We wanted to enhance the diagnosis process that the medical students learn in their coursework at school and through experience. Making a correct diagnosis for a patient is a skill that is developed over time as one gets exposed to more scenarios, and we wanted to use this idea to help students better understand medical diagnosis.
What it does
It is an application that allows students to toggle different diagnosis metrics for different organs, which then shows the likelihood of the disease based on training done on huge datasets.
How we built it
Challenges we ran into
It was our first time using machine learning algorithms to predict real world data and exploring the classifiers and using it in a real world context, and so we can get a better understanding of the current data we have by analyzing large sets of data
Accomplishments that we're proud of
We are proud of building an app that will greatly enhance the way that medical students can simulate diagnosis, and that we were able to develop an app that is easy to use. We are proud that we were able to simulate data and get reasonable accuracy, given our limited exposure to machine learning.
What we learned
We learnt immensely throughout the whole project. We started off with a lot of unfamiliarity due to machine learning and how to analyze large data sets, and learned immensely in using different python libraries to help us out.
What's next for MedTeach
We would like to create an API to the python and the web app, to finish the application. We would want to employ help of other algorithms to make our predictions more accurate, and develop more on the UI to make it friendly for different devices and platforms.