Lung Cancer Detection via Deep Learning
YOLOv3 applied to non-small cell lung cancer nodule detection in CT scans, trained on NIH Imaging Data Commons data to reduce diagnostic error and speed up radiologist workflows. Co-led with Robyn An.
Early and accurate detection of non-small cell lung cancer (NSCLC) in CT scans significantly improves patient outcomes, but manual reading is time-intensive and subject to variability between radiologists. This project applies YOLOv3 to automate nodule detection directly from CT imagery.
CT images were sourced from the National Institutes of Health Imaging Data Commons (IDC) and manually annotated using labelImg. The model was trained on Google Colab using GPU resources. The goal was to reduce diagnostic error and accelerate the detection pipeline for clinical use.
Project co-led with Robyn An.