Algorithmic Challenge: Lung Cancer

The Lung Cancer Challenge was a two part challenge run by TopCoder in collaboration with the Harvard Catalyst Translational Innovator program and the Laboratory for Innovation Science (LISH) at Harvard University to create and test automatic delineation algorithms that can improve treatment of cancerous lung tumors.

Successful treatment depends heavily on the radiation oncologist’s ability to make accurate measurements of the tumor’s shape and its responsiveness to interventions. Manual delineation of tumors is very time consuming when performed by highly trained experts and is prone to inconsistency and bias. Automatic delineation also has issues because it depends heavily on the training data sets and tends to make errors, which would have been easily detected by experts.

The lung cancer algorithmic challenge tasked competitors with producing an automatic tumor delineation algorithm that parallels the lung tumor delineation accuracy of an average expert, while exceeding the expert in processing speed and delineation consistency. The challenge ran in two stages. The first targeted the accuracy of the algorithm, which had to reproduce the delineation by the expert as closely as possible. The second targeted credibility of the algorithm by using expert feedback to further train the algorithm to avoid the types of errors that a trained human would never make.