Objective To study the application of correlation statistical analysis model in IMRT planning optimization. Methods A correlation statistical analysis model was established. IMRT plans of 5 prostate cancer patients were randomly chosen from the ADAC Pinacal 7.6 version planning system. The beam parameters, objective functions of the target area and optimization parameters of the primary plan were kept unchanged. The main optimization parameters including EUD parameter a,weight (w) and EUD_(max) of OAR were adjusted in optimization procedure successively. The correlation analysis (CF) and optimization efficiency analysis (OF) were carried out on the results of the optimization. The optimal value of parameters (MORt{}) with the best dose distribution was obtained and substituted in the corresponding primary plan.After optimization, the dose distributions of the two IMRT plans were compared. Results There were signif-icant differences with different optimization methods. The EUD parameter a and weight factor (w) almost had no effect on volume dose of OAR (OF < 0.01), though had obvious effect on the maximum dose and mean dose (OF≈1). The CF analysis showed that the correlation between the PTV ( V_(95)) volume and OAR dose was different when the EUD parameter a and w were applied for optimization. Meanwhile, the difference was proportional to the distance between the target area and OAR. The mean dose of OAR was decreased and the mean dose of PTV was more close to the prescription dose when the optimization parameters of the prima-ry plan were substituted with MOR{a} ,MOR{w} and MORt{EUD_(max)}. Besides, the DVH and isedose dis-tribution of the secondary plan were better. Conclusions The correlation statistical analysis model can be used to accurately determine the scope of optimization parameters in the IMRT planning procedure in prostate cancer, and an IMRT plan which fulfils the clinical requirement can be obtained efficiently.

Objective To analyze the biophysical dosimetric characteristics and clinical application ability of VMAT technology for breast cancer post-mastectomy.Methods 28 patients with breast cancer (10 at left side and the other at right side) were planned in different ways respectively.One was two 90 degree arc VMAT plan and the other were 5 beam IMRT plan.The dosimetric parameters of two different plans including tumor control probability (TCP),conformity index(CI),homogeneity index (HI),V95and V110 in target,normal tissue complication probability (NTCP),V5,V20,V30 for ipsilateral lung,NCTP,D V25 for heart,D for the contralateral breast in OARs,MU and times were compared.Results The average tumor control probability (TCP) in VMAT and IMRT group was(96 ±2)％ and (90 ±2)％ (t =-6.28,P ＜ 0.01),respectively.The PTV dose average homogeneity index (HI) of VMAT plans was better than that of IMRT plan (0.15 ±0.04 vs 0.22 ±0.02,t =13.29,P ＜0.01).For cancer position in left side,the mean dose of heart was decreased by 433.24 cGy in the VMAT plan.The NTCP of the hearts in VMAT plans had statistically significant difference compared with IMRT plans [(1.00±0.12)％ vs (1.70±0.13)％,t =2.14,P ＜0.05].For plans of right breast cancer,the average mean dose of hearts in two control group was (3.27 ± 0.26) Gy and (6.00 ± 0.47) Gy (t =9.21,P＜0.01).The total monitor unit (MU) was 530.7 in the VMAT arm and 693.9 in the IMRT arm (t =9.58,P ＜0.01).The treatment time was shorter in VMAT arm (t =8.40,P ＜0.05).Conclusions VMAT plans have better clinical value and more superior biophysical dosimetric characteristics for breast cancer post-mastectomy.

Objective To investigate the clinical effect of lung equivalent uniform dose (LEUD)-based predictive model for radiation pneumonitis (RP) induced by volumetric modulated arc therapy (VMAT) and to determine the optimal a value.Methods A total of 65 patients with primary lung cancer who received VMAT from July 2015 to February 2016 were divided into RP group and non-RP group according to the presence or absence of RP after radiotherapy.Their dose-volume histogram (DVH) data and other data were obtained and analyzed by the self-compiled numerical analysis program.The LEUD values in the two groups were calculated at a=[-50, 50], and then the a value was identified when the relative difference of LEUD between the two groups was maximal.The paired t test was used for analyzing the differences in V5, V20, V30, minimum lethal dose (MLD), and LEUD (aoptimal) between the two groups.A Pearson correlation analysis was used to determine the correlation of Vdose and LEUD (aoptimal) with RP.The logistic regression method was used to establish the predictive model of RP.Results The maximum relative difference in LEUD between RP group and non-RP group was obtained at a=0.3(627.94 cGy vs.510.23 cGy, relative difference[R]=23.07%).R decreased slowly at t=[-50,-5], increased sharply at t=[-5, 0], and reached the maximum value at a=0.3.After a rapid decrease at a=[0.3, 4], R decreased slowly at a=[4, 50].The correlation analysis of the traditional physical volume dose threshold also showed that the LEUD (at a=0.3) was correlated with V5, V10, V20, and MLD (r=0.929, P<0.05).Conclusions For patients receiving VMAT for thoracic cancer, LEUD (at a=0.3) can distinguish between patients with and without RP.Therefore, LEUD is recommended to be<510 cGy.A combination of LEUD and conventional physical dose has a good clinical predictive value for RP under non-uniform irradiation.

