Comparison Study Of Diagnosis Of Kidney Tumor Among Color Doppler Ultrasound, Computed Tomography And Image Fusion Of US And CT

ObjectiveKidney neoplasm mainly includes renal cell carcinoma, nephroblastoma, tumor of renal pelvis, metastatic carcinoma, renal oncocytoma, angiomyolipoma and so on. Renal malignant tumors, as one of the most common primary malignant tumors, consist of the main part of kidney tumor, about86%. Its incidence trends to rise year by year in the world. The most common renal benign tumor is angiomyolipoma. However, renal tumors contain a variety of types and lack of typical clinical manifestation, which means difficult to differentiate malignant from benign tumors. An early, accurate diagnosis is of much importance to the clinical treatment and prognosis.Ultrasound, the preferred imageological examination of renal tumors, has shown to supply useful information for finding and diagnosing early renal tumors. It is reported that over2/3early kidney tumors were found by ultrasound occasionally. Because of the cross-over in ultrasonography and low positive rate of low velocity blood flow by Color Doppler, ultrasound cannot efficiently differentiate malignant from benign tumors.CT (Computed Tomography), the golden standard of imageological examination of renal tumors, has shown obvious advantages in evaluating the degree of metastasis of retroperitoneal lymph node, invasion in perirenal lymph node, cancer embolus in renal vein and inferior vena cava. But it tends to mix up malignant with benign tumors, if the tumor diameter is less than3cm or hemorrhage necrosis, cystic change occur inside the lesion.Image fusion is an emerging imageological examination technology. During real-time US, the ultrasound machine generates multiplanar reconstruction images from the CT volume data, corresponding to the US images of the same cross-section side-by-side.In this study, a comparison of diagnosis of renal tumors among ultrasound, CT and image fusion has been made to analyze the differences between malignant and benign tumors.Material and methods1. Patients dataThe present study included54patients (38men and16women, age range:12to80, mean age:56.37±17.23), who had renal lesions from March,2013to February,2014. all patients were evaluated by ultrasound, CT and image fusion of ultrasound and CT.2. Methods2.1UltrasoundThe ultrasound equipment used in this study was GE LOGIQ E9, equipped with a3.5-4.0MHz convex array probe. All patients had US examination to evaluate the size, shape and structure of kidneys. Record the location, number, size, shape, border, hemorrhage and echo of lesions. Therefore the primary diagnosis was made.2.2CTThe CT inspection in this study used GE High speed16row spiral CT and GE High speed64row spiral CT. All patients were accessed to plain scan and enhanced scan (5mm collimation and slice). Contrasted medium were injected i.v (1ml/kg,3-4ml/s,20s) and3phrases were taken.2.3Image Fusion of US and CTIn this study, real-time virtual US (RVS) system comprised an ultrasound scanner (GE LOGIQ E9), a magnetic probe motion-tracking device and a personal computer. Firstly, the acquired CT volume data were loaded onto the US system. During real-time US, the ultrasound equipment generated reconstruction images from the CT volume data, corresponding to the US images of the same cross-section side-by-side. To align the RVS image and the conventional US image, renal artery and renal vein were used as fiducial landmarks. After the US image and RVS image were overlaid, distance of the same anatomic structure in different images was supposed to be measured. If the distance was less than5mm, image fusion was successfully done.3. Statistical analysisAll data acquired in this study was saved as a database in excel. SPSS17.0statistic software was used. The results of image fusion of US and CT were compared with that of US and CT. Chi-square test should be applied in count data. Taking P<0.05for the difference was statistically significant.Results1. Pathology TypesAmong34renal malignant tumors, there were30renal clear cell carcinomas (88.2%),2papillary renal carcinomas (5.9%),1chromophobe cell tumor (2.9%) and1granular cell carcinoma of kidney (2.9%). Among20renal benign tumors, there were17angioleiomyolipoma (85.0%) and3nephradenoma (15%).2. Ultrasound ExaminationAmong54renal lesions, there were53found by US. The positive rate of US was98.1%. Compared with postoperative pathological diagnosis, the accuracy of US diagnosis was81.5%(44/54).10patients were misdiagnosed by US. The accuracy of US in malignant tumors was85.3%(29/34), in benign tumors was75.0%(15/20).3. CTIn CT plain scan of54renal lesions,24lesions were low density (44.4%),7lesions were appeared equidensity (13.0%),3cases were slightly high density (13.0%),16nodules were mixture density (29.6%). In contrasted enhanced CT scan, there were19nodules appeared hyper echoic (35.2%),15tumors isoehoic (27.8%),19appeared hypotonic (35.2%) and1no enhancement (1.8%). Compared with postoperative pathological diagnosis, the accuracy of CT diagnosis was85.2%(46/54).8patients were misdiagnosed by CT. The accuracy of CT in malignant tumors was85.3%(29/34), in benign tumors was85.0%(17/20).4. Image Fusion of US and CT6lesions that were not showed in ultrasound could be located accurately in real-time ultrasound system. Compared with postoperative pathological diagnosis, the accuracy of image fusion diagnosis was96.3%(52/54). The accuracy of image fusion in malignant tumors was97.1%(33/34), in benign tumors was95.0%(19/20).5. Comparison of Sensitivity and Specificity among US, CT and image fusion of US and CTAmong54renal lesions, there were53found by US. The positive rate of US was98.1%. There were54found by CT. The positive rate of CT was100%. It is the same with image fusion of US and CT. The sensitivity of US in renal malignant tumors was85.3%(29/34), the specificity is75.0%(15/20). The sensitivity of CT in renal malignant tumors was85.3%(29/34), the specificity is85.