Detection and Attention: Diagnosing Pulmonary Lung Cancer from CT by Imitating Physicians
This addresses the problem of efficient and accurate lung cancer diagnosis for radiologists, though it is incremental with novel methods for known bottlenecks.
The paper tackled lung nodule detection from CT scans by treating it as an object detection on video problem, achieving a state-of-the-art FROC score of 0.892 on the LUNA16 dataset with detection speeds under 20 seconds per patient.
This paper proposes a novel and efficient method to build a Computer-Aided Diagnoses (CAD) system for lung nodule detection based on Computed Tomography (CT). This task was treated as an Object Detection on Video (VID) problem by imitating how a radiologist reads CT scans. A lung nodule detector was trained to automatically learn nodule features from still images to detect lung nodule candidates with both high recall and accuracy. Unlike previous work which used 3-dimensional information around the nodule to reduce false positives, we propose two simple but efficient methods, Multi-slice propagation (MSP) and Motionless-guide suppression (MLGS), which analyze sequence information of CT scans to reduce false negatives and suppress false positives. We evaluated our method in open-source LUNA16 dataset which contains 888 CT scans, and obtained state-of-the-art result (Free-Response Receiver Operating Characteristic score of 0.892) with detection speed (end to end within 20 seconds per patient on a single NVidia GTX 1080) much higher than existing methods.