eduzhai > Applied Sciences > Engineering >

PneumoXttention A CNN compensating for Human Fallibility when Detecting Pneumonia through CXR images with Attention

  • king
  • (0) Download
  • 20210506
  • Save

... pages left unread,continue reading

Document pages: 10 pages

Abstract: Automatic Chest Radiograph X-ray (CXR) interpretation by machines is animportant research topic of Artificial Intelligence. As part of my journeythrough the California Science Fair, I have developed an algorithm that candetect pneumonia from a CXR image to compensate for human fallibility. Myalgorithm, PneumoXttention, is an ensemble of two 13 layer convolutional neuralnetwork trained on the RSNA dataset, a dataset provided by the RadiologicalSociety of North America, containing 26,684 frontal X-ray images split into thecategories of pneumonia and no pneumonia. The dataset was annotated by manyprofessional radiologists in North America. It achieved an impressive F1 score,0.82, on the test set (20 random split of RSNA dataset) and completelycompensated Human Radiologists on a random set of 25 test images drawn fromRSNA and NIH. I don t have a direct comparison but Stanford s Chexnet has a F1score of 0.435 on the NIH dataset for category Pneumonia.

Please select stars to rate!


0 comments Sign in to leave a comment.

    Data loading, please wait...