Covid19 Xray Image classification using resnet50 CNN model on 2days of data

I will be using 2days of X-ray data derived from over a throusand patients, to classify the covid19 disease. The first X-ray image will be taken at the admission time and the next after the admission. The motivation of this problem is to determine whether these the days of data will provide an accurate classification. Once the model provides a classification, it will be clear as to whether a patient needs to be sent to the ICU or not.

The code

There are 4 files.

First the "dataset_man.py" is used to manipulate the dataset using the metadata.csv, to obtain & sperate out the positive and negative covid Xray images.

Second, "dicom_decompressor.py" is used to decompress the dicom images to .png image.

Third, "2daysdata.py" is used to obtain only the first two day's of images, from all the patient's data.

Fourth, finally "2daysmodel.py" contains the resnet50 CNN model, that takes the 2daysdata.py's output as input, and runs the model.

HealthInfoLab/COVID19-Pari

About

Resources

Stars

Watchers

Forks

Releases

Languages