Covid19 Detection from Chest-X-Ray

Coronavirus disease 2019 (COVID-19) is a highly infectious disease caused by severe acute respiratory syndrome coronavirus2 (SARS-CoV-2)

Currently Reverse transcription polymerase chain reaction (RT-PCR) is used for diagnosis of the COVID-19. X-ray machines are widely available and provide images for diagnosis quickly so chest X-ray images can be very useful in early diagnosis of COVID-19.

Large scale implementation of the COVID-19 tests which are extremely expensive cannot be afforded by many of the developing & underdeveloped countries hence if we can have some parallel testing procedures using Artificial Intelligence & Machine Learning, it will be extremely helpful.

Here, We have used Transfer Learning approach to build a Deep Learning based model to speed up the process of testing using chest x-ray. We have used the concept of explainable AI named as GradCAM to support our findings.

1. Densnet
2. EfficientNetB7
3. VGG16

We have used Chest X-ray (Covid-19 & Pneumonia) dataset from kaggle. Dataset contains Chest X-ray images of COVID-19, Pneumonia, & Normal patients.

Transfer Learning

It is a method that allows us to use knowledge gained from other tasks in order tackle new but similar problems quickly and effectively. It is used by loading a generic and well trained image classification network for feature extraction, and then adding few layers (head) to be trained for the target task.

Data Description & Visualization

We can observe that images are of different sizes.

Data preprocessing

Images are scaled to a size of 244 by 244, Normalized to values (0,1) & augmented by simple zoom and rotation to enhance the generalization.

VGG16 loss and accuracy curves

Output of VGG16