Automated 3d lung nodule detection and segmentation from ct scans
NoduleFlow: Automatic 3D Lung Nodule Detection and Segmentation
Computer Vision and Pattern Recognition
Summary
Finding lung nodules in CT scans is important for lung cancer screening but can be slow and inconsistent when done by humans. The authors propose NoduleFlow, a fully automatic system that detects, classifies, and outlines lung nodules in 3D CT images without needing manual help or prior information about where nodules might be. Their method also tracks potential nodules across multiple scans over time to improve detection. They tested NoduleFlow on several different datasets, showing it can work well in diverse screening and diagnostic settings.
What this means in practice
- •For radiology teams: Automatically detect and outline lung nodules in CT scans to support faster and more consistent cancer screening workflows.
- •For medical imaging software developers: Build integrated lung nodule analysis tools that automatically process CT scan series without manual input or lesion annotations.
Authors
Giulia Romoli, Filippo Ruffini, Francesco Di Feola, Valerio Guarrasi, Matteo Tortora, Paolo Soda
Abstract
Low dose Computed Tomography (CT) is the standard imaging modality for lung cancer screening. On CT scans, lung cancer typically manifests as pulmonary nodules, whose manual identification is time-consuming and prone to observer variability, image quality limitations, and suboptimal reading conditions. Moreover, subtle temporal changes are often difficult to quantify reliably in CT scans from the same patient across follow-up examinations. Deep learning-based approaches for the automatic detection and segmentation of lung nodules have demonstrated improved accuracy and efficiency. However, most existing approaches provide only coarse region-of-interest localisation, rely on prior lesion annotations or human supervision, or treat detection and segmentation as separate tasks. In this work, we present NoduleFlow, a framework for lung nodule detection, classification, and segmentation directly from series of CT volumes. Our proposal consists of four interconnected modules and operates in an automated manner at inference time, without requiring prior lesion annotations, manual initialisation, or region-of-interest pre-selection. We further investigate whether nodule localization benefits from the propagation of candidate regions across follow-up scans through inter-scan registration fields, as opposed to conventional independent per-scan analysis. The generalisability of the framework is systematically evaluated across four independent datasets. The results underscore the potential of automated pipelines for lung nodule analysis in screening and diagnostic settings.