Font Size: a A A

Image Quality Evaluation Of Diffusion-weighted MRI In Chest And Study Of DWI In The Differential Diagnosis Of Lymph Nodes With Lung Cancer

Posted on:2014-08-30Degree:MasterType:Thesis
Country:ChinaCandidate:X ZhangFull Text:PDF
GTID:2254330398965035Subject:Imaging and nuclear medicine
Abstract/Summary:PDF Full Text Request
ObjectiveFirst:To evaluate the image quality of diffusion-weighted MRI under different bvalues in lung cancer and optimize b value.Second:To explore the value of diffusion-weighted imaging (DWI) under different b values in differentiation metastatic from nonmetastatic mediastina and hilar lymph nodes with lung cancer.Materials and MethodsPart One Image quality evaluation of diffusion-weighted MRI under different b values in chest.42patients with lung cancer were confirmed with pathology methods (operation, bronchoscope and aspiration biopsy) and examined with conventional MR and DW imaging(b=0s/mm2,800s/mm2,1000s/mm2,1200s/mm2).DW images under different b values were obtained. SNR and CNR of each lung tumor were analyzed and calculated from DWI under different b values. Image quality was graded I-V. One-factor analysis of variance to compare the differences between groups of SNR,CNR under different b values,and chi square test to evalue image quality differences of DWI under different b values.Then, to optimize b value.Part Two Application of DWI with under b values in the differential diagnosis of lymph nodes with lung cancer.25lung cancer with mediastinal and (or)hilar lymphadenectasis before surgical operation were examined with conventional MR and DW imaging(b=0s/mm2,800 s/mm2,1000s/mm2,1200s/mm2). DW images and ADC maps under different b values were obtained. Minor axis and ADC value of each mediastinal and (or) hilar lymph nodes were calculated from T2WI and ADC maps under different b values. According to operative and pathological findings,78lymph nodes were divided into two groups:43nonmetastatic lymph nodes,35metastatic lymph nodes. Student’s t test to compare the difference between metastatic and nonmetastatic lymph nodes’minor axis. One-factor analysis of variance to compare the differences between groups of ADC values under different b values, Student’s t test to compare the difference between metastatic and nonmetastatic lymph nodes’ADC value, and ROC curve was used to analyze the diagnostic performance of the ADC under different b values in differentiation metastatic from nonmetastatic lymph nodes.Then select the optimal b value and confirm the diagnostic threshold. ROC curve was used to analyze the diagnostic performance of the ADC and minor axis in differentiation metastatic from nonmetastatic lymph nodes.ResultsPart OneSNR and CNR were gradually reduced with increasing b value, and the differences were statistically significant(FSNR=50.464,P=0.000; FCNr=8.349,P=0.000). The SNR differences under different b values between groups were statistically significant (P=0.000). The CNR difference between800s/mm2and1000s/mm2was not statistically significant (P=0.063). The CNR differences between800s/mm2and1200s/mm2, between1000s/mm2and1200s/mm2were statistically significant (P=0.000, P=0.029). Image quality differences of DWI under different b values were not statistically significant(P>0.05).The optimal image quality was0,800s/mm2, because its SNR was maximal.Part TwoThe minor axis differences between metastatic and nonmetastatic lymph nodes were statistically significant(t=-4.480,P=0.000). The ADC values of metastatic and nonmetastatic lymph nodes were gradually reduced with increasing b value, and the differences were statistically significant(F非=27.454,P=0.000; F转=6.487,P=0.002). The ADC value differences of non-metastatic lymph nodes under different b values between groups were statistically significant(P=0.000,P=0.000,P=0.002). The ADC value difference of metastatic lymph nodes between800s/mm2and1000s/mm2was not statistically significant (P=0.553). The ADC value differences of metastatic lymph nodes between800s/mm2and1200s/mm2and between1000s/mm2and1200s/mm2were statistically significant (P=0.003, P=0.030). The ADC of metastatic lymph nodes was obviously lower than that of nonmetastatic lymph nodes(t800=-5.888,P<0.05; t1000=-4.964,P<0.05; ti2oo=-5.139,P<0.05).ROC analysis shows the area under curve(AUC) were0.827,0.805,0.797(b=800s/mm2,1000s/mm2,1200s/mm2respectively), with diagnosis value(AUC>0.5).The difference among them was not statistically significant(P>0.05).The threshold ADC value were2.210×10-3mm2/s,1.862×10-3mm2/s,1.625×10-3mm2/s when b value was0,800s/mm2,0,1000s/mm2,0,1200s/mm2.Sensitivity,specificity, positive predictive value and negative predictive value was82.9%,69.8%,69.0%,83.3%;74.3%,74.4%,70.3%,78.0%;82.9%,69.8%,69.0%,83.3%; respectively. ROC analysis shows the AUC of ADC (b=800s/mm2) and minor axis in differentiation metastatic from nonmetastatic lymph nodes were0.827,0.771, respectively. The difference between them was not statistically significant(P>0.05).ConclusionPart One The optimal image quality was0,800s/mm2.Part TwoThe ADC values under different b values were diadynamic criteria in differentiation metastatic from nonmetastatic mediastina and (or) hilar lymph nodes with lung cancer. DWI can provide proper diagnosis value in differentiation metastatic from nonmetastatic lymph nodes with lung cancer.
Keywords/Search Tags:Diffusion-weighted imaging, Apparent diffusion coefficient, Lung cancer, Metastatic, Lymph nodes
PDF Full Text Request
Related items