Date: May 2024
Developed a series of Convolutional Neural Network (CNN) based models, including YOLOv5, DETR, and VNet, for the purpose of detecting complex and irregular lung nodules in computer tomography (CT) scans. Implemented a shape modeling algorithm and integrated it with the most effective model among the three, YOLOv5, to address the irregular and complex shapes commonly found in lung nodules. Successfully achieved an impressive accuracy rate of over 70% using the YOLOv5 model.