Objective To compare the position,volume and matching index (MI) of patientspecific internal gross tumor volume (IGTV)delineated by 4 different approaches based on three- dimensional and four - dimensional CT ( 3 DCT and 4 DCT ) image for primary esophageal cancers.Methods Thirteen patients with primary esophageal cancer underwent 3DCT and 4DCT simulation scans during free breathing,and the patient were divided into group A (tumor located in the proximal thoracic esophagus) and B (tumor located in the mid-and distal thoracic esophagus).IGTV were delineated using four approaches: The gross tumor volume (GTV) contours from 10 respiratory phases were combined into IGTV10 ;IGTV2 was acquired by combining the GTV from 0％ and 50％ phases; IGTVMIP was the GTV contour delineated from the maximum intensity projection (MIP) ;IGTV3D was acquired from the enlargement of 3 DCT-based GTV by each spatial direction on the motion amplitude measured in the 4DCT.ResultsTarget movement in lateral (LR),anterio-posterior (AP),superio-inferior (SI) directions showed no statistically significant difference (0.11 cm,0.09 cm,0.18 cm,respectively; χ2 =1.06,P=0.589),and there was no statistically significant difference in centroid positions between IGTV10 and IGTV2 or IGTV3D in group A (t =-2.24,-0.00,P =0.089,- 1.000 ),MI between IGTV10 and IGTV2,IGTV10 and IGTV3D were 0.88,0.54,respectively. For group B, target movement amplitude in SI direction was bigger than in LR, AP ( 0.47 cm,0. 15cm,0. 12 cm,X2= 12.00,P = 0.002).Therewasno significantdifference betweenIGTV10 andIGTV3D inLR, AP, SI ( t =- 0.80.- 0.82,- 1.16,P = 0.450.0.438.0.285 ), MI was 0.59 ; but the target center coordinates was demonstrated significant difference in SI between IGTV10 and ICTV2 for group B ( t = 2.97.P = 0.021 ), Mlwas 0.86.Thevolume of IGTVMIPwassmaller thanIGTV10 ( t =- 2.84,P = 0.025 ), but the position of IGTv10 and ICTVMIp were with no statistically significant difference in the LR,AP,SI ( t =- 0.25,0. 84. - 1.22,P = 0. 809,0.429.0.263 ) ,MIbetweenIGTV10andIGTVMIp was 0.78.Conclusions Patient-specific IGTV can be acquired from 4DCT with correct target coverage while avoiding a geographic miss for the thoracic esophageal cancer,but IGTV2 and IGTVMIP can not contain all the information about primary tumor position,shape.and size at different phases of the respiratory cycle.

Objective To analyze artifacts of gross tumor volume (GTV) and correlated factors in each phase images of four dimensional CT (4DCT) for peripheral lung cancer based on three dimensional CT (3DCT) assisted with active breathing control.Methods Nineteen patients with peripheral lung cancer underwent 3DCT (CTFB) and 4DCT simulation scans during free breathing and then underwent 3DCT simulation scans in end inspiration hold (CTEIH) and end expiration hold (CTEEH) assisted with active breathing control.The relative deviations (Devref) between the reference GTV (GTVref) and the GTVsdelineated based on CTFB (GTVFB) and all phases of 4DCT were calculated respectively.Correlations between GTVref and Devmax and between the tumor motion in the cranio-caudal (CC) direction and Devref were analyzed.Results The maximum median Devref of GTV was GTVFB with 17.83％,and the maximum median Devref of the GTV in all phases of 4DCT was GTV30 with 17.20％.A significant negative correlation was found between GTVEIH and Devmax (r =-0.691,P =0.001).The Devref was crrnelated with the tumor motion amplitude in the CC direction (r =0.323-0.617,P =0.005-0.150).The partial regression coefficient of influence of GTVref size and motion amplitude in the CC direction to the tumor Devmax were -0.500 and 0.583,P =0.002 and 0.001,respectively.Conclusions The GTV artifacts in different phase of 4DCT for the peripheral lung cancer were different to each other,and the influence of target displacement to artifacts was larger than that of target volume,so artifacts could be reduced by controlling breathing to reduce target displacement.

