| Purpose：In this study, we performed the multibreath-hold two-dimensional susceptibility-weighted imaging technique on abdominal application. We aimed：1. To investigate the diagnostic performance of susceptibility weighted imaging (SWI)for detection of siderotic nodules (SN) in patients with liver cirrhosis, and to evaluate thepotential of SN numbers for assessing the degree of hepatic iron deposition, liver function,and liver fibrosis stage.2. To characterize the imaging manifestations of hepatocellular carcinoma (HCC) usingmulti-breath-hold two-dimensional susceptibility-weighted imaging (SWI) and compare toconventional liver magnetic resonance imaging (MRI).3. To assess the ability of2D multislice breath-hold susceptibility weighted imaging(SWI) to detect Gamna-Gandy bodies (GGBs) in the spleens of patients with PH and toevaluate the potential role of GGB number as a non-invasive marker of PH and esophagealvarices (EV).4. To establish a baseline of susceptibility-weighted imaging (SWI) phase value as a meansof detecting iron abnormalities in cirrhotic liver and to analyze its relationship with R2*.Methods：The study was divided into four parts：1. Ninety-one patients with chronic liver cirrhosis, who underwent megnetic resonanceimaging (MRI) scanning in our department between November2010and April2011, wereincluded in the study. A3.0T MRI scanner was used to acquire T1WI, T2WI, T2*WI, andSWI images. The number of nodules, signal intensity ratio (SIR), and contrast noise ratio(CNR) were recorded and analyzed by chi-square and ANOVA statistical tests. Correlation analysis was performed to evaluate the correlations between the number of SN andChild-Pugh classification, ferritin and hyaluronic acid levels.2. Forty-three patients with histopathologically confirmed HCC underwent conventionalliver MRI, multi-breath-hold two-dimensional SWI, and contrast enhanced CT preoperatively.The T1-weighted imaging (WI), T2WI, and SWI images were evaluated in consensus by twoexperienced radiologists. The tumour boundaries, blood products in the tumour, venousvessels, and non-tumour liver parenchyma were compared. The SIR and CNR were calculatedfor T1WI, T2WI, and SWI images.3. T1-, T2-and T2*-weighted imaging and SWI were performed on135patients withPH and on37control individuals. Platelet counts were collected from all portal hypertension(PH) patients. Two radiologists analyzed all magnetic resonance imaging (MRI) data, andmeasured the portal vein diameter, splenic index (SI), and platelet count/spleen diameter ratio.The numbers of patients with GGBs in the spleen were determined, and the numbers of GGBwere counted in the four MRI sequences in GGB-positive patients. The portal vein diameter,SI, platelet count, and platelet count/spleen diameter ratio of control individuals werecompared with those of GGB-negative and GGB-positive patients on SWI images. Thecorrelations among GGB numbers, the portal vein diameter, the SI, the platelet count, and theplatelet count/spleen diameter ratio were analyzed.4. Sixteen MnCl2phantoms, thirty-seven healthy individuals and87cirrhotic patientswere performed SWI and multi-echo T2*-weighted imaging, and the signal processing inNMR (SPIN) software was used to measure the radian on SWI phase images and the R2*onT2*maps. The mean minus two times standard deviation (SD) of Siemens Phase Unit (SPU)in healthy individuals was designated as a threshold to separate the regions of interest (ROIs)into high-and low iron areas in healthy participants and cirrhotic patients. The SWI phasevalues of high-iron areas were calculated. The R2*values was measured in the same ROI inboth healthy participants and patients.Results：1. The sensitivity of SWI, T1WI, T2WI, and T2*WI for detecting SN was62.5%,12.1%,24.2%and41.8%, respectively. SWI detected significantly more nodules than routine T1WI,T2WI, and T2*WI procedures (P <0.05). The SIR was the lowest in SWI (0.361±0.209), as compared to T1WI (0.852±0.163), T2WI (0.584±0.172), and T2*WI (0.497±0.196). The CNRwas the highest in SWI (13.932±5.637), as compared to T1WI (9.147±5.785), T2WI(9.771±5.490), and T2*WI (11.491±4.573). The correlation coefficients of the number of SNwith ferritin, Child-Pugh classification, and hyaluronic acid levels were0.672,–0.055, and0.163, respectively.2. SWI demonstrated significantly better tumour boundary detection than T1WI andT2WI imaging (67.4and25.6%, respectively). The detection rate for intra-tumoural bloodproducts using SWI was higher than that of T1WI and T2WI (76.7and16.3%, respectively).The detection rate for tumour venous vessels using SWI was72.1%, while none was detectedwith conventional T1WI and T2WI. The detection rate for siderotic nodules in non-tumourliver parenchyma using SWI was higher than that of conventional T1WI and T2WI (65.1and20.9%, respectively). The SIR of SWI, T1WI, and T2WI were0.76±0.13(mean±standarddeviation),2.22±0.93,1.38±0.69, respectively (F=24.82, P<0.05). CNR of SWI, T1WI, andT2WI were14.56±10.73,28.56±17.38,10.81±9.96, respectively(F=10.67，P<0.05).3. The GGB detection rate and the detected GGB number by using SWI weresignificantly greater than those by using T1-, T2-, and T2*-weighted images. The number ofGGBs in the SWI images correlated positively with the portal vein diameter and SI andcorrelated negatively with the platelet count and platelet count/spleen diameter ratio.4. SWI phase values correlated linearly with R2*values in cases of MnCl2concentrations lower than2.3mM in vitro (r=20.996, P<0.001). The mean value and SD of37healthy participants were2003and15(SPU), respectively. A threshold of1973SPU(20.115radians) was determined. The SWI phase value and R2*values had a negativecorrelation in the cirrhotic patients (r=20.742, P<0.001). However, no similar relationshipwas found in the healthy individuals (r=0.096, P=0.576). Both SWI phase values and R2*values were found to have significant correlations with serum ferritin concentrations in42patients with blood samples (r=20.512, P=0.001and r=0.641, P<0.001, respectively).Conclusions：1. The sensitivity and contrast of SWI for detecting SN in patients with liver cirrhosis arehigher than conventional MRI. The number of SN can help to assess the degree of irondeposition in patients with liver cirrhosis. 2. SWI can provide more detailed information than conventional liver MRI in evaluationof tumour boundaries, blood products, venous vasculature, and non-tumour liver parenchyma.SWI is a valuable complement to conventional liver MRI. The image quality of SWI was nothigher than T1WI and T2WI.3. SWI provided more accurate information of GGBs in patients with PH. The number ofGGB may be a noninvasive predictor of improving the selection for endoscopic screening ofPH patients at risk of EV.4. SWI phase values had significant correlations with R2*after the establishment of abaseline on the phase image. SWI phase images may be used for non-invasive quantitativemeasurement of mild and moderate iron deposition in hepatic cirrhosis in vivo. |
PDF Full Text Request