Video Link (Youtube)
EMTs and other medical personnel spend too much time filling out paperwork. In the midst of a pandemic and shortage of both medical supplies and workers, our team decided to create a project to increase the amount of people our healthcare workers can save, however much is possible.
What it does
This project allows EMTs and other medical personnel to focus on saving lives while our application fills out the paperwork. By parsing their speech, MIRI gathers critical information about the patient and creates a preliminary patient report while our expert is still out on the field. This includes identification, blood pressure and heart rate measurements, mechanisms of injury, trauma, patient history, as well as parsing out miscellaneous information and creating notes at the bottom.
How we built it
Based on our previous experiences, our team decided to build a web application on Flask and use Python with Google-cloud for the logic.
Challenges we ran into
Natural language processing is a problem that still hasn't been solved. Although the scope of language in our project is small compared the the expanse of the English language, there's no tools which process such a specific subset of language. Thus, our team had to combine Google's speech-to-text with various patterns to parse out the information we wanted.
Accomplishments that we're proud of
A working prototype which is able to parse relevant information into a neat package to ship to the pdf-filler.
What we learned
Working with all the new technologies our project required, filling out PDFs and speech-to-text.
What's next for Miri
As Miri continues to be refined, we strive to increase the accuracy of reporting and save even more of our life saving personnel's precious time by pre-populating fields or automatically sending the form to our hospitals. Additionally, our app could expand it's functionality through a dedicated mobile app which may be better than a simple webpage.
Try It out
flask, google-cloud, google-web-speech-api, pyaudio, python