MediScribe

User Persona
Landing Page
Main Page
Transcription and Highlighting
Medical Term Definition and Translation
Previous Transcription

Inspiration

Around 1 in 5, British Columbians do not have a family doctor. That is about 1 million people. Not only is it difficult to find a family doctor, but it can also be difficult to find a doctor that can speak your native language if English isn’t your first language. Especially when medical jargon is introduced, it can be overwhelming trying to understand all the medical terms and for older immigrant parents, this disconnect of language can serve as a deterrent to seeking medical aid. We want to create a solution that allows users to have a better relationship with their healthcare professionals and a better understanding of their own health.

What it does

Our solution “MediScribe” is a web application that transcribes and stores what the doctor is saying while highlighting medical terms and providing simpler definitions and offering translations in different languages. The raw transcriptions can be accessed at a later time and can be sent to others via email.

How we built it

We used MongoDB, ExpressJS, ReactJS, and NodeJS to handle the frontend and backend. We also incorporated multiple APIs such as the Microsoft Speech-to-Text API to transcribe the audio, Merriam-Webster Medical Dictionary API to define medical terms, Microsoft Translator API to translate medical terms to different languages, and the Twilio SendGrid Email API to send the transcript to the person the user chooses.

Challenges we ran into

We ran into difficulty configuring the speech-to-text API and having it listen non-stop. The tech stack was relatively new to us all, so we were spending a lot of time reading documentation to figure out the syntax.

Accomplishments that we're proud of

We were able to create all the functionality we planned for and our web application looks very similar to the Figma draft. We were able to learn on the fly and navigate a tech stack that we were not very familiar with. Furthermore, we were proud of the end product that we were able to produce and the experiences we gained from it.

What we learned

We learned the divide and conquer method of working on a project where we work on our individual components, then figure out how to integrate all the parts. We also learned how to use various APIs and the different methods to configure them.

What's next for MediScribe

We are planning to increase the translation capabilities to more languages and offer a setting so that all the text on the website can be translated into a language of choice.