A deep learning framework for detecting lesions in CT scans from Deep Lesion dataset as a part of the Machine Learning: Deep Learning course offered at Johns Hopkins University, Spring 2019.

To convert Pascal VOC annotations to COCO dataset format, run generate_xml_list.py to generate a text file containing a list of all the .xml files in the annotation folder run voc2coco.py to generate a single .json file

Our pytorch implementations of PANet and EncoderNet are present in panet.py .

The final report is present in final_report.pdf

We are grateful to Dr Mathias Unberath (course instructor), Jie Ying Wu, Gao Cong (Course TAs), Intuitive Surgical and Google Cloud (Course Sponsor) for making it possible to complete this project.