Mental Health Support System

The project is aimed at addressing mental support gaps in India using AI

Comment

Inspiration

In India, most of them do not understand nor are they aware about individual mental issues. Most individuals go through some kind of mental challenges. Children committing suicide when they fail, farmers committing suicide when they cannot pay loan. Family members, elders usually resort to yelling, ignoring, beating as a solution. Doctors are not available in this mental medical field. India needs help to stay mentally happy and healthy. Hence I was inspired to fill this gap in mental support.

What it does

We want the mental health support system to talk, see, hear like a psychologist and be there to guide, support towards any mental challenges using AI.

How we built it

We will have the following components in the system

  1. Data capturing using NLP
  2. Data capturing using Computer Vision
  3. Data transferring from local edge devices like mobile or computer or smart watch to main server using AWS 5G
  4. Data Engineering service to collect, clean, format the data using AWS Glue, SQS, SNS, Event Bridge kind of tools and send it to down stream processes
  5. Data Analyst and BI service depicting the clean data in visual intuitive format
  6. Custom Gen AI model to collect and process all data using AWS Bedrock using existing dataset or our own source data
  7. Comparison on capabilities of any existing relevant AI tools in health and come up with a possible design using them all together as needed

Challenges we ran into

The needed AWS services are not free and it's a financial hurdle due to which we could not try things on AWS itself. So we tried our prototype with free tools that were available. We did not have enough device resources (eg. EEG etc.), skills and team strength to try out all modules.

Accomplishments that we're proud of

  1. High Level Architecture was created after studying AWS platforms and tools Architecture Video: https://youtu.be/PaikUEDkD7g?si=W4Z6JnlDFrbKi0fM&t=47

  2. Mental Health Machine Learning Linear Regression Model was implemented using pre-generated dataset consisting of features sleep,nutrition,brain,physical health,hydration and predicted label being mental health Demo Video: https://youtu.be/PaikUEDkD7g?si=rMAj3EQdyov-LFKu&t=104 Technology: Machine Learning in Python, Google Colab GitHub: https://github.com/8SLabs/Mental-Health-Support-System/tree/master/machinelearningmodel

  3. A preliminary chat bot was implemented that could chat with the user Demo Video: https://youtu.be/PaikUEDkD7g?si=8itZnZNmgG4GOIYp&t=181 Low Level Design video: https://youtu.be/Kimu674g1nk?si=5WwCsOQhR8HBotLg Technology: Google Gemini GitHub: https://github.com/8SLabs/Mental-Health-Support-System/tree/master/chatbot/chatbot

  4. A simple web site was created to gather profile information about the user Technology: NextJS on Vercel Cloud, MongoDB on MongoDB cloud, Vercel cloud to host Chatbot-FrontEnd Integration GitHub: https://github.com/8SLabs/Mental-Health-Support-System/tree/master/webinterface

  5. Studied how Python Medical Library can be utilized in our module Education Video: https://youtu.be/B35DDGAyJrw?si=N9qybQmeBDwZpUYb

What we learned

The linear regression model, chatbot, a simple website gave us a little confidence that this project is feasible, if we continue working on this product

What's next for Mental Health Support

We would like to continue working on this solution , test it in collaboration with a mental health supporting NGO and make it into a product offering sponsored via NGO or government of India. At some point, we want to migrate our solution into AWS to support scalability, performance, agility requirements across India.

Built With

Share this project:

Updates