Objective To study the geometric difference between six different ITV generation methods from 4DCT for patients with non?small cell lung cancer ( NSCLC) treated with stereotactic ablative radiotherapy technique ( SABR) . Methods Between Dec. 2013 and Mar. 2014,16 patients were enrolled in this retrospective study. All patients underwent imaging with 4DCT scans. The MI and DI index were evaluated between six ITV generation methods:combining GTV from all 10 respiratory phases ( ITV10 );combining GTV from four respiratory phases, including two extreme phases ( 0% and 50%) plus two intermediate phases ( 20% and 70%) ( ITVYeo ) which was proposed by Seung?Gu Yeo of Soonchunhyang University;combining GTV from two extreme phases ( ITVEI+EE ) . And combining GTV from five odd phases (10%,30%,50%,70%,90%)(ITVodd).Accordingly the ITVEVEN which was combined from the remaining five even phases (20%,40%,60%,80%,0%),and ITVAVG,ITVMIP were contoured from two reconstructed 4DCT sequences,finally,a method which was not sensitive to the tumor volume and motion characteristic was selected for clinical use. Data were compared using a variance analysis followed by Student?Newman?Keulsa test both in same group or between groups. At the same time, the volume and the three dimensional movements of the tumor, the relativity of MI and DI were analyzed by Multi?parameter regression analysis. Results The mean (range) tumor motion (RLR,RAP,RCC,and R3D) are 3. 5 mm (1. 4?8. 4 mm),4. 5 mm (1. 1?8. 6 mm),9. 5 mm (0?10 mm),12. 3 mm (2. 5?55. 3 mm) respectively. The IGTVx volume are Underestimated by 25. 7%,35. 6%,17. 9%,12. 8%,3. 6%,4. 8%( P=0. 000) respectively. The MI index comparisons between six ITV generation methods and ITV10 showed statistical significance:0. 69,0. 62,0. 80, 0. 86,0. 93,0. 91 ( P=0. 006 ) . The DI index showed no statistical significance:0. 98, 0. 98, 0. 97, 0. 97, 0. 99,0. 98(P=0. 130).The tumor size and motion amplitude were certified not the independent factors for the MI index of ITVodd and ITVEVEN . Conclusions IGTVODD/EVEN based on odd or even 4DCT phases is not sensitive to the tumor size or motion characteristic and is proved to have a good marching with ITV10 meanwhile maintaining a reasonable contouring efficiency,it can be recommend to the institutions which was not equipped with the deformable registration systems.

Objective To investigate the relationship between the body mass index (BMI)/body mass index improved (BMIIMPd) and the dose of the small intestine as well as the acute radiation colitis in the intensity-modulated radiation therapy after cervical cancer surgery.Methods Thirty-nine cervical cancer patients underwent postoperative adjuvant radiotherapy.All patients received Philips large bore CT scan for enhanced CT scan,target delineation and organ at risk.All patients were treated with a single arc 10 MV VMAT plan.The correlation between the radiation dose of the small intestine and the acute radiation enteritis and BMI/BMIIMPd was analyzed.Results The BMI was calculated as (22.23±2.80) kg/m2,BMIIMPd was (21.49±3.95) kg/m2,the small intestine volume VSI was (1 155.71 ± 419.33)cc3.The volume of the small intestine received more than 10 Gy (V10_SI) VMAT was (66.50± 27.01) ％,and the equivalent uniform dose (EUD) and normal tissue complication probability (NTCP) were (4 098.87± 184.93) cGy and (7.98±8.73)％.One way ANOVA demonstrated that under the VMAT technology,the BMIIMPd,V30,V40,EUD (or=50) and NTCP in the small intestine were the influencing factors of the occurrence of acute radiation enteritis.Conclusions If the improved BMIIMPd is utilized to distinguish the BMI,the high dose area of the small intestine will be larger and the incidence of acute radiation enteritis will be higher for patients with BMIIMPd between 10.1 and 16.9(normal and thin).Conventional BMI cannot be utilized as a basis for the prediction of the incidence of acute radiation enteritis in patients with cervical carcinoma.

Objective To evaluate the difference of set-up errors between thermoplastic mask and breast bracket in patients receiving intensity-modulated radiotherapy after breast conserving surgery,and the impact of clinical factors associated with set-up errors.Methods A total of 34 patients treated with intensity-modulated radiotherapy after breast conserving surgery from January 2016 to June 2018 were reviewed.Eighteen patients were fixed with thermoplastic mask,and sixteen were with breast bracket.Weekly CBCT scan records were used to analyze set-up errors,and group systematic and random errors were computed.The influence of clinical factors on set-up errors was also analyzed.Results The immobilization technique with thermoplastic mask showed great superior in comparison with breast bracket;however,only in the Ty(translation) and Ry(rotation),the differences had significance.Based on group systematic and random errors,PTV margins in Tx,Ty and Tzwere 2.65,4.36 and 2.87 mm in thermoplastic mask group,as well as 5.71,6.07 and 4.20 mm in breast bracket group,respectively.Multi-factor regression analysis showed that BMI was independent factors affecting set-up errors.Conclusions Compared with breast bracket,the immobilization technique with thermoplastic mask has the potential of reducing set-up errors and PTV margins in patients receiving intensity-modulated radiotherapy after breast conserving surgery,especially in patients with high BMI.

Radiomics, a technology based on multimodal medical image processing and analysis, is able to extract automatically and analyze massive data from computed tomography (CT), magnetic resonance imaging (MRI), positron emission tomography/computed tomography (PET/CT) via high-performance computer algorithm in order to pursue early diagnosis of disease, benign and malignant tumor discrimination, dynamic evaluation of disease treatment, and individualized precision therapy. To date, many studies demonstrate that radiomics not only has great potential in early diagnosis of lung cancer and prediction of genotype, treatment efficacy, as well as prognosis but also is based on imaging methods that are noninvasive, inexpensive, and repeatable. It does demonstrate precious values in guiding the clinical diagnosis and treatment of lung cancer, especially in the personalized and precise treatments and researches of lung cancer. However, the consistency and reproducibility of radiomics and the selection of robust characteristics still warrant further researches.
.