DeepViewer软件自动勾画乳腺癌保乳术后患者心脏结构的应用

Application of DeepViewer in Automatic Delineation of Heart Structure for Patients with Breast Cancer After Breast-Conserving Surgery

目的评估DeepViewer软件自动勾画乳腺癌保乳术后患者心脏结构的可行性与准确性。方法选取20例乳腺癌保乳术后患者进行研究,其中左侧乳腺癌10例,右侧乳腺癌10例。通过比较DeepViewer软件自动勾画和手工勾画的体积差异、位置差异、形状一致性和勾画时间,同时对左侧和右侧乳腺癌的数据进行比较研究,评估DeepViewer软件自动勾画心脏结构的可行性和准确性。结果自动和手工勾画两种方式心脏结构的百分体积差异(ΔV%)为-21.43%±5.52%,总位置差异(ΔL)为(0.37±0.19)cm,相似性指数为0.84±0.03,自动勾画比手工勾画节约了大约84%的勾画时间。左侧乳腺癌和右侧乳腺癌患者的DSC值分别为0.83±0.03和0.85±0.02,两组差异无统计学意义(P>0.05)。结论使用DeepViewer软件自动勾画乳腺癌保乳术后患者的心脏结构能够达到满意的勾画结果,并可缩短心脏结构的勾画时间,提高放疗工作效率。

Objective To evaluate the feasibility and accuracy of DeepViewer software in the automatic delineation of heart structure for patients with breast cancer after breast-conserving surgery.Methods Twenty patients with breast cancer after breast-conserving surgery were enrolled in this study,including 10 patients with left breast cancer and 10 with right breast caner.In order to evaluate the feasibility and accuracy of DeepViewer for automatic delineation of heart structure,differences in volume and position,dice similarity coefficient(DSC),and delineation time were assessed between the automatic delineation and manual delineation.Meanwhile,the data of left and right breast cancer were compared.Results The volume difference(ΔV%),total position difference(ΔL)and DSC between the automatic delineation and manual delineation were-21.43%±5.52%,(0.37±0.19)cm,0.84±0.03,respectively.Automatic delineation saved about 84%of delineation time compared to manual delineation.The DSC values were 0.83±0.03 and 0.85±0.02 for patients with left and right breast cancer,respectively,with no statistical difference between the two groups(P>0.05).Conclusion Using DeepViewer software in automatic of heart structure for patients after breast cancer conserving surgery can achieve satisfactory results,shorten the delineation time,and improve the work efficiency of radiotherapy.