The Study Of Uncertainty In Grass Tumor Volume Delineation Of Lung Cancer And Its Influence To Normal Tissue Complication Probability

Part I Subjective Variability in Gross Tumor Volume Delineation of Lung Cancer and Its Influence to Normal Tissue Complication ProbabilityObjective: To compare the differences of the gross tumor volumes(GTV) of lung cancer defined by radiologists and radiation oncologists and to evaluate the influence to normal tissue complication probability (NTCP) of heart.Methods: The CT images of ten patients with lung cancer eligible for 3D-CRT were reviewed. Three radiologists and three radiation oncologists were invited to delineate the GTV. The value and variability of GTV volume were calculated, and the differences from different physicians were compared. Then the three dimensional conformal radiotherapy plans based on the GTVs were optimized. The time-dose-fractionation was 60Gy in 2 Gy perfraction,5 times a week. NTCPs of heart were calculated by binomial model.Results: Significant differences between physicians were observed (p<0.05). Compared to radiation oncologists, radiologists tended to delineate smaller volumes. Though the differences of NTCPs of heart were not statistically significant ,the differences of NTCPs in some individual patients were pretty big. Conclusion: Major discordances were observed between the radiation oncologists'and the radiologists'delineations, which lead to the differences in NTCPs and these differences are not negligible. A better training of radiation oncologists in thoracic imaging and collaboration between radiation oncologists and radiologists should decrease this variability.Part II Influence of CT Window Techniques in Gross Tumor Volume Delineation of Lung CancerObjective:To find out the influence by different window wide(WW) and window level(WL) in target volume delineation of lung cancer.Methods: (1) Ten lung cancer cases for 3-D conformal radiation therapy were selected for simulation CT scanning, then GTVs were delineated with twelve different WW and WL condition while the CT scan imaging were transferred to treatment planning system(TPS). The average and variability of GTV volume values were calculated. (2)To scan the wax ball under the same conditions which was made to simulate tumour in lung, then draw the outline of ball with different WW and WL used before in TPS system. Calculate testing volumes and evaluate whether they were equal to real ball volume.Results: (1)A high variability of GTV volume in different WW and WL was found, the ratio of the largest to the smallest volume was more then 2 in five patients. About the change range of GTV volume compared with its mean value in each patient, there were 28%volume values larger than±20%.(2)The testing volumes of wax ball change with WW and WL. The ratio of the largest to the smallest volume was 1.35.Conclusions: The influences in target volume delineation caused by CT window technique are not negligible, so optimal WW and WL is significant for accurate target volume delineation.