Preliminary Study Of Intravoxel Incoherent Motion Diffusion-weighted Imaging Of The Pancreas

PartⅠ Reliability of intravoxel incoherent motion diffusion-weightedimaging of the pancreas.Purpose: To test the measurement reliability of parameters in the pancreas derived byintravoxel incoherent motion (IVIM) diffusion-weighted imaging (DWI). Materialsand Methods: Thirty-two healthy volunteers underwent respiratory triggered DWI withtwo sets of11b-values on a3.0-T system twice. Perfusion fraction (), pseudodiffusioncoefficient (D*) and pure diffusion coefficient (D) were calculated by IVIM-DWI.Parameters derived by IVIM-DWI were compared between repeated acquisitions andtwo sets of b-values by using the paired t test. Reproducibility of parametersmeasurement on repeated acquisitions were determined by using95%Bland-Altmanlimits of agreement and intraclass correlation coefficients (ICCs). Consistency ofparameters measurement on two sets of b-values were determined by using ICCs. Theeffects of b-values on the reproducibility of the parameters measurements were assessedby comparing ICCs by using the Z test. Results: There were no significant differencesin the parameters between repeated acquisitions (P>0.05), however, significantdifferences in the parameters between two sets of b-values were observed (P<0.05). Dmeasurement demonstrated good reproducibility (ICCs0.806-0.864;95%limits ofagreement12.8-13.6%), measurement demonstrated moderate reproducibility (ICCs0.659-0.711;95%limits of agreement26.8-27.5%), and D*measurement showed poorreproducibility (ICCs0.449-0.630;95%limits of agreement28.3-33%) on repeatedacquisitions. Reproducibility of parameters measurement between two sets of b-valuesdemonstrated no significant differences. There were good consistency of and D ontwo sets of b-values (ICCs0.825), while poor consistency of D*were observed (ICC0.557). Conclusion: There were good to moderate measurement reliability of D and in pancreas, while poor measurement reliability of D*were observed. The limits of measurement error of D, and D*were14%,28%and33%, respectively. Part Ⅱ Intravoxel incoherent motion diffusion-weighted imaging for thedifferentiation between pancreatic carcinoma, chronic pancreatitis and normalpancreas--preliminary studyPurpose: To assess usefulness of parameters derived by IVIM-DWI in differentiationbetween pancreatic carcinoma (PC), chronic pancreatitis (CP), and normal pancreas.Materials and Methods: Twelve patients with pancreatic carcinoma, nine patients withchronic pancreatitis and thirty-two healthy volunteers underwent respiratory triggeredDWI with11b-values on a3.0-T system. D*, and D were calculated by IVIM-DWI,while ADCtotwas calculated by using all b-values DWI. Parameters derived byIVIM-DWI and ADCtotof pancreatic carcinoma, chronic pancreatitis and normalpancreas were compared by the independent-samples t test. Results: There weresignificant differences in between PC, CP and normal pancreas (P<0.01). D*in PCwas significantly lower than CP and normal pancreas (P<0.05), while difference of D*between CP and normal pancreas showed no statistical significance (P=0.200). D in CPwas significantly lower than PC and normal pancreas (P<0.05), while difference of Dbetween PC and normal pancreas showed no statistical significance (P=0.998). ADCtotvalue in normal pancreas was significantly higher than PC and CP (P<0.01), whiledifference of ADCtotbetween PC and CP showed no statistical significance (P=0.815).Conclusion: Differences of ADC value between pancreatic carcinoma, chronicpancreatitis and normal pancreas partly caused by perfusion. Perfusion fraction () mayhelp to differentiate pancreatic carcinoma, chronic pancreatitis and normal pancreasfrom each other.