Developing a deep learning model for automatical detection of Covid-19 phneumonia from frontal chest X-Ray/CT scans. The code was written on Google Colab, so everyone can use the free GPU resources available online to improve the model.
Example Covid-19 and healthy patient chest X-Rays used for training (Note that the healthy examples are pediatric patients)
Loss curves
Confusion Matrix (accuracy 98%, sensitiviy 93% and specificity 100%)
Correct Healthy Patient Detection (True Negatives): 13\ Incorrect Covid-19 Detection (False Positives): 1\ Incorrect Healthy Patient Detection (False Negatives): 0\ Correct Covid-19 Detection (True Positives): 28\ Total Patietns Diagnosed with Covid-19: 28