Applied Research Of GE HDCT Adaptive Statistical Iterative Reconstruction And Spectral Imaging

Part OneComparison of adaptive statistical iterative and filtered back projection reconstruction techniques in brain CTPurpose:To compare image quality and visualization of normal structures and lesions in brain computed tomography (CT) with adaptive statistical iterative reconstruction (ASIR) and filtered back projection (FBP) reconstruction techniques in different x-ray tube current-time products. Materials and methods:In this IRB-Approved prospective study, forty patients (nineteen men, twenty-one women; mean age69.5±11.2years) received brain scan at different tube current-time products (300and200mAs) in64-section multi-detector CT (GE, Discovery CT750HD). Images were reconstructed with FBP and four levels of ASIR-FBP blending. Two radiologists assessed all the reconstructed images for visibility of normal structures, lesion conspicuity, image contrast and diagnostic confidence in a blinded and randomized manner. Volume CT dose index (CTDIvol) and dose-length product (DLP) were recorded. All the data were analyzed by using SPSS13.0statistical analysis software. Results:There was no statistically significant difference between the image qualities at200mAs with50%ASIR blending technique and300mAs with FBP technique (p>0.05). While between the image qualities at200mAs with FBP and300mAs with FBP technique a statistically significant difference(p<0.05) was found. Conclusions:ASIR provided same image quality and diagnostic ability in brain imaging with greater than30%dose reduction compared with FBP reconstruction technique. Part TwoSpectral CT imaging application in the differentiation of early stage lung cancer and mediastinal lymph nodule Chapter OneComparison of the spectral CT parameters between solid, part-solid and ground-glass lung cancerPurpose:To compare the spectral parameters between the solid, part-solid and ground-glass lung cancer and to observe whether they have similar enhanced characteristics in CT spectral imaging and to provide theoretical basis for the following study about differential diagnosis of lung cancer and benign lung disease. Materials and methods:This prospective study was approved by IRB and all patients applied contrast-enhanced CT imaging for the reason of abnormal finding in the lung X-ray or pre-contrast CT examination. They were scanned by three lung series, including a low-dose pre-contrast routine helical scan of the whole lung, a contrast-enhanced spectral imaging of the lesion and another low-dose contrast enhanced routine helical scan of the whole lung. A total of257patients underwent lung spectral imaging between Oct2011and Dec2012. According to the pathological results,64patients were included in this study, and a total of21solid nodules (Group Al, n=21),23part-solid nodules (Group A2, n=23) and20ground-glass nodules (Group A3, n=20) were detected. All those nodules were proved as stage I lung cancer by surgical pathology. The CT numbers of the lesion from40keV to140keV, the water concentration (WC), the iodine concentration (IC) and the effective atomic number (Zeff) with three ROIs of different slices were recorded. The variation of HU from40keV to140keV (vHU) were calculated. We compared the mean value of each parameter and the CT numbers at40keV,70keV and140keV by one-way variance analysis using SPSS13.0software, p<0.05was considered as to have statistical significance. Same methods were conducted in the normal lung and chest wall muscle tissue. Results:For lesion, the differences of the CT numbers and the WC between the three groups had statistical significance (p<0.05), the differences of the vHU and the IC between the three groups had no statistical significance (p>0.05); the differences of the Zeff between Group A1and Group A2had no statistical significance (p>0.05), while both had statistical significance when compared to Group A3respectively (p<0.05). For normal lung and chest wall muscle, the differences of all the spectral parameters between the three groups had no statistical significance (p>0.05). Conclusions:The solid, part-solid and ground-glass lung cancers have similar spectral HU curve slope and IC. Spectral parameters between vHU and IC and between CT number at70keV and WC have positive correlation. Chapter TwoDifferentiation of early stage lung cancer and benign lung lesion with Spectral CTPurpose:To undertake an initial research with spectral CT for differentiation between early stage lung cancer and benign lung lesion. Materials and methods:This prospective study was approved by IRB. The patients’source, scanning methods and the spectral parameters’acquisition were same with Chapter One. According to the pathological and follow-up results, eighty-eight patients were included in this study, and a total of64stage I lung cancer nodules (64patients, Group A, n=64) and35benign lung lesions (25patients, note:one patient was detected a lung