Pulmonary Parenchyma Perfusion Imaging Using256iCT And The Application In Patients With COPD

【Background】Chronic Obstructive Pulmonary Disease is a disease characterized by progressive airflowlimitation, and not fully reversible. COPD is the fourth leading cause of death in the world and600million people worldwide suffer from COPD. World Health Organization report thatthere was a average of300million people died of COPD every year in the world. In China,about100million people died of COPD each year. Therefore, COPD has causedtremendous social and economic burden, and has become the major health problems. Earlydetection, early diagnosis, early treatment is key to reduce mortality. At present, clinicaldiagnosis of COPD is through a comprehensive assessment of smoking history and otherrisk factors, clinical symptoms and pulmonary function tests. But the less extent ofimpaired lung function in early COPD patients and the symptoms are not typical, andeasily be neglected. Although pulmonary ventilation function parameters FEV1wasextensively used, it does not reflect the local lung function and is not sensitive to the earlychanges of lung function and does not detect changes in lung ventilation of early COPD.Chronic Obstructive Pulmonary Disease Global Initiative for Chronic Obstructive LungDisease (GOLD) promulgated by the American Thoracic Society (ATS) in2004, reportthat FEV1is not fully reflect all the clinical situation of COPD patients, especially there areobvious limitations to the partial assessment of the degree of impairment of lung functionof uneven distribution of emphysema. Early COPD only changes in lesions of pulmonaryinterstitial fibrosis and in the local alveolar and alveolar wall and in local interstitialcapillaries and small artery changes. Routine chest X-ray examination is not easy to find.CT examination can evaluate the changes of COPD local structures such as airways,lung parenchyma and vascular. The high-resolution CT (HRCT) can find the change of thelocal anatomical details, such as the changes of the small airways and lung parenchyma.CT functional imaging, such as lung perfusion imaging, pulmonary ventilation imaging,can find the abnormality of local ventilation, abnormal regional blood flow distribution.Inour study we applied256iCT to have perfusion imaging in COPD and optimized perfusionimaging scanning parameters first, then evaluate the application value of early diagnosis inCOPD. Pulmonary parenchyma perfusion imaging using256iCT: the effect of radiationdose on image qualityObjective: To evaluate the effect of different radiation dose on image quality of perfusionimaging using256iCT and to obtain better perfusion images with radiation dose as low aspossible. Materials and Methods: thirty healthy volunteers underwent256iCTexamination. All volunteers were divided into six groups randomly: group1to3, tubevoltage120KV, tube current of200,100,50mAs; group4to6, tube voltage80KV, tubecurrent of200,100,50mAs. Image quality of iCT parenchymal perfusion of each groupwas recorded subjectively and objectively. At the same time, a single radiation dose wascalculated. Statistical analysis was performed using the software of SPSS18.0version.Results:(1) Subjective evaluation: there was no significant difference between group1and2(P=0.677), and between group2and4(P=0.098); No statistically significantdifference was found among group3to5(P>0.05); With regard to group1to5, each groupshowed a statistically significant difference from group6(P<0.05).(2) Objective evaluation:there was no significant difference between group1and2(P=0.409); No statisticallysignificant difference was found among group3to5(P>0.05); With regard to group1to5,each group showed a statistically significant difference from group6(P<0.05).