Identify Pneumothorax Disease using UNet

Pneumothorax can be caused by a blunt chest injury, damage from underlying lung disease, or most horrifying—it may occur for no obvious reason at all. On some occasions, a collapsed lung can be a life-threatening event. Pneumothorax is usually diagnosed by a radiologist on a chest x-ray, and can sometimes be very difficult to confirm. An accurate AI algorithm to detect pneumothorax would be useful in a lot of clinical scenarios. AI could be used to triage chest radiographs for priority interpretation, or to provide a more confident diagnosis for non-radiologists.

In this repository a model is developed using UNet with the help of Convolution Neural Networks that takes a chest x-ray image as input and predicts whether the given image has a pneumothorax or not. The SIIM_ACRDataset class in data.py performs preprocessing and augmentation on the images and the masks from the dataset before feeding it to the model for training. The dataset is available at https://www.kaggle.com/vbookshelf/pneumothorax-chest-xray-images-and-masks.

Criterion for using this repository:

• This project was tested on Python 3.7 using Conda Distribution
• Choose the num_workers for parallelizing the training and validation dataset using pytorch based on your system specification.
• If your cuda memory gets full, try resizing the images using img_resize argument and also try decreasing the batch size for training.

Requirements

Install the following packages:

• NVIDIA Apex for mixed precision training. If you have a single GPU then normal loss.backward() should suffice, therefore you can avoid this installation.

git clone https://github.com/NVIDIA/apex
pip install -v --no-cache-dir --global-option="--cpp_ext" --global-option="--cuda_ext" ./apex

• Albumentation library for image augmentation

pip install albumentations

References

• https://github.com/abhishekkrthakur/approachingalmost

• https://www.kaggle.com/vbookshelf/pneumothorax-chest-xray-images-and-masks