A segmentation tool for pulmonary nodules in lung cancer screening: Testing and clinical usage

• Published in: Physica medica 2021-10, Vol.90, p.23-29
• Main Authors: Garau, Noemi, Paganelli, Chiara, Summers, Paul, Bassis, Danny, Lanza, Carolina, Minotti, Marta, De Fiori, Elvio, Baroni, Guido, Rampinelli, Cristiano
• Format: Article
• Language: English
• Subjects: Automatic segmentation; Clinical validation; Deep learning; Lung cancer screening
• ISSN: 1120-1797; 1724-191X
• DOI: 10.1016/j.ejmp.2021.08.011

•Optimization of labelled data collection for lung cancer screening research.•Integration of 2D-Unet into GUI for segmentation in a clinical context.•Good generalizability of the net in two training-independent datasets.•Limited inter-observer variability in terms of volume and volume doubling time. With the future goal of defining a large dataset based on low-dose CT with labelled pulmonary lesions for lung cancer screening (LCS) research, the aim of this work is to propose and evaluate into a clinical context a tool for semi-automatic segmentation able to facilitate the process of labels collection from a LCS study (COSMOS, Continuous Observation of SMOking Subjects). Considering a preliminary set of manual annotations, a segmentation model based on a 2D-Unet was trained from scratch. Contour quality of the final 2D-Unet was assessed on an internal test set of manual annotations and on a subset of the public available LIDC dataset used as external test set. The tool for semi-automatic segmentation was then designed integrating the tested model into a Graphical User Interface. According to the opinion of two clinical users, the percentage of lesions properly contoured through the tool was quantified (Acceptance Rate, AR). The variability between segmentations derived by the two readers was estimated as mean percentage of difference (MPD) between the two sets of volumes and comparing the likelihood of malignancy derived from Volume Doubling Time (VDT). Performance in test sets were found similar (DICE ~ 0.75(0.15)). Accordingly, a good mean AR (80.1%) resulted from the two readers. Variability in terms of MPD was equal to 23.6% while 2.7% was the VDTs percentage of disagreement. A semi-automatic segmentation tool was developed and its applicability evaluated into a clinical context demonstrating the efficacy of the tool in facilitating the collection of labelled data.