(3)Radiation dose: the dose of radiation in group1,2,4,3,5,6was13.3mGy,9.4mGy,3.8mGy,3.3mGy,1.9mGy,1mGy, respectively; therefore, although the group1and2havethe best image quality, the radiation dose of them were the top among all groups. Therewas medium image quality in group3-5and no statistically significant difference wasfound among them. The mean dose of radiation in group5was relatively lower with adecrease of42.4%and50%compared with that in group3and group4. The meanradiation dose of group6was the lowest, but it had the worst image quality. Conclusions:the group5(80kv,100mAs) can provide better image quality of parenchyma perfusionimaging and had lower radiation dose using256iCT. The study of pulmonary perfusion imaging in patients of COPD using iCTObjectives: To evaluate the morphological characteristics of pulmonary perfusion with256iCT in patients with Chronic Obstructive Pulmonary Disease (COPD) phenotypes andto semi-quantitative analysis perfusion defects. Then we assess the correlation of theperfusion defects with pulmonary function test (PFT). To assess the relationship betweenlung parenchymal destruction and pulmonary perfusion defects on a lobar level using256iCT in patients with COPD. The value of early diagnosis of iCT pulmonary perfusionimaging for COPD was preliminary investigated. To assessed the difference of iCTPulmonary Perfusion Imaging in Patients With Chronic ObstructivePulmonary Disease and Controls. To quantitatively evaluate the parenchymalperfusion imaging using256iCT and to analysis the correlation with pulmonary functiontest (PFT) and emphysema index in patients with chronic obstructive pulmonarydisease(COPD). Materials and Methods:1.63patients with COPD underwent PhilipsBrilliance256iCT perfusion examination. According to the dominancy of emphysema andthe presence of bronchial wall thickening evaluated by chest high-resolution computedtomography (HRCT), all patients were classified into three phenotypes: A phenotype,absence of emphysema, with little emphysema with or without bronchial wall thickening;E phenotype, emphysema without bronchial wall thickening; and M phenotype,emphysema with bronchial wall thickening; Then we observed the patterns of PD(wedge-shaped, circum-scribed but not wedge-shaped, and patchy) on pulmonaryperfusion imaging on a lobar level for COPD phenotypes and calculated the CT valuedifference ratio (RHU) between the perfusion defects and the normal lung(RHU=HUabnormal/HUnormal). Statistical analysis was performed using the software of SPSS18.0version.Analysis of variance and Bonferroni test was applied to analysis the differencesamong the RHUof COPD phenotypes. The Pearson correlation analysis was used to assessthe correlation RHUwith pulmonary function test (PFT).2.63patients with chronicobstructive pulmonary disease(COPD) underwent inspiratory3D-HRCT and256iCTperfusion. All patients with COPD were divided into three groups according to PFT:group of mild COPD (GOLD Stage I, sixteen patients), group of moderate COPD(GOLDStage II， eighteen patients), group of severe COPD (GOLD Stage Ⅲ-Ⅳ,twenty-ninepatients). Quantitatively analysis of emphysema using3D-HRCT data is most often done by threshold based methods (density mask technique), where all voxels with hounsfieldunits (HU) below950HU. Emphysema was categorized in five clusters with differentvolumes.3D-HRCT and lung perfusion were visually analyzed for five lobes on wholelung using a five-point-score to grade the abnormalities on CT (0: No emphysema;1: thevolume of emphysema <25%of the lobe;2: the volume of emphysema is in25-50%ofthe lobe;3: the volume of emphysema is in50–75%of the lobe;4: the volume ofemphysema is in>75%of the lobe) and CT perfusion (0: normal perfusion without anydefects;1: Small perfusion defects,<25%of the lobe;2: Large perfusion defects,25–75%of the lobe;3: Large perfusion defects,50–75%of the lobe not perfused;4: Loss ofperfusion>75%of the lobe). Statistical analysis was performed using the software of SPSS18.0version. For agreement between3D-HRCT and CT perfusion results weightedKappa-Analysis was performed.3.63patients with COPD and20controls underwent256iCT perfusion examination. All subjects were divided into five groups according toPFT: group of controls (normal,twenty subjects), group of mild COPD (GOLD Stage I,sixteen patients), group of moderate COPD (GOLD Stage II， eighteen patients), group ofsevere COPD (GOLD Stage Ⅲ, eleven patients), group of extremely severe COPD(GOLD Stage Ⅳ, eighteen patients).If the presence of perfusion abnormality, the contrastbetween normal lung and perfusion defects was quantified by calculating the CT valuedifference ratio (RHU) between the perfusion abnormalities and the normal lung on a lobarleve（lRHU=HUabnormal/HUnormal）. Statistical analysis was performed using the software