The Feasibility Study Of Mono-exponential And Bi-exponential Model Diffusion Weighted Imaging Based On GRASE Sequence In Evaluating Solitary Pulmonary Lesions

PART Ⅰ Intravoxel incoherent motion diffusion-weighted imaging of solitary pulmonary lesions:initial study with gradient-and-spin echo（GRASE）sequenceObjective:The aim of the current study was to assess the imaging quality and the repeatability and reproducibility of related parameters of the GRASE-IVIM sequence in solitary pulmonary lesions（SPLs）compared with TSE-IVIM and EPI-IVIM sequences.Methods:The clinical and imaging examination related data of Forty-two patients with forty-two SPLs who underwent chest MRI examination in the Affiliated Hospital of Jining Medical College from February 2021 to March 2022 were analyzed prospectively（38 malignant and 4 benign）.Chest MRI scanning sequences include axial T1WI,axial T2WI and axial multi-b-value IVIM-DWI（based on sequences of TSE-IVIM,GRASE-IVIM and EPI-IVIM,respectively）.Intra-and interobserver agreements were evaluated using the intraclass correlation coefficient（ICC）.The SNR,CNR and DR of TSE,GRASE and EPI-IVIM sequences were compared by Friedman test.The coefficient of variation（CV）of repeated measurement was used to evaluate the stability of repeated measurement of ADC and IVIM-derived parameters between two MRI scans.The data of the first scanning were taken,and the repeatability of ADC,D,D^*and f values between TSE-IVIM,GRASE-IVIM and EPI-IVIM sequences was evaluated by Bland-Altman.Results:Compared with TSE-and GRASE-IVIM sequences,EPI-IVIM had higher SNR,lower CNR,and higher DR（P<0.05）,but there was no significant difference in SNR,CNR,and DR between TSE-IVIM and GRASE-IVIM（P>0.05）.The reproducibility of ADC,D,D^*and f values of TSE-IVIM,GRASE-IVIM and EPI-IVIM sequences was from good to excellent（ICC>0.75）.Compared with the D^*and f values of TSE-IVIM sequences（ICC lower limit>0.90）,the D^*and f values of GRASE-IVIM and EPI-IVIM sequences are relatively poor in reproducibility（ICC lower limit<0.90）.With regard with repeatability analysis of derivative parameters of each sequence between two repeated scans.The repeatability of ADC and D values obtained by EPI-IVIM sequence was good（CV,10.03%～10.20%and 11.30%～11.57%）,while the repeatability of ADC and D values of TSE-IVIM and GRASE-IVIM sequence was excellent（TSE-IVIM:CV,1.93%～2.96%and 2.44%～3.18%;GRASE-IVIM:CV2.56%～3.12%and 3.21%～3.51%）.The repeatability of D^*and f values of all sequence perfusion parameters was relatively poor（TSE-IVIM:CV,30.03%～30.67%and 21.10%～21.15%;GRASE-IVIM:CV,31.06%～37.39%and 21.10%～21.15%;EPI-IVIM:CV,36.76%～36.77%and22.09%～23.34%）.Bland-Altman analysis showed wide limits of agreement of ADC and IVIM-derived parameters of the three sequences.Conclusions:Compared with EPI-IVIM sequences,TSE-IVIM and GRASE-IVIM sequences could reduce the image distortion rate and improve the stability of ADC and D value repeated scanning.In addition,compared with the TSE-IVIM sequence,the scanning time of GRASE-IVIM was much shorter than TSE-IVIM sequence.Therefore,GRASE could be used as a new alternative technique in lung IVIM imaging.However,there were great differences in ADC and IVIM related parameters among TSE-IVIM,GRASE-IVIM and EPI-IVIM sequences,so it is not recommended to use different sequences for the evaluation and follow-up of SPLs patients in the same cohort.PART Ⅱ The value of mono-exponential and bi-exponential model diffusion weighted imaging in differentiating benign and malignant solitary pulmonary lesions based on GRASEIVIM sequenceObjective: To explore the diagnostic performance of monoexponential and bi-exponential model DWI derived parameters in differentiating benign and malignant solitary pulmonary lesions（SPLs）based on GRASE-IVIM sequence.Methods:The clinical and imaging examination related data of 71 patients who underwent lung MRI examination for SPLs in the Affiliated Hospital of Jining Medical College from June 2021 to December 2022 were retrospectively analyzed.The MRI scanning sequences of lung include axial-T1 WI,axial-T2 WI and axial-multi-b-value GRASE-DWI.According to the pathological results,all patients with SPLs were divided into benign group and malignant group,with 19 cases in benign group and 52 cases in malignant group.Intra-class correlation coefficient（ICC）was used to evaluate the consistency among observers.The apparent diffusion coefficient（ADC）,true diffusion coefficient（D）,pseudo diffusion coefficient（D*）and perfusion fraction（f）between benign group and malignant group were compared by independent sample t-test or Mann-Whitney U test.ROC curves were drawn to evaluate the diagnostic performance and optimal threshold of mono-exponential and biexponential model DWI derived parameters in differentiating benign from malignant SPLs.Results: The ADC and D values in malignant SPLs group were lower than those of benign SPLs group [（1.14±0.05）× 10^-3mm²/s vs.（1.20±0.05）× 10^-3mm²/s,（0.97±0.07）× 10^-3mm²/s vs.（1.09±0.05）× 10^-3mm²/s,P < 0.05] with statistical significance.There was no significant difference in D^* and f values between the two groups（P > 0.05）.The AUC of ADC and D values in differentiating benign and malignant SPLs were 0.818 and 0.926,respectively,with significant difference（Z=2.048,P=0.041）.When ADC < 1.17 × 10^-3 mm²/s or D < 1.05 × 10^-3 mm²/s,a diagnosis of malignant SPLs was preferred.The diagnostic sensitivity of ADC and D value was 78.95%,and the specificity was 80.77% and 94.23% respectively.Conclusion: The mono-exponential and bi-exponential model DWI could be used to distinguish between benign and malignant SPLs based on GRASE-IVIM sequence,and the diagnostic performance of biexponential DWI model was preceded by that of mono-exponential models.