# Evaluation of Specific Contour Propagation Tool Accuracy for Lung Tumor Lesions

**Authors:** Hideaki Matsukawa, Eiji Shiba, Ryohei Kuroki, Yoshio Iwai, Takayuki Ohguri, Takatoshi Aoki

PMC · DOI: 10.7759/cureus.81913 · Cureus · 2025-04-08

## TL;DR

This study evaluates the accuracy of an automatic contouring tool for lung tumors and nearby organs in radiation therapy planning.

## Contribution

The study provides a clinical evaluation of Adapt Anatomy's performance for lung cancer treatment planning using CBCT and rCT images.

## Key findings

- Adapt Anatomy showed high accuracy for most structures, with DSC and OI values above 0.8.
- Heart and esophagus contours had lower accuracy compared to manual contours.
- CBCT and rCT image results were comparable in terms of DSC and OI.

## Abstract

Introduction: Adapt Anatomy is an automatic propagation tool installed in the Monaco® treatment planning system v. 5.5.1 (Elekta AB, Stockholm, Sweden). Few studies have reported the usefulness of Adapt Anatomy for adaptive radiation therapy (ART) in clinical cases. This study aimed to investigate the accuracy of Adapt Anatomy for lung tumor lesions.

Methods: Ten patients with lung cancer were treated at our hospital, and treatment plans were formulated again during the radiotherapy course. Each patient’s gross tumor volume (GTV), lung, cord, heart, and esophagus were contoured manually by radiation oncologists and automatically by Adapt Anatomy on replanning computed tomography (rCT) images. The structures were also contoured on cone-beam computed tomography (CBCT) images obtained on the same days as the rCT images, using Adapt Anatomy. The structural volumes contoured using Adapt Anatomy were compared with those contoured by the radiation oncologists. The dice similarity coefficient (DSC) and overlap index (OI) were calculated for each structure. Additionally, the DSC and OI calculated for the GTV on the CBCT images were compared with the values calculated from the rCT images.

Results: The automatic contour volumes for all of the structures did not differ significantly from the manual contour volumes, except for heart and esophagus volumes. The DSC and OI values for all of the structures were >0.8, except for the esophagus. The DSC and OI values for the esophagus were 0.73±0.070 and 0.68±0.10, respectively. The DSC and OI values calculated for the CBCT images did not differ significantly from those calculated for the rCT ones (p=0.208, 0.401, respectively).

Conclusion: Adapt Anatomy showed sufficient accuracy compared to the other software solutions, and we concluded that there is a possibility that Adapt Anatomy may be useful for performing ART on CBCT images. However, careful reviews by radiation oncologists are warranted in clinical cases because its accuracy varies significantly in certain structures such as the esophagus.

## Linked entities

- **Diseases:** lung cancer (MONDO:0005138)

## Full-text entities

- **Diseases:** tumor (MESH:D009369), Lung Tumor Lesions (MESH:D008175)
- **Species:** Homo sapiens (human, species) [taxon 9606]

## Full text

_Full body text omitted from this summary view._ Fetch the complete paper as Markdown: https://tomesphere.com/paper/PMC12061514/full.md

## Figures

3 figures with captions in the complete paper: https://tomesphere.com/paper/PMC12061514/full.md

## References

26 references — full list in the complete paper: https://tomesphere.com/paper/PMC12061514/full.md

---
Source: https://tomesphere.com/paper/PMC12061514