cancer nodule and a benign lung nodule simultaneously) were detected. Benign lung lesions were further divided into the active inflammatory disease (13patients, Group B, n=20) and the benign nodules (12patients, Group C, n=15). Because the CT numbers, WC and Zeff between solid, part-solid and ground-glass lung cancer had statistical significance while the vHU and IC didn’t. So when we conducted the statistical analysis for the CT numbers, WC and Zeff, we divided Group A into three subgroups as solid nodules (Group A1, n=21), part-solid nodules (Group A2, n=23) and ground-glass nodules (Group A3, n=20), while the vHU and IC did not. Statistical analysis was finished by one-way variance analysis using SPSS13.0software, p<0.05was considered as to have statistical significance. We drew the receiver operating characteristic curves (ROC curve) for differential diagnosis between early stage lung cancer and benign lung diseases. The area under the curve (AUC), the diagnostic threshold, sensitivity and specificity were recorded Results:The differences of the lesion CT numbers and the WC between Group Al, Group B and Group C had no statistical significance when compared to each other (p>0.05), but when Group A2and Group A3were compared to Group B and Group C respectively, the differences had statistical significance (p<0.05). The differences of the vHU and the IC between Group A and Group C and between Group B and Group C had statistical significance (p<0.05), the AUC of the vHU and IC were0.865and0.835respectively between Group A and Group C, the differences between Group A and Group B had no statistical significance (p>0.05). The differences of the Zeff between Group A1, Group A2, Group B and Group C had no statistical significance when compared to each other (p>0.05), but had statistical significance when compared to Group A3(p<0.05). When the threshold was set112.55for vHU or14.61for IC, the sensitivity and specificity were82.4%and80%respectively for the differential diagnosis between early stage lung cancer and benign lung nodule. Conclusion:Spectral CT imaging provides more diagnostic information that the CT number, spectral parameters between vHU and IC and between CT number at70keV and WC have positive correlation. The vHU and the IC could help to differentiate the early stage lung cancer from the benign lung nodule. Chapter ThreeDifferentiation of benign and malignant mediastinal lymph nodule with Spectral CTPurpose:To compare the differences of the spectral CT parameters between benign and malignant mediastinal lymph nodule and to investigate the value of spectral imaging in their differential diagnosis. Materials and methods:This prospective study was approved by IRB. The patients’source and scanning methods were same with Chapter One. According to the follow-up results, a total of30patients were enrolled in this study (we considered the patients’radiation dose fully, and in the spectral imaging we only chose the lung lesion as the scanning target, so the number of the lymph nodules was a little small.) One experienced radiologist correspond the location of lymph nodule in the surgical pathology and the CT imaging, at last, we found20malignant nodules and27benign nodules. The inclusion criteria were as follows:lymph nodule length≥10mm,7malignant lymph nodules were confirmed by pathology,13malignant lymph nodules were confirmed by CT (with criteria: malignant lung tumor+multiple mediastinal lymph nodules with length≥20mm),12benign lymph nodules were confirmed by pathology,6benign lymph nodules were confirmed by pathological benign lung disease,9benign lymph nodules were confirmed by CT (the lung lesion disappeared or contracted significantly after formal anti-inflammatory therapy in two weeks). We compared the vHU, IC, WC, Zeff and the CT numbers at40keV,70keV and140keV by independent-samples t test using SPSS13.0software,p<0.05was considered as to have statistical significance. Results:The differences of the spectral parameters for the normal lung, normal muscle and lymph nodule between the two groups had no statistical significance (p>0.05). Conclusions:Our results show negative results for the spectral imaging differential diagnosis of malignant and benign lymph nodule. The possible reasons are as follows:first, the superior vena cava contrast agent hardening artifacts may impact the spectral parameters’measurement of the2th and4th regional lymph nodules; second, some region may be found more than one metastatic lymph nodules while the other lymph nodules were benign, in this case, it was difficult to complete one-to-one correspondence for CT and pathology, and we chose the larger one as the malignant lymph nodule, this may produce the false positive results; third, there may be a certain missing rate of lymph node drawn, this may produce the false positive or negative results; forth, the relative small number of cases proved by pathology in this study may also produce negative results.