ofSPSS18.0version and Analysis of variance and Bonferroni test was applied.4.63patientswith COPD underwent256iCT perfusion examination. If the presence of perfusion defectson a lobar, we calculate the CT value difference ratio (RHU) between the perfusionabnormalities （HUabnormal）and the normal lung （HUnormal）on a lobar level(RHU=HUabnormal/HUnormal). From3D-HRCT data, emphysema index(EI) was the volumefraction of the lung below-950Housefield Units was assessed. The correlation of the RHUwith EI and PFT was assessed using Spearman correlation analysis. Results:1.Therewas38patients of A phenotype,11patients of E phenotype,14patients of M phenotypein COPD patients. The pattern of PD with A phenotype was mainly circum-scribed but notwedge-shaped. The pattern of PD with E and M phenotype was mainly patchy. There wassignificant difference of RHUbetween A phenotype and E phenotype（P=0.003），Nosignificant difference of RHUwas found between A phenotype and M phenotype, betweenM phenotype and E phenotype（P>0.05）. The Spearman correlation test showed that RHU of the A phenotype positively correlated with forced expiratory volume in1second (FEV1)（R=0.48，P=0.002）, FEV1%Predicted（R=0.47，P=0.002）, forced vital capacity (FVC)（R=0.57，P=0.01）, FEV1/FVC（R=0.42，P=0.008）. The Spearman correlation testshowed that RHUof the E phenotype positively correlated with FEV1/FVC （R=0.68，P=0.042）, and then RHUof the M phenotype positively correlated with FEV1（R=0.72，P=0.019）, FVC （R=0.65，P=0.043）, FEV1/FVC（R=0.63，P=0.039）.2. For groupof mild COPD, a total of80lobes were evaluated. At3D-HRCT, the score was1for25lobes,2for28,3for9,4for, and5for7lobes.At CT perfusion, the score was1for11lobes,2for18,3for26,4for14, and5for11lobes. Matching of lung parenchymaldestruction and reduced perfusion was found in34lobes(weighted kappa=0.286).Regional heterogeneity of emphysema on3D-HRCT did not match with the decreasedperfusion on iCT in this group; For group of moderate COPD, a total of90lobes werevaluated. At3D-HRCT, the score was1for9lobes,2for21,3for37,4for15, and5for8lobes.At CT perfusion, the score was1for5lobes,2for15,3for30,4for21lobes, and5for19lobes. Matching of lung parenchymal destruction and reduced perfusion was foundin53lobes(weighted kappa=0.46). For group of severe COPD, a total of145lobes werevaluated. At3D-HRCT, the score was1for11lobes,2for16,3for24,4for37, and5for57lobes.At CT perfusion, the score was1for9lobes,2for7,3for25,4for44lobes, and5for60lobes. Matching of lung parenchymal destruction and reduced perfusion was foundin113lobes (weighted kappa=0.46).3. With regard to mild COPD to extremely severeCOPD, each group showed a statistically significant difference from the controls (P<0.05).There was significant difference between mild COPD and severe COPD (P=0.022), andbetween mild COPD and extremely severe COPD (P=0.001); At the same time,statistically significant difference was found between moderate COPD and extremelysevere COPD (P=0.003); There was no significant difference between othergroups(P>0.05).4. RHU、FEV1、FEV1%Predicted、FVC、FEV1/FVC、EI were0.356±0.16,1.59±0.84liters,56.71±26.25(%),2.54±0.94liters,60.66±14.02(%),10.6±7.58(%), respectively. The Spearman correlation test showed that RHUpositivelycorrelated with forced expiratory volume in1second (FEV1)(R=0.53, P=0.000),FEV1%Predicted (R=0.514, P=0.000), forced vital capacity (FVC)(R=0.41, P=0.002),FEV1/FVC (R=0.466, P=0.000) and negatively correlated with EI (R=-0.47, P=0.000).Conclusions:1.This study shows that there was difference of patterns of PD and RHUofsemi-quantitative parameter among COPD phenotypes. And the RHUof the phenotypes always positively correlated with FEV1/FVC.2.3D-HRCT and CT perfusion show a highlobar agreement between parenchymal destruction and reduction of perfusion in patientwith severe COPD. However, the similar result was not found in patients with mild andmoderate COPD. CT pulmonary perfusion has the advantage in discovering early changesof pulmonary hemodynamics.3. There was obvious difference of RHUof Semi-quantitativeparameter between COPD and controls. RHUmay become a predictor of severity ofCOPD.4. the RHUmeasured on CT perfusion in patients with COPD, correlates withparameter of PFT and emphysema index. Therefore RHUcould become a predictor of locallung function.