Differentiaion Between Primary Benign And Malignant Ovarian Tumors At Spectral CT:a Preliminary Study

Objective:The aim of this retrospective study was to investigate the diagnostic performence of energy spectrum CT to distinguish primary benign from malignant ovarian tumors by analying the results of energy spectrum CT and pathologic findings of patients with primary ovarian tumor.Methods:58 female patients(19-64 years)with pathologically-proved primary benign or malignant ovarian tumor who underwent dual phase contrast-enhanced energy spectrum CT scan from May 2015 to August 2016 in our department were included in this study,and were divided into two groups:benign group(n=30)and malignant group(n = 28).Inclusion criteria:?women;?lesions have solid components:?not allergic to idionated contrast agent;?have patholigic diagnostic results performed by our hospital or other tertiary hospital.All patients underwent pelvic flat-scan and dual-phase contrast-enhanced scanning on Gemstrom Spectrum CT(Discovery CT 750 HD.GE Heahhcare,Milwaukee,USA).The scope of the scan included at least from the iliac to the lower edge of the pubic symphysis.Both phases used energy spectrum scanning mode(GSI).Tube voltage was high and low energy(80/140 kVp)instantaneous(0.5ms)switch,tube current 375mAs.speed 0.5s/circle,detector width 40mm Contrast medium(iohexol,300 mg I/ml.1.2 ml/kg)was injected intravenously at a rate of 3.0 ml/s with a high-pressure syringe.The arterial phase images were obtained at 25-30s after the injection of contrast medium,and the venous phase images were obtained at 60-75s.Image were reconstructed using standard algorithms.Informed consent was obtained from every patients before scanning.The data of energy spectrum image was reconstructed into the slice thickness and increment of 1.25mm.The single energy images were transferred to the AW4.4 workstation for energy spectrum analysis,and the energy spectrum imaging software(GSI Viewer)wes utilized for displaying single energy images and quantitative measurement.The region of interest(ROI)was placed at the level which reveales as much solid component as passible of the tumor.ROI placement criteria:?select the level which reveales as much solid component as passible and place the same size of the circular or oval ROI,and then select the upper and lower levels for repeat measurement,the three measurement results were average as the final results.?ROI area of 50?60mm2;?measurement should be performed at the same two-level to the great extent;?ROI shoul be placed to avoid the cystic,necrostic.liquified portion and vascular area of the tumor.The average CT value of 40?100keV single energy(interval 10keV),IC content(ICg/L),water content(WCg/L).slope of energy spectrum curve of 40?100keV and the ROI energy spectrum curve in the region of interest were obtained.The slope of the energy spectrum curve betw een 40 and 100 keV is calculated from the following equation:?=CT(40keV)-CT(100keV)/| 40keV-100keV |.The results of the two groups were compared with the corresponding pathological results.Student single-sample t or rank sum test was utilized for comparing measurements,?2 test was utilized for comparing rate or ratio.For evaluating the association between the results of energy spectrum CT and pathologic findings.Kappa test was utilized.SAS 9.4 statistical software was utilized for all the statistical analyses.P value of<0.05 indicatede a statistically significant difference.Result:In both the arterial venous phases,the curve of the energy curve of the two groups showed arc-shaped decrease with the increase of single energy.The lower the energy,the greater the difference of CT value.Arterial phase single energy 40?100keV.vein 60?140keV between the two groups of CT values were statistically significant(P<0.05).The CT value of the benign group was lower than that of the malignant group.The CT value of the benign group was significantly higher than that of the malignant group.In the benign group,the slope of the energy spectrum curve and the iodine content in the arterial phase were(1.43±0.58),(11.82±4.89)100g/L,(2.43±0.29).(20.71.90)and 100g/L in the malignant group.respectively.The difference between the two groups was statistically significant(P<0.05).The water content in the arterial phase was(1030.37±10.56),(1025.96±6.10)100g/L,(1032.34±11.44),(1029.06±6.25)100g/L in the two groups of venous phase.The slope of iodine content and energy spectrum curve in benign group were(17.78±5.75),(1.98±0.65),(15.85±2.02)and(2.0±0.38)100g/L in malignant group.The difference was not statistically significant(P>0.05).The detection rates of benign and malignant tumors were 86.67%(26/30)and 85.71%(24/28),respectively,and the sensitivity was 86.67%,80.00%and 80.00%,respectively,and the specificity was 80.00%and 86.67%respectively.Pair design ?2 test display there was no statistically significant difference between the CT and CT(x2=30.39,P<0.0001),and Kappa test showed a good association between the results of energy spectrum CT and pathologic findings(Kappa value was 0.72).Conclusion:The CT value,slope of the energy spectrum curve and iodine content in the arterial phase of the binign group were significant lower than those of' the malignant group,and a good association was observed between the results of energy spectrum CT and pathologic findings.Indicating that the multi-parameter analysis of energy spectrum CT can be utilized for the differentiation between primary benign and malignant ovarian tumors.