ava

we're leveraging voice recognition technology to enable better access to medical assistance

Comment

Inspiration

Current statistics show that >50% of mental health disorders go undiagnosed. Mental heath still remains a stigmatized area, with little to no penetration.

Cognitive Behavioral Therapy (CBT) for mental health disorders is expensive and requires you to connect with the therapist on a personal level. Research has shown that well administered exposure therapy is as effective as CBT. Scientifically designed exposure therapy can be administered using VR.

What it does

With Ava, we provide users an easy to use general and mental health checkup assistant.

The user visits our website/app and is taken through a simple questionnaire where he’s asked a set of questions designed by medical professionals to ascertain the user’s mental state, and information about the user’s lifestyle and health. The data entered by him will be collected and stored as part of his user profile. Every interaction the user has with the intelligent assistant “Ava” builds upon that. If the user preliminarily shows signs of distress, then they are requested to visit a healthcare provider to undergo further testing/diagnosis. If the user showed a likelihood for suspected mental health issues, certain questions that aid their diagnosis, designed by psychologists are presented to the user on the website while collecting EEG data through the Muse Headband. The EEG data points are used to generate a 3D brain activity heat map, which can be used to identify potential fear responses to certain stimuli, aiding in diagnosis. Finally, this also serves as a platform to help alleviate certain conditions through special exposure therapy designed on the VR platform, in conjunction with progress indicated via lower fear responses collected via the Muse Headband.

Ava can also give you information about possible interaction between FDA approved drugs relying on a web API of published medical studies.

How we built it

We used Bootstrap to create a responsive website, Node.js to interface the Google App Engine, along with DialogFlow, and the Cloud Speech API. We used Muse headband to collect EEG data, and the Oculus headset to run a simulation of an exposure therapy.

Challenges we ran into

We were unable to integrate the developed personal assistant to the website. We need to understand how to use the data from the Muse headset in realtime and perform analysis.

Accomplishments that we're proud of

Creating Ava using the Google Cloud Products, with speech input and output. Developing a Oculus VR exposure therapy instance. Being able to analyze Muse data, and generate a 3D brain activity heat map. Using an API to provide the user with data about adverse drug interaction possibilities.

What we learned

We learnt how to efficiently use bootstrap, integrate APIs into the virtual assistant, how to use Google Cloud services to create a personalized virtual assistant.

What's next for Ava?

Scaling up AVA to include more functionalities, to work with psychologists to create tests and therapies tailored to user specific mental health issues

Built With

Share this project:

Updates