马上注册,阅读更多内容,享用更多功能!
您需要 登录 才可以下载或查看,没有账号?立即注册
×
A novel pulmonary nodule localization method based on characteristic curves of lung segment boundaries compared with puncture localization, amulti-center randomized study
Objective: We found that injecting indocyanine green(ICG) after blocking the arteries or veins of lung segments will result in personalized boundaries. If it is possible to obtain the adjacent relationship between lung segment boundaries and pulmonary nodules before surgery, we can identify the position of the nodule during surgery, thereby achieving accurate localization.
Methods: We designed a lung segmentation model(WTS) in 3D reconstruction based on vessels using Convolutional Neural Network and Reinforcement Learning. The modeling group included 3,862 cases preoperative CT-dicom and corresponding surgical video recording in 2017-2025. For validation, We have initiated a multicenter Randomized controlled study (NCT06404164) to compare the accuracy of three-dimensional reconstruction of segmental boundaries versus percutaneous pulmonary needle localization for guiding wedge resection of pulmonary nodules, with the primary endpoint being the one-time R0 resection rate.
Results: The study included a total of 549 patients form 8 sites. The following populations were excluded: patients with solid pulmonary nodules, >2cm or showing pleural changes. Among the included population, in the WTS group and the Puncture group, nodules located in the outer one-third of the lung field accounted for 93.2% and 91.7%, respectively, while those in the middle one-third accounted for 6.8% and 8.3%.
The R0 resection accuracy rates for the WTS group and the Puncture were 95.2% (261/274 cases) and 93.6% (257/275 cases), respectively (non-inferiority P=0.007, superiority test P=0.23). In terms of median procedure time, the WTS group had a median time of 24 minutes, while the Puncture group had a median time of 28 minutes.
Regarding procedural complications, the average postoperative chest tube indwelling time was 1.3 days for the WTS group and 1.2 days for the Puncture group. The rates of grade 3 or higher complications were 0.8% and 0.5%, respectively.
Conclusions: The positioning efficiency of WTS based on the boundary of lung segments and the adjacent relationship of lung nodules is not inferior to that of puncture positioning, which can relieve patients of psychological pressure and pain during percutaneous lung puncture, and also avoid the occurrence of pneumothorax after positioning.
Zihao Chen (1), Wenzhao Zhong (2), (1) Guangdong Provincial People's Hospital, Guangdong, (2) Guangdong Lung Cancer Institute, Guangdong Provincial People's Hospital, Guangzhou, NA
|