Automated detection of pulmonary embolism from CT-angiograms using deep learning

We show that a deep learning model trained with a relatively small, weakly annotated dataset can achieve excellent performance results in detecting PE from CTPAs.