Using Deep learning to diagnose Pneumonia and CoronaVirus.

This study proposes a convolutional neural network model trained from scratch to classify and detect the presence of pneumonia from a collection of chest X-ray image samples. Unlike other methods that rely solely on transfer learning approaches or traditional handcrafted techniques to achieve a remarkable classification performance, we constructed a convolutional neural network model from scratch to extract features from a given chest X-ray image and classify it to determine if a person is infected with pneumonia. This model could help mitigate the reliability and interpretability challenges often faced when dealing with medical imagery.

An interesting thing about The model, is it has predicted "pneumonia" on X-ray images with patients suffering from CoronaVirus(COVID-19). This means although the patient does not have pneumonia, he/she is likely to be suffering from another respiratory illness like corona in our case. This is unproven, but can be very useful especially during the current pandemic where scientists aren't sure of CoronaVirus Symptoms! Check out the following story where a Chest X-ray helped diagnose a Vietnamese man with CoronaVirus. To try out the application, click on the try it out button, click on POST, then click try it now. Upload a chest x-ray and execute! The results is in terms of log loss, so multiply by 100 to get probability.

More information

Developer's say

We developed a model to detect and classify pneumonia from chest X-ray images taken from frontal views at high validation accuracy. The algorithm begins by transforming chest X-ray images into sizes smaller than the original. The next step involves the identication and classifcation of images by the convolutional neural network framework (keras), which extracts features from the images and classifes them

One of the biggest challenges of using AI for disease diagnosis is the lack of labeled data. Data is not readily available and only medical professionals are qualified to label the data. Furthermore, data collection is often biased to specific demographics and the hospital equipment used to collect the data. You can download x-ray images at for testing the application.

Check me out on Github!