1.Intravoxel Incoherent Motion Diffusion Weighted MR in Rabbits of Liver Fibrosis Model
Lisui ZHOU ; Yong DU ; Tao PENG ; Lihua SONG ; Yumei DUAN ; Saiqun LV ; Xiangke NIU ; Guangnan QUAN
Chinese Journal of Medical Imaging 2017;25(6):414-417
Purpose To evaluate the value ofintravoxel incoherent motion (IVIM) imaging in diagnosis of liver fibrosis staging in rats.Materials and Methods Rabbit models of liver fibrosis at different stages were established.All rabbits were divided into four groups based on the pathological results of fibrosis grading as S1-S4.The 1VIM imagings with 8 b-values (0,50,100,200,300,800,1000,1200 s/mm2) were performed.The diffusion coefficient (D),perfusion-related coefficient (D*),and perfusion fraction (f) were calculated and compared between control (only injection of saline) and S 1 group,S2 and S3 group.Results The D value was significantly lower in S1 group compared with control group (P<0.05),but the D* and f values showed no significant difference between the two groups (both P>0.05).With the progression of liver fibrosis,the D,D* and f value decreased gradually;the D* value showed significant difference between S2 and S3 group (P<0.05),but the D and f values showed no significant differences between the two groups (both P>0.05).Conclusion The D value is useful for differentiation of normal liver and hepatic fibrosis of S1 stage,while the D* is valuable for differentiation of hepatic fibrosis of S2 and S3 stage.However,the f value neither could detect early fibrosis,nor could differentiate hepatic fibrosis staging.IVIM imaging provides a noninvasive method for early and accurate staging of liver fibrosis,which may be of great help in clinical diagnosis and treatment.
2.Application of combination 3D arterial spin labeling and diffusion kurtosis imaging in the grading of gliomas
Lisui ZHOU ; Yuan QU ; Tao PENG ; Xiangke NIU ; Liang SUN ; Guangnan QUAN ; Wei XIONG
Journal of Practical Radiology 2018;34(12):1842-1845,1892
Objective To evaluate the diagnostic value of combining 3D arterial spin labeling (3D ASL)and DKI in grading cerebral gliomas. Methods Thirty-four patients with pathologically confirmed gliomas were enrolled in this study.3D ASL and DKI (3 b-values were used:1,1 000,2 000 s/mm2and 15 diffusion directions)were acquired and tumor blood flow (TBF),mean kurtosis(MK),mean diffusivity (MD) and fractional anisotropy of kurtosis (FAk)were measured in the solid part of the tumor.The contralateral normal appearing white matter was used to normalize these parameters.The parameters of 3D ASL and DKI before and after normalization of high-grade and low-grade gliomas were compared.Receiver operating characteristic (ROC)curve analyses were also conducted for all parameters to determine the sensitivity and specificity for grading.Results TBF and MK values were significantly higher in the high-grade gliomas than the low-grade gliomas (All P<0.001),but there were no differences of MD and FAkbetween the two groups.The area under curve (AUC)of combination TBF and MK for glioma grading was 0.951.After normalization,the AUC of combination TBF and MK was 0.986 with a sensitivity of 94.7% and specificity of 94.3%.Conclusion The multi-parameter combined analysis of 3D ASL and DKI further improves the sensitivity and specificity of preoperative glioma grading.
3.Comparison of different post label delay time in arterial spin labeling for grading of gliomas
Yuan QU ; Lisui ZHOU ; Guangnan QUAN
Journal of Practical Radiology 2018;34(1):114-117,132
Objective To evaluate the diagnostic value of arterial spin labeling(ASL)in gliomas grading with different post label delay(PLD)time.Methods A total 30 patients with postoperative histo-pathology confirmed gliomas were included in this study.3D ASL acquisition with PLD of 1 525 ms and 2 025 ms were analyzed retrospectively.Tumor blood flow(TBF)values,as well as the ratio between TBF and the mirror ROI CBF(M-TBF),the ratio between TBF and normal contra-lateral gray matter(G-TBF)were analyzed respectively.T test between HGG and LGG groups was performed with P<0.05 regarded as statistically significant different. Receiver operating characteristic curve(ROC curve)analyses were also conducted for all parameters to determine the sensitivity and specificity for grading.Spearman rank analysis was finally used to evaluate the correlation between these parameters and glioma grades.Results For 1 525 ms PLD,TBF,M-TBF and G-TBF were all significantly higher in HGG[(88.7 ± 22.9)mL/100 g·min, 1.66 ± 0.36,1.75 ± 0.48)]than that in LGG[(68.0 ± 11.9)mL/100 g·min,1.23 ± 0.19,1.17 ± 0.21)](all P<0.01).While for 2 025 ms, TBF,M-TBF and G-TBF also showed significantly higher in HGG[(80.9 ± 18.1)mL/100 g· min,1.68 ± 0.32,1.73 ± 0.34]than that in LGG[(58.0 ± 11.7)mL/100 g·min,1.21 ± 0.16,1.15 ± 0.17)](all P<0.001).For 1 525 ms PLD,AUC values of TBF,M-TBF and G-TBF were 0.806,0.875 and 0.921,respectively.While for 2 025 ms,AUC values of TBF,M-TBF and G-TBF were 0.866,0.949 and 0.977 respectively.TBF and M-TBF showed moderate positive correlation with tumor grade both in 1 525 ms and 2 025 PLD. While G-TBF showed strong positive correlation with tumor grade for 2 025 ms PLD comparing with 1 525 ms PLD,which was moderate positive correlation.Conclusion The normalized G-TBF in 2 025 ms PLD obtains the best diagnostic performance than other parameters, which can be used as a more accurate pre-operative imaging method in grading of gliomas.