2024 World Conference on Lung Cancer (WCLC) - Posters-P2.05B.04 Development of a CT Image-Based Virtual Atelectasis Simulation Model and Noninvasive Lung Nodule Localization System

P2.05B.04 Development of a CT Image-Based Virtual Atelectasis Simulation Model and Noninvasive Lung Nodule Localization System

Back to course

Pdf Summary

The study detailed in this document focuses on the development of a non-invasive lung nodule localization system, specifically utilizing a CT image-based atelectasis simulation to aid in surgical procedures. Conducted by a team from Korea University, the research involves converting 2D CT images into 3D models using the Interactive Lobe Segmentation feature in the 3D Slicer Chest Imaging Platform.<br /><br />The process includes rotating the lung model into a surgical posture, creating a concave thoracic cavity shape, and utilizing a virtual descent simulation with mesh points. A critical part of this simulation is the Virtual Atelectasis Lung Model (VALM), which is enhanced with the use of the Intel Realsense T265 Sensor to provide real-time 3D localization of nodules during procedures such as Video-Assisted Thoracic Surgery (VATS).<br /><br />The effectiveness of this simulation is evaluated using various image comparison metrics, specifically the Dice Similarity Coefficient (DSC), Jaccard Similarity Coefficient (JSC), and Hausdorff Distance (HU), which measure the agreement between simulated and real conditions. Results showed a high level of agreement between VALM and VATS images, with an average nodule concurrency of 6.39mm, suggesting that this non-invasive system can enhance precision in resection margins during surgery.<br /><br />This innovative approach is fully supported by the National Research Foundation of Korea and was ethically approved by the Korea University Institutional Review Board. The research underscores a significant advance in thoracic surgery, offering greater accuracy in identifying and localizing lung nodules while potentially reducing invasiveness in procedures.

Asset Subtitle

Jinwook Hwang

Meta Tag

Speaker Jinwook Hwang

Topic Pulmonology & Staging

Keywords

non-invasive lung nodule localization

CT image-based atelectasis simulation

3D Slicer Chest Imaging Platform

Interactive Lobe Segmentation

Virtual Atelectasis Lung Model

Intel Realsense T265 Sensor

Video-Assisted Thoracic Surgery

Dice Similarity Coefficient

Jaccard Similarity Coefficient

Hausdorff Distance