Detection and Attention: Diagnosing Pulmonary Lung Cancer from CT by Imitating Physicians

TL;DR

This paper introduces a new CT-based lung nodule detection system that mimics radiologist reading, using sequence analysis techniques to improve accuracy and speed, achieving state-of-the-art results on the LUNA16 dataset.

Contribution

The paper presents two novel methods, MSP and MLGS, for analyzing CT scan sequences to enhance lung nodule detection accuracy and reduce false positives and negatives.

Findings

Achieved state-of-the-art FROC score of 0.892 on LUNA16 dataset.

Detection process completes within 20 seconds per patient.

Outperforms existing methods in accuracy and speed.

Abstract

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.