Lung Cancer diagnosis from radiology imaging
The goal of the project is to effectively identity lung cancer from radiology imaging available with NLST lung screening data. We will use unsupervised and supervised deep learning methods to classify images. Accuracy of 95% for true positives and elimination of false positives by 95%.
Identify whether image quality is acceptable
Classify presence or absence of malignant tumor or inconclusive
Param Namboodiri, Vishal Shah