Application of machine vision in fractionated radiotherapy
10.3760/cma.j.cn112271-20230627-00212
- VernacularTitle:机器视觉在分次放射治疗中的应用研究
- Author:
Xiaolin CHEN
1
;
Yangchao XIE
;
Xingfu LIN
;
Fenpen HUANG
;
Tingying CHEN
;
Wanquan CHEN
;
Shaofeng WANG
Author Information
1. 厦门医学院附属第二医院肿瘤放疗科，厦门 361021
- Keywords:
Machine vision;
Radiotherapy;
Real-time monitoring;
Set-up errors
- From:
Chinese Journal of Radiological Medicine and Protection
2024;44(3):202-206
- CountryChina
- Language:Chinese
-
Abstract:
Objective:To monitor intra-fractional set-up errors in tumor radiotherapy using a real-time intelligent capture system for precision displacement.Methods:A simulated radiotherapy environment was created in both the laboratory and the treatment room. A three-axis ( xyz) displacement platform (LD60-LM) and dial gauges were used as displacement measurement tools. Moreover, a real-time intelligent capture system for precision displacement was developed for displacement monitoring. With 23 patients treated with radiotherapy enrolled in this study, the above system was employed to monitor their intra-fractional set-up errors in fractionated radiotherapy. Descriptive analyses were conducted on the deviations between the data captured by cameras and the actual displacement, obtaining the mean values and standard deviation. Results:The monitoring calibration data from the laboratory revealed displacement differences of ≤ 0.5 mm within 20 mm and a maximum displacement difference of 1.47 mm for 50 mm. In contrast, the calibration result from the treatment room exhibited deviations of ± 0.2 mm on the y- z axes, as displayed by both the left and right cameras, and ± 0.31 mm on the x- z axes, as displayed by the middle camera. During 37 radiotherapy sessions in 23 patients, the monitoring result from the middle camera revealed five deviations exceeding the threshold of 5 mm, with the maximum deviation duration and displacement of 57.2 s and 9.24 mm, respectively. Conclusions:The real-time intelligent capture system for precision displacement based on machine vision can achieve real-time monitoring of set-up errors during tumor radiotherapy. Nevertheless, further improvements and service testing are necessary for this system.