Objective To investigate CT appearances of abdominal primary malignant fibrous histiocytoma(MFH).Methods The CT characteristics,clinical features and pathological data of 1 7 patients with MFH proved pathologically were analyzed retrospectively. Results The lesions located in retroperitoneum were 6,in liver were 5,in kidney were 2,in superior mesentery was 1,in greater omentum was 1,in stomach was 1,in ileum was 1.The lesions are oval shape,lobulated,nodule shape,and the size of these lesions were large. 2 cases of MFH located in gastrointestinal tract were slightly low density,and the remaining were uneven high density due to necro-sis.In CT contrast enhanced scan,the solid portion and internal divisions showed progressive or continuous enhancement,and the nec-rosis were not enhanced in MFH located in the retroperitoneum,the greater omentum,the superior mesentery and the liver.MFH in kidney was poorly circumscribed and showed mild progressive enhancement lower than normal renal parenchyma.The stomach and ileum lesions showed uniform and continuous enhancement with normal gastrointestinal mucosa in corresponding parts.Conclusion Imaging features of retroperitoneal MFH were the same as those of interstitial tumors,and most tumors showed features of progres-sive and persistent enhancement,but have different imaging appearances with the malignant lesions in corresponding parts.

Objective To investigate the variations of the spatial position and overlap ratio for gross tumor volume (respiratory phase 50％) (GTV50) and internal gross tumor volume (IGTV) of primary thoracic esophageal cancer during conventional fractionated radiotherapy based on repeated four-dimensional computed tomography (4DCT) scans.Methods Thirty-three patients with thoracic esophageal cancer underwent contrast-enhanced 4DCT scans before radiotherapy and at the 10th and 20th fractions of radiotherapy.Scans were registered to the baseline 4DCT scan using bony landmarks.The GTV50 was delineated by the same radiotherapist on each 4DCT imaging data set,and the IGTV was constructed accordingly.The target volume,degree of inclusion (DI),and matching index (MI) were compared in different phases.Results The volumes of GTV50 and IGTV decreased along with treatment course.No significant changes in the centroid position were observed for the GTV50 and IGTV.The median DIs of the target volumes at the 10th and 20th fractions in the original target volume were 0.75 and 0.63(P =0.000) for GTV50 and were 0.79 and 0.66(P=0.000) for IGTV,while the median MIs were 0.61 and 0.56(P=0.002) for GTV50 and were 0.68 and 0.58 (P =0.005) for IGTV.A positive correlation between the variation of volume ratio and the variation of DI was found for GTV50 and IGTV (r =0.632,r =0.783),and the variation of volume ratio was also positively correlated with the variation of MI (r =0.387,r =0.483) ;the 3D vector was negatively correlated with the MI (r =-0.455,r =-0.438).Conclusions During conventional fractionated radiotherapy,the variation of spatial position is less than 0.8 cm for GTV50 and IGTV of primary thoracic esophageal cancer,and the decline of the target leads to varying degrees of decreases in DI and the MI.

ObjectiveTo explore the influence of the CT scanning mode on the gross tumor volume (GTV) delineation of solitary pulmonary lesion (SPL),and to evaluate the feasibility of the spiral CT scan in CT simulation.Methods Sixteen patients with SPL underwent axial scan,spiral scan during free breathing.Compare the target position,volume between GTVS and GTVA (paired t-test).The matching index (MI) between GTVS and GTVA and correlations between MI and the tumor volume were calculated ( bivariate correlation analysis).ResultsGTVS and GTVA volume was 8.95 cm3 and 9.38 cm3 ( t =0.43,P =0.667),respectively.The centroid position for GTVS and GTVA in x,y and z axises were 6.80 cm and 6.81 cm (t =0.27,P =0.794),36.19 cm and 36.05 cm (t =0.37,P =0.717),and 4.99 cm and 4.96cm (t =0.65,P =0.526),respectively.There were also no statistically significant difference in the distance between the centroidal position and origin of coordinates for GTVS and GTVA (38.31 cm∶ 38.23 cm,t =0.47,P =0.646 ).MI between GTVS and GTVA was 0.36 ( range 0-0.77 ),correlated with the tumor volume (r =0.587,P =0.017).ConclusionsThere was no significant difference between the axial scan and spiral scan in the GTV volume and position for SPL,but MI between GTVS and GTVA were small.A correlation was found for the MI between GTVS and GTVA with the tumor volume.Spiral CT scan was more timesaving,and was feasible in CT simulation scan.

