SkinDiseaseDetector

• 

  Home Page

• 

  Uploaded Image

Inspiration

The inspiration of our project was that we saw a lack of AI in the medical field online and we had a teammate who was in that field. We found that almost everyone have experienced some type of scare or worry from a random rash appearing on us. This made it our goal to create a tool that could empower individuals to assess their skin health with just a singular picture, promoting early diagnosis and keep everyone live a longer and happier life.

What it does

What the project does is that it allows users to login in and make an account where they can upload a photo of their skin. Then it would detect what was on your skin and determine if you had any skin disease. Then that photo and the results would be stored into a database for you to look at whenever you want. After uploading the image, we used gen ai to generate treatments for which disease our model has deemed most likely.

How we built it

We built this project using Next.js, Tailwind CSS, Flask, Tensorflow, Docker and MongoDB We used Next.js and Tailwind CSS for the frontend because we had previous experience and we found it most easiest to develop with, then we used Flask because for our backend. We used TensorFlow to create our model to detect the skin disease. We used Docker to containerize the application allowing for quick development and we used MongoDB to store the user's uploaded images and store their results

Challenges we ran into

Some challenges we ran into was that our model did not have good accuracy and there was a problem with sessions where it would kick us out when we logged in. There was also a problem with connecting everything together with our docker file to run everything simultaneously. Another problem we ran into was that we never used OpenAI's API and it was really confusing for us and how to use it.

Accomplishments that we're proud of

We were able to make it kind of work despite being very new to programming. Most of our members were creating a website for the first time and learning how it functions. A huge accomplishment that we're proud of is that we can see the progress we've made and how much we've learned.

What we learned

What's next for SkinDiseaseDetector

What's next for SkinDiseaseDetector is that we could make an app for it as it's easier to upload a photo using your phone camera. We could partner up with dermatologists or clinics to retrieve data that would allow our model to train allowing for more accuracy and precision. We can also enhance our Gen AI to allow for more specific and tailored treatments and could have conversations with the user to help them find out what's happening on their skin.

Built With

• flask
• mongodb
• next.js
• openai
• tailwind
• tensorflow