Related Experiment Video
Updated: Mar 1, 2026
07:53
Author Spotlight: Advancing 3D Modeling for Enhanced Diagnosis and Treatment of Pulmonary Nodules in Early-Stage Lung Cancer
Published on: October 13, 2023
2.2K
Application of the 3D slicer chest imaging platform segmentation algorithm for large lung nodule delineation
Stephen S F Yip1, Chintan Parmar1, Daniel Blezek2
1Department of Radiation Oncology, Dana-Farber Cancer Institute, Brigham and Women's Hospital, and Harvard Medical School, Boston, MA, United States of America.
Plos One
|June 9, 2017
Summary
The semiautomatic Chest Imaging Platform (CIP) segmentation algorithm offers faster and more robust lung nodule segmentation than manual methods. While it requires adjustments for most nodules, it provides an efficient starting point, potentially reducing physician workload.
Area of Science:
- Medical Imaging
- Radiology
- Computational Pathology
Background:
- Accurate lung nodule segmentation is vital for developing imaging biomarkers to predict malignancy.
- Manual segmentation is time-consuming and prone to inter-observer variability.
Purpose of the Study:
- To evaluate the robustness and accuracy of the semiautomatic 3D Slicer Chest Imaging Platform (CIP) segmentation algorithm.
- To compare its performance against manual segmentation for lung nodules.
Main Methods:
- Utilized CT images of 354 manually segmented nodules from the LIDC database.
- Assessed robustness using region of uncertainty (δ) and Dice Similarity Index (DSI).
- Compared semiautomatic CIP segmentation with manual segmentation using Wilcoxon signed-rank test and DSIAgree.
Main Results:
- CIP segmentation was significantly more robust (median δ = 14ml, DSI = 99%) than manual segmentation (median δ = 222ml, DSI = 82%).
- Median agreement between CIP and manual segmentations (DSIAgree) was 60%.
- CIP segmentation required manual adjustment for 87% of nodules, performing poorly on subtle, non-/sub-solid nodules.
Conclusions:
- Semiautomatic CIP segmentation offers computational efficiency and superior stability compared to manual segmentation.
- It can potentially reduce physician workload by providing a preliminary contour for 13% of nodules.
- Manual adjustments are often necessary, but CIP serves as a valuable starting point.
