| Objective:To explore the application value of Knowledge-Based Iterative Model Reconstruction technique in the diagnosis of low-contrast lesions in brain tissue.Material and method:1.research object A total of 30 patients who Simultaneously accept CT and MRI examinations in my hospital from June 2018 to December 2018 were enrolled.There are 15 new cases of cerebral infarction and 15 cases of white matter demyelination;19 males and 11 females.Age from 47 to 62 years old,the mean age 54.5 years old.2.equipment and parameters The Philips 256 MSCT and the Siemens 1.5T MRI were used for scanning.The scanning baseline is orbitomeatal line.CT scanning parameters: tube voltage 120 k V,Automatic milliamperage regulation of tube current,collimator width 128×0.625 mm,pitch 0.91 mm,FOV23cm×20cm,layer thickness 3mm.MRI scanning parameters:FOV 23 cm × 20 cm,layer thickness 3mm,scanning sequence is T1 WI,T2WI,DWI.3.image post processing In the post-processing workstation,the FBP image of the parsing algorithm,the i Dose4 image of the Hybrid Iterative algorithm,and the IMR image of the Knowledge-Based Iterative Model Reconstruction are respectively reconstructed.The reconstructed layer thickness is 3mm.The window width and window level is 80,40.4.image quality evaluation Objective evaluation: firstly,the basal ganglia level is selected as the region of interest.The CT value and noise value(SD)of gray matter,white matter and cerebrospinal fluid are respectively measured on FBP,i Dose4 and IMR images.Secondly,thesignal-to-noise ratio(SNR)of gray matter,white matter,cerebrospinal fluid and the contrast-to-noise ratio(CNR)of the image are calculated according to the calculation formula.Subjective evaluation: two imaging diagnosticians were asked to independently score FBP,i Dose4 and IMR image quality,including gray matter contrast,subjective noise and overall image quality.5.low-contrast lesion analysis Firstly,the maximum level of low-contrast lesions was selected,and the difference between lesion density and surrounding normal brain tissue density(D)was measured.The results were divided into three groups: D≥10Hu,10>D≥5Hu,<5Hu.Secondly,Two imaging diagnosticians were asked to independently observe all lesions and score according to the extent of lesion display.Finally,the results of subjective scoring were analyzed according to the above three groups.6.statistical analysis One-way ANOVA was used to compare the objective evaluation indexes of FBP,i Dose4 and IMR images;When there were statistical differences,LSD comparisons were taken.Kappa test was used to test the consistency of the subjective evaluation indicators of the two doctors.The one-way ANOVA method was used to compare the subjective evaluation indexes of FBP,i Dose4 and IMR images;when there were statistical differences,LSD comparisons were taken.Consistency tests were performed on the lesion observations of the two physicians using the Kappa test.One-way ANOVA was used to compare the evaluation results of FBP,i Dose4 and IMR images in each group.When there were statistical differences,LSD comparisons were taken.Result:1.objective evaluation analysis There were no significant differences in CT values of gray matter and white matter between FBP,i Dose4 and IMR in conventional radiation doses(p>0.05).There were no statistical difference in CT values of cerebrospinal fluid between FBP and i Dose4(p>0.05);CT values of cerebrospinal fluid between IMR and FBP,i Dose4were statistically significant(p<0.05),and IMR of cerebrospinal fluid CT values were lower than FBP,i Dose4.Under conventional radiation dose,the noise values of FBP,i Dose4 and IMR were statistically significant(p<0.05).The noise of gray matter,white matter and cerebrospinal fluid on IMR image was lower 62%,64%,63% than FBP image,compared with i Dose4 image noise reduced by 44%,43%,46%.Under the conventional radiation dose,the difference of signal-to-noise ratio and contrast-to-noise ratio of FBP,i Dose4 and IMR were statistically significant(p<0.05).2.subjective evaluation analysis The subjective scores of the two radiologists were tested for consistency.The Kappa values of gray matter contrast,subjective noise and overall image quality respectively were 0.74,0.88,0.71.The subjective image quality evaluation of FBP,i Dose4 and IMR reconstructed images was statistically significant(p<0.05),and the subjective image quality scores of IMR images were significantly higher than i Dose4,FBP images.3.low-contrast lesion analysis Consistency test was performed on the findings of the lesions of two radiologists,and Kappa value is 0.85.When D≥10Hu,three different reconstructed images of FBP,i Dose4 and IMR showed no significant difference in lesion display(p>0.05),all three images can clearly show lesion.When 10>D≥5Hu,there were statistically significant between the three different reconstructed images of FBP,i Dose4 and IMR(p<0.05),and the IMR images showed significantly better than FBP,i Dose4.When D<5Hu,three different reconstructed images of FBP,i Dose4 and IMR showed no significant difference in lesion display(p>0.05),and none of the three images can show lesion.Conclusion:1.Under conventional radiation dose conditions,compared with FBP and i Dose4 technologies,IMR technology can significantly improve subjective-objective image quality and reduce image noise.2.Under the conventional radiation dose condition,IMR technology can significantly improve the resolving power of soft tissue and increase the detection rate oflow-contrast lesions of brain tissue(CT value is from 5HU to 10HU),which provides the possibility of early diagnosis of low-contrast lesions. |