当前位置： SCI文献检索 > PHYSICS IN MEDICINE AND BIOLOGY期刊下所有文献 > A novel geometry-dosimetry label fusion method in multi-atlas segmentation for radiotherapy: a proof-of-concept study.

A novel geometry-dosimetry label fusion method in multi-atlas segmentation for radiotherapy: a proof-of-concept study.

Abstract：

:Multi-atlas segmentation (MAS) has been widely used to automate the delineation of organs at risk (OARs) for radiotherapy. Label fusion is a crucial step in MAS to cope with the segmentation variabilities among multiple atlases. However, most existing label fusion methods do not consider the potential dosimetric impact of the segmentation result. In this proof-of-concept study, we propose a novel geometry-dosimetry label fusion method for MAS-based OAR auto-contouring, which evaluates the segmentation performance in terms of both geometric accuracy and the dosimetric impact of the segmentation accuracy on the resulting treatment plan. Differently from the original selective and iterative method for performance level estimation (SIMPLE), we evaluated and rejected the atlases based on both Dice similarity coefficient and the predicted error of the dosimetric endpoints. The dosimetric error was predicted using our previously developed geometry-dosimetry model. We tested our method in MAS-based rectum auto-contouring on 20 prostate cancer patients. The accuracy in the rectum sub-volume close to the planning tumor volume (PTV), which was found to be a dosimetric sensitive region of the rectum, was greatly improved. The mean absolute distance between the obtained contour and the physician-drawn contour in the rectum sub-volume 2 mm away from PTV was reduced from 3.96 mm to 3.36 mm on average for the 20 patients, with the maximum decrease found to be from 9.22 mm to 3.75 mm. We also compared the dosimetric endpoints predicted for the obtained contours with those predicted for the physician-drawn contours. Our method led to smaller dosimetric endpoint errors than the SIMPLE method in 15 patients, comparable errors in 2 patients, and slightly larger errors in 3 patients. These results indicated the efficacy of our method in terms of considering both geometric accuracy and dosimetric impact during label fusion. Our algorithm can be applied to different tumor sites and radiation treatments, given a specifically trained geometry-dosimetry model.

journal_name

Phys Med Biol

journal_title

Physics in medicine and biology

authors

Chang J,Tian Z,Lu W,Gu X,Chen M,Jiang SB

doi

10.1088/1361-6560/aa5ed9

subject

Has Abstract

pub_date

2017-05-07 00:00:00

pages

3656-3667

issue

eissn

0031-9155

issn

1361-6560

journal_volume

pub_type

杂志文章

在线工具

A novel geometry-dosimetry label fusion method in multi-atlas segmentation for radiotherapy: a proof-of-concept study.