Speech-To-Text Technology: The mental health angle (Innovative Ideas Challenge)

Introduction

When I discovered that there was a category for innovative ideas in the Deepgram Hackathon on DEV, my joy knew no bounds. I have always believed that only with innovative and fresh ideas can technology, and by extension, the world make progress. This is because I believe innovation is the cornerstone of sustained economic growth and prosperity. That was why I applied for this category of the hackathon. Although prior to the Deepgram x DEV Hackathon, I had not encountered Deepgram, given my interest in Artificial Intelligence, I am not new to the concept of speech recognition technology. It is worthy of note that I am not only interested in Artificial Intelligence. However, my social impact activities has made me grown an interest in the field of mental health. Mental health is the state of well-being in which the individual realizes his/her own abilities, hence being able to cope with the normal stresses of life, work productively and fruitfully and above all, is able to make a contribution to his or her community. Without an iota of doubt, the issue of mental health is an important one that should not be treated with levity. Ergo, as much as I can, I endeavour to advocate for open conversations along with better awareness and understanding of mental health issues. This is why this particular submission resonates with me.

My Deepgram Use-Case

For context, in 2021, I was [opportuned to be amongst the 6% of undergraduate students] worldwide selected as a United Nations Millennium Fellow. The United Nations Academic Impact stipulates that selected fellows are required to work on a project of their choice in tandem with a Sustainable Development Goal(SDG) of their choice. I worked on the third Sustainable Development Goal (Good health and Well-being).

In general, there is a global shortage of mental health workers, with demand outstripping service provision. In a country like mine (Nigeria), reports suggest that we have as few as 0.1 psychiatrists for every 1,000,000 people. As an innovative young mind, whilst many see a problem, I see an opportunity. I opine that the insufficient number of mental health workers should prompt the utilization of technological advancement to meet the needs of the people who are affected by mental health conditions. Furthermore, the laissez faire approach to mental health in some parts of the world reduces the chances of persons with mental health disorders to get treatment.

It is estimated that mental health represents around 34% of the global disease burden, and with this predicted to increase, the NHS faces more pressure to meet these demands as ever. One solution that appears to be growing in popularity is the use of chatbots for the screening, diagnosis and treatment of mental health conditions.

Dive into Details

Chatbots are systems that are able to converse and interact with human users using spoken (our use case), written and visual languages. However, my idea would be focusing on the spoken languages use-case mainly for the ease and anonymity it offers. My belief is that chatbots have the potential to be useful tools for individuals with mental disorder, especially those who are reluctant to seek mental health advice due to stigmatization.
Given the growing availability of large health databases and corpora, a chatbot can be designed using several evidence-based therapies such as cognitive behavioral therapy, behavioral reinforcement and mindfulness to target symptoms of depression for users.
The major benefactors of this idea are people with depression, anxiety, schizophrenia, dementia, phobic disorders, stress and eating disorders.

Using Deepgrams SDKs which are supported for use with the Deepgram API, it becomes possible to create a chatbot that can convert the speech input made by users to text. This chatbot will depend only on a well-versed decision tree that will generate their responses. This implementation will be a web-based chatbot as opposed to stand-alone software for two reasons. Firstly, to use web-based chatbots, users do not need to install a specific application to their devices, thereby reducing the risk of breaching the privacy. Secondly, web-based chatbots are more accessible than stand-alone chatbots.

Conclusion

From participating in Deepgram hackathon Innovative Ideas challenge, I have learnt more about speech-to-text technology. Furthermore, I am super glad I could use this opportunity to advocate for open conversations about mental health issues whilst being innovative with challenges that relates to mental health.