Objecttve To compare the positional and volumetric differences of planning target volumes(PTVs)based on axial three-dimensional CT(3D-CT)and four-dimensional CT(4D-CT)for the primary tumor of non-small cell lung cancer(NSCLC).Methods Sixteen NSCLC patients with lesions located in the upper lobe and 12 patients with lesions in middle and lower lobes,totally 28 patients, initially underwent three-dimensional CT scans followed by 4D-CT scans of the thorax under normal free breathing.PTVvector was defined on gross tumor volume (GTV) contoured on 3D-CT and its motion vector. The clinical target volumes(CTVs)were created by adding 7 mm to GTVs,then, internal target volume (ITVs)were produced by enlarging CTVs isotropically based on the individually measured amount of motion in the 4D-CT,lastly,PTVs were created by adding 3 mm setup margin to ITVs. PTV4D was defined on the fusion of CTVs on all phases of the 4D data.The CTV wag generated by adding7 mm to the GTV on each phase.then,PIVs were produced by fusing CTVs on 10 phases and adding 3 mm setup margin.The position of the target center,the volume of target and the degree of inclusion(DI)were compared reciprocally between the PTVvector and the PTV 4D The difference of the position,volume and degree of inclusion of the targets between PTVvecter and PTV4D were compared,and the relevance between the relative characters of the targets and the three-dimensional vector was analyzed based on the groups of the patients. Results The median of the 3 D motion vector for the lesions in the upper lobe was 2.8 mm, significantly lower than that for the lesions in the middle and lower lobe ( 7.0 mm, z = - 3. 485, P < 0. 05 ). In the upper lobe group there was only significant spatial difference between the PTVvector and PTV4D targets in the center coordinate at the x axe (z = -2. 010, P < 0. 05 ), while in the middle and lower lobes there was only significant spatial difference between the PTVvector and PTV4D targets in the center coordinates at the z axe (z = -2. 136,P <0.05). The median of ratio of PTV4D and PTVvector, of the upper lobe group was 0. 75, significantly higher than that of the middle and lower lobes group (0. 52, z = - 2. 949, P < 0. 05 ).A significant correlation was found for the motion vector and the ratio of PTV and PTV4D in both groups ( r = - 0. 638, - 0. 850, P < 0. 05 ). For all patients, the median of D[ of PTV4D in PTVvector was 66. 39% ,while the median of DI of PTVvector, in PTV4D was 99. 55% , both showed a positive significant correlation with the motion vector (r = -0. 814,0. 613 ,P < 0. 05). Conclusions PTV4D defined based on 4D-CT simulation images is obviously less than PTV defined based on 3D-CT simulation images. The ratio and DI of both targets are related with the three-dimensional motion vector of the tumor.

