| Objective: cirrhosis(cirrhosis of liver) is one of the most common clinical digestive gland disease, it is a chronic, progressive increase of diffuse liver disease, the majority of early asymptomatic patients with cirrhosis and that China is a big country with cirrhosis, Cirrhosis after hepatitis is most. Currently, imaging diagnosis of cirrhosis of the liver remains to observe morphological changes、extrahepatic revascularization and signs of ascites and a series of analysis and judgment. This study collected and analyzed a large number of clinical patient data has been diagnosed with cirrhosis using 64-slice spiral CT dynamic enhanced scan, collect more patient’s parameters after CT scanning, and its after statistical tests, then further fisher discriminant, try to from the quantitative study CT in patients with cirrhosis were analyzed indicators research, and ultimately provide a new method for imaging diagnosis of cirrhosis.Methods: The clinical diagnosis of 86 cases of liver cirrhosis, 55 patients were male and female patients 31 cases, and the establishment of the normal control group, using the GE Optima 64-slice spiral CT, after fasting, in the shallow upper bolus iohexol injection speed 2.8-3.5ml/s, while the patient calm under preset breath breathing equipment in CT contrast injection 65-75 s after the start when the portal vein scan, scan range since the diaphragm to the iliac crest and even top line level. Scanning preferences for: thickness 5mm, speed 400 ms, tube voltage 120 kv, tube current 250 m As, window width and window level were set to 320-340 HU, 50-70 HU. Observe the resulting 1.25 mm thick original image reconstruction, spread Advantage Workstation 4.6 workstation for post-processing, the subjects were collected every measure portal vein diameter and length, diameter of the right portal vein, left portal vein diameter and length, around the portal vein branch angle, diameter and length of the splenic vein, diameter of superior mesenteric vein, the diameter of the gastric coronary vein(left gastric vein), for a total 10 indicators, SPSS13.0 statistical software for data processing, P<0.05 was statistically significant for having excluded the two was not statistically significant indicators, the remaining 8 groups of indicators Fisher discrimination.Result:1 cirrhosis group and the normal control group income index selectionCirrhosis total have 83 cases, male 53 cases, female 30 cases, the normal control group a total of 60 cases of samples, 30 males, 30 females. Were measured in the portal vein diameter and the length, the left portal vein and the right branch diameter and length et total indicators 10, the statistical software analysis, the results were statistically significant(P<0.05) index entry for eight, namely:(1) portal vein diameter,(2) portal vein length,(3) splenic vein diameter,(4) mesenteric vein diameter,(5) left portal vein diameter,(6) the length of the left portal vein,(7) right portal vein diameter,(8) left gastric vein diameter. Two showed no statistical significance, namely: splenic vein diameter and the left and right portal vein branch angle. Therefore, no significant rounding 2, there are 8 significant indicators items as Fisher discriminant.2 Fisher discriminant function equations and determine the correct income rateMale and female subjects there are some differences in the measured, values of male subjects in the measured vessel diameters and lengths were slightly higher than female subjects. So the cirrhosis group and the normal control group were divided into male and female groups, respectively its judgment, the results are as follows: Using the Fisher discriminant method to discriminate cirrhosis disease eventually come yAB(male group) = 0.1387, and to determine the male group discriminant function equation is: y(male) = 0.0008x1 + 0.0011x2 + 0.005x3 + 0.0029x4-0.0060x5 + 0.0015x6-0.0031x7 + 0.0063x8, which is x1 for main portal vein diameter; x2 for portal vein length; x3 for splenic vein diameter; x4 for superior mesenteric vein diameter; x5 for left portal vein diameter; x6 for left portal vein length; x7 for right portal vein diameter; x8 for left gastric vein diameter. Use this equation to measure for new samples data back-generation verification, get y(new) value, if y(new)> yAB(male group), the new sample belongs cirrhosis group, that is, the sample of patients with cirrhosis, if y(new) ≤yAB(male group), the new sample is a normal group, that is, the sample is not in cirrhotic patients, the diagnostic accuracy of validation results show male group reached 85.5%, a sensitivity of 86.7% and a specificity of 84.9%.Final results yAB(female group) = 0.3799, and to determine the female group function equation: y(F) = 0.0024x1 + 0.0005x2-0.0036x3 + 0.0152x4+0.0060x5+0.0064x6-0.0041x7-0.0032x8,empathyback enerations verify the accuracy of the display female group prevalence reached 96.7%, a sensitivity of 96.7% of its judgment that determines the specificity of 96.7%. The prevalence of cirrhosis comprehensive judgment accuracy was 91.1%, a sensitivity of 91.7%, specificity of 90.8%.Conclusion: Fisher discriminant method combines the image data can be a good indicator of liver cirrhosis diagnostic imaging to help, and get a high diagnostic accuracy, and this method is simple and effective, with a high clinical value. |
PDF Full Text Request