diverse and billion-plus country like India. Owing to poor mental health professionals to population ratio and a meager amount of mental healthcare budget, urban-rural disparity leads to an imminent threat to society. There are four psychiatric nurses, two each of psychologists, and psychiatric social workers per ten million populations. Besides, only about 0.06% of the total healthcare budget is spent on mental healthcare. The rural areas that account for nearly 70%–80% of the population would have at the most two mental health professionals per million populations. Thus, technological help is an immediate necessity to overcome such an upcoming catastrophic situation. In these contexts, artificial intelligence (AI) can help in public mental health service delivery in India with judicious use of workforce and consuming resources.
| Artificial Intelligence and Psychiatry|| |
In AI, one can learn automatically with minimum human intervention and improve learning through more observations, examples, and experiences. Psychiatric diseases have multifactorial causations and high individual variability, and it can be altered by the epigenetic effects. Thus, in response to the emergency requirements of efficient and optimal psychosocial models, AI can specifically be the chalice of psychiatry. Here, we are going to discuss some potential roles of AI in India [Figure 1] for improving mental health services in the aftermath of the COVID pandemic.
|Figure 1: Illustrating potential roles of artificial intelligence in post-COVID-19 mental health service delivery in India|
Click here to view
| Telepsychiatry|| |
Teleconsultation is on the rise, especially in psychiatry, due to comparatively lesser need of physical interaction, increased need for follow-up visits, etc., However, there are some inherent limitations in using those sites and platforms.
AI-based classification algorithms can help to evaluate ones' mental health condition with the help of their case/follow-up history stored in a particular database at the end of the consultants in a specific teleconsultation platform. Here, the electronic health record of patients can be considered as the input to the aforementioned classification algorithms. In addition to the predictive models, smartphone app-based on AI can be developed for real-time monitoring of contextual information and treatment delivery. Such mental health monitoring systems may give additional information during consultation.
| Diagnostic Aids|| |
Other than the prevailing syndromic classification, many other models such as research domain criteria, cross-cutting model, and nonclinical biomarker specifiers are coming up. Even with the implementation of diagnostic and statistical manual of mental disorders (DSM–5), the validity of psychiatric diagnosis remains a pressing issue. To validate the predicted diagnosis by AI-based tools, available clinical and research neuroimaging data as well as databases (e.g., Attention Deficit Hyperactivity Disorder -200, OASIS, and many more) of worldwide patients suffering from various mental and neurological disorders may be analyzed.
Nowadays, recent deep learning-based AI algorithms, such as a convolutional neural network, recurrent neural network, autoencoders can be utilized to classify different mental disorders with their distinguishing characteristics (features) based on aforementioned databases with sufficiently good accuracy than conventional machine learning (ML)-based algorithms. For example, it may be mentioned that such a deep learning framework can successfully diagnose dementia and ADHD, as cited by recent studies.
| Risk Prediction|| |
There is accumulating evidence that AI can predict the risk of disease occurrence, particularly, when suicide is on the rise. Digital psychiatric tools based on AI and ML can detect high-risk individuals at an early stage. As a result, it helps in deciding efficient and timely preventive measures to avert major crisis and mortality. These applications can efficiently predict the probable suicidal tendency of an individual early enough to save one's life. These can also predict relapse in chronic and episodic diseases by measuring comprehensive screening score and endophenotype detection.
| Personalized Medicine|| |
Personalized medicines with appropriate doses at different stages of a particular mental disorder and their treatment responses can optimally be predicted by AI-based predictive models based on both supervised and unsupervised classification algorithms, such as support vector machine, random forest, and fuzzy C-regression model clustering among others. Thus, these predictive models can help general physicians as well as psychiatrists.
| Psychotherapy|| |
As a prerequisite to psychotherapeutic intervention, the therapist needs to gain substantial knowledge of the clinical history, personality structure, and thorough process of a particular patient. Such knowledge is essential to find novel treatment solutions in lesser time. However, a high degree of commitment on the part of therapist and patient, time-consuming sessions, and high cost make psychotherapy difficult to implement in India. To overcome such constraint, AI-based application coupled with Santiago's theory of cognition, general system theory, and many other psychological theories can understand patients' susceptibility and genesis of the index problems. Therefore, an embodied AI-based system, chatbot-based psychotherapy have the potential to deliver psychotherapeutic interventions with little involvement of a specialist and thus can be implemented cost-effectively to deal with the upcoming postpandemic mental threat in India.
| Public Health|| |
Recent advancements in the field of ML for natural language processing can provide useful insight to understand the public sentiment (for example, fear, panic, anger, and uncertainty among others) by analyzing social media data. A recent study has conceptualized such aforementioned application to develop two computational models based on social media information, such as the medical-mental-social-global model and assertive team-time limited-peer support model. Such methodologies inspired the Indian psychiatric society to develop an AI-based android mobile application to screen individuals suffering from postpandemic mental illness and encourage them to seek psychiatric intervention.
In conclusion, AI-based diagnosis and treatment model scan become a cost-effective alternative to face-to-face psychiatric consultation in the postpandemic era. AI-based tools and algorithms can efficiently predict diagnosis and suggest suitable intervention based on the past assessments of thousands of patients and the experience of mental health professionals. However, AI is still at a very fetal stage and should ideally never replace face-to-face clinical interview or intervention; rather, it can aid to minimize human error, allows a well-maintained patient to avoid unnecessary travel for his/her follow-up, saves money, and diminish the magnitude of community spread of infection in the present scenario. We pitch for large scale field studies to substantiate the effectiveness of AI in a developing country like India.
We thank Uttirna Halder, B.Tech, Bengal Institute of Technology, Kolkata for the illustration in the manuscript.
| References|| |
Chatterjee SS, Vora M, Malathesh BC, Bhattacharyaa R. Worried well and Covid-19: Re-emergence of an old quandary. Asian journal of psychiatry. 2020.
Das SK, Conti M, “Editorial: Pervasive and Mobile Computing, Volume 52,” Pervasive and Mobile Computing, vol. 52, pp. A131-A134, Elsevier B.V., Jan 2019.
Durstewitz D, Koppe G, Meyer-Lindenberg A. Deep neural networks in psychiatry. Mol Psychiatry 2019;24:1583-98.
Ćosić K, Popović S, Šarlija M, Kesedžić I. Impact of human disasters and COVID-19 pandemic on mental health: Potential of digital psychiatry. Psychiatr Danub 2020;32:25-31.
Xi K, Huiyu L, Xiaohua H, Meihua W, Guangfeng W, Fang Z, et al
. Fuzzy regression model clustering algorithm based assessment, analysis and acceptance of palliative care and psychological behaviors of elderly patients with cancer pain. Cluster Computing 2019;22:4813-20.
Mello FL, Souza SA. Psychotherapy and artificial intelligence: A proposal for alignment. Front Psychol 2019;10:263.
Singh O. Chatbots in psychiatry: Can treatment gap be lessened for psychiatric disorders in India. Indian J Psychiatry 2019;61:225.
] [Full text]
Samuel J, Ali GG, Rahman M, Esawi E, Samuel Y. Covid-19 public sentiment insights and machine learning for tweets classification. Information 2020;11:314.
Wang M, Jiang W, Cheng P. Mental health peer to peer support via social media practice reference during COVID-19 Pandemics. PsyArXiv. May 3. doi:10.31234/osf.io/u4pqy.
Malathesh BC, Chatterjee SS, Das S. Overview of mental health issues of COVID-19: Need of the hour. General Psychiatry 2020;33:e100233.