Objective To compare positional and volumetric differences of internal gross tumor volume (IGTV) delineated separately by three approaches based on four-dimensional CT (4DCT) for the primary tumor of non-small cell lung cancer (NLCLC). Methods Twenty-one patients with NLCLC underwent big bore 4DCT simulation scan of the thorax. IGTVs of the primary tumor of NSCLC were tumor on the MIP images were delineated to produce IGTVMIP. The position of the target center, the volume of target, the degree of inclusion (DI) and the matching index (MI) were compared reciprocally between IGTV10, IGTVEI+EE and IGTVMIP. Results Average differences between the position of the center of IGTVs on direction of x,y and z axes were less than 1 mm, with no statistically significant difference. The volume of IGTV10 was larger than that of IGTVEI+EE, the difference was statistically significant (t=2.37,P=0.028);the volume of IGTV10 was larger than that of IGTVMIP, but the difference was not statistically significant(t=1.95 ,P=0.065). The ratio of IGTVEI+EE with IGTV10, IGTVMIP with IGTV10 were 0.85±0.08 and 0.92±0.11, respectively. DI of IGTVEI+EE in IGTV10, IGTVMIP in IGTV10 were 84.78% ± 8. 95% and 88.47% ±9.04%. MI between IGTV10 and IGTVEI+EE, IGTV10 and IGTVMIP were 0.85 ±0.09, 0.86±0. 09, respectively. Conclusions The center displacement of the IGTVs delineated separately by the three different techniques based on 4DCT images are not obvious; IGTVEI+EE and IGTVMIP can not replace IGTV10 , however , IGTVMIP is more close to IGTV10 comparing to IGTVEI+EE . The ratio of GTVEI+EE with IGTV10 is correlated to the tumor motion vector. As the vector increases, the ratio of GTVEI+EE with IGTV10decreases, especially for small tumors.

Objective To investigate the correlations of the whole breast displacement in different respiratory cycle during free breathing (FB) following breast-conserving surgery to the displacement of selected skin marker,nipple,and selected surgical clip based on four-dimensional computed tomography (4D-CT).Methods Thirteen breast cancer patients who had undergone breast-conserving surgery received whole breast intensity-modulated radiotherapy (IMRT).Respiration-synchronized 4D-CT image data were gathered during FB and were exported to the Varian Eclipse treatment planning system,and the whole breast target,nipple,superior clip,and metal marker on the skin at the anterior body midline were delineated on the CT images of ten phases of the respiratory cycle by the same radiotherapist based on the same delineating criteria.The displacement distances of the delineated target in the mediolateral (x),anteroposterior (y),and superoinferior (z) axles were achieved,and the correlations of the whole breast target displacement to the displacement of the clip,nipple,and skin marker were analyzed.The ipsilateral lung was delineated on the CT images of every phase of the respiratory cycle,and the changes in ipsilateral lung volume were analyzed during the respiratory cycle relative to the displacement of the breast.Results The maximal displacement distances of the whole breast target in the x,y,and z axles during FB were 0.71,0.76 and 1.29 mm,respectively ( F =5.755,P ＜ 0.05 ).There was no relationship between the three-dimensional (3D) displacement of the whole breast and the volume of the whole breast (r =-0.264,P ＜ 0.05 ),and there was no relationship between the displacemeat of the whole breast and the volume change of the ipsilateral lung ( r =0.346,P ＜ 0.05).There was no significant difference among the mean target displacement distances in 3 axles,and among 2 selected successive end-inspiration (EI) phases and 3 selected successive end-expiration (EE) phases.There was no significant difference between the volumes of the whole breast targets at the selected El and EE phases.There was no relationship between the displacement of the whole breast target and the displacement of the nipple,skin marker or superior clip in the cavity along the x- and z-axles.Along the y-axle,8/13,7/11 and 9/13 of the patients showed displacement of the whole breast target relative to the displacement of the nipple,skin marker and superior clip respectively.However,according to a population-based analysis,the displacement of the whole breast target was only significantly associated with the displacement of the superior clip ( r =0.657,P ＜ 0.05 ).Conclusions The clip registration is more credible and sensitive than a skin marker or the nipple for measuring and correcting the displacement of the whole breast target during radiotherapy.

Objective To investigate the variations of the spatial position and overlap ratio of the internal target volume (ITV) and planning target volume (PTV) of primary thoracic esophagus carcinoma using repeated four-dimensional computed tomography (4D-CT) scanning during conventional fractionated radiotherapy.Methods Thirty patients with thoracic esophageal carcinoma were included whose 4D-CT scans were conducted before radiotherapy and between every ten fractions.The gross tumor volumes (GTVs) were delineated by the same radiation oncologist on each 4D-CT image phase,and the ITV and PTV were constructed afterwards.Results No significant difference of the isocenters was observed for the ITV and PTV during the treatment course,yet both the volumes of the ITV and PTV decreased.The median DI (the degree of inclusion) of the target acquired subsequently in the original target were 0.85,0.77 (Z=-3.10,P ＜0.05) for ITV and 0.86、0.82(Z =-2.49,P ＜0.05)for PTV respectively during entire treatment.The variation of volume ratio correlated strongly with the variation of DI (the DI of the target acquired subsequent in the original target) (rITV =0.71,rPTV =0.77,P ＜0.05).The variation of volume ratio and the variation of the matching index (MI) were positively correlated (rITV =0.47,rPTV =0.59,P ＜ 0.05).The 3D vectors of ITV and PTV motions were negatively correlated with the corresponding MI (rITV =--0.52,rPTV =-0.36,P ＜ 0.05).If the initial PTV was used for treatment planning,8.80％ and 6.37％ of the target volume would be missed at the tenth and twentieth fraction (Z =-0.55,P ＞ 0.05),respectively.In the meanwhile,11.45％ and 18.49％ of the normal tissues would be wrongly irradiated at the corresponding time points (Z =-2.49,P ＜ 0.05).Conclusions The variations of the spatial position of all targets were all less than 0.6 cm.The DI and the MI of the target decrease by various degrees during the treatment course,which lead to target mispositioning and normal tissue irradiation at different levels.

Objective:To investigate the associations between small diffusion-weighted imaging (DWI) hyperintensities lesions and total cerebral small vessel disease (cSVD) burden and the influence on prognosis in patients with acute intracerebral hemorrhage (ICH).Methods:Consecutive patients with acute spontaneous ICH from January 2018 to June 2021 were recruited in the Stroke Center of Zhengzhou People′s Hospital. Magnetic resonance imaging was performed to quantify DWI hyperintensities lesions and cSVD imaging markers, including white matter hyperintensities, enlarged perivascular spaces, lacunes and cerebral microbleeds, which were calculated for the total cSVD burden (0-4 points). The prognosis was assessed with the modified Rankin Scale (mRS) at discharge and 90-day. Multivariable Logistic regression models were adopted to explore the associations between DWI lesions and total cSVD burden and clinical outcome.Results:Of 283 included patients, 59 (20.8%) had small DWI lesions, 32 (11.3%) had multiple lesions. They were mostly punctate, mainly located in the cortical and subcortical regions, and scattered in multiple vascular territories. With the increase of cSVD burden, the number of DWI lesions gradually increased. Spearman correlation analysis showed that the total cSVD burden was positively correlated with the number of DWI lesions ( r=0.21, P<0.001). In multivariable regression analyses, the total cSVD burden was independently associated with DWI lesions ( OR=1.63, 95% CI 1.23-2.15, P=0.001). The 90-day poor outcome (mRS scores≥4) in patients with DWI lesions was significantly higher than those without DWI lesions (39.3% vs 16.3%, χ 2=14.38, P<0.001), while there was no statistically significant difference in the poor outcome of discharge between the two groups (26.5% vs 17.7%, χ 2=3.06, P=0.080). With the increase in the number of DWI lesions, the 90-day poor outcome increased significantly (trend chi-squared test χ 2=11.50, P=0.001). Multivariable analyses showed that DWI lesions ( OR=4.39, 95% CI 1.92-10.03, P<0.001) and their number ( OR=1.42, 95% CI 1.06-1.90, P=0.018) were independently associated with the 90-day poor outcome. Conclusions:Higher total cSVD burden is an independent risk factor for small DWI lesions in patients with ICH. Small DWI lesions were independently associated with the 90-day poor outcome, but not with the discharge outcome.