Tumor Detection using Airways Asymmetry

A novel tumor detection technique on CT scan images of the neck area is detailed in this paper. This technique is based on an airways´ symmetry evolution within slices. The algorithm proposes to the physician a set of three slices where a tumor (if it exists) should mostly be located. Then, he will just have to browse the three slices instead of almost 100 in a CT scan. Our method is very effective and shows no false alarms within the patients in our database. In each of our tests the tumors were found to be close to one of the three proposed slices