COVID detection using AI

Inspiration

Due the the increasing rise in covid cases and the technological barrier we decided to make a model to make the common man understand the signs of covid using a ct scan by passing it through our model.

What it does

Our problem statement is detection of different kind of lung diseases using retina net module and tell the accuracy of classified covid patients Since different kind of models already exists, we decided to collectively take data base from various resources and compiled for our program We are using Ct scan images of size of 512 to 512 px and 16-bit gray scale resolution we pass the data through the first algorithm which discards few images from the data set which are of not useful for us After that we r using featured pyramid network to classify images based on infection points At the end of network classification, we pass it through another network for precision and the resultant data produces up to 91 % of accuracy of predicting a covid patient

How we built it

We use retina net to interpret the data and give the amount of correct classified covid patients data We are using Ct scan images of size of 512 to 512 pixels and 16-bit gray scale resolution

Challenges we ran into

We weren't able to input few modules while running the code which included keras.application.resnet50

Accomplishments that we're proud of

What we learned

we learnt a lot of data set extraction and few models which led to the creation of this project.

What's next for COVID detection using AI

we try to expand the domain by including various other diseases such as pneumonia etc. and hope to make ease for a common man to understand the symptoms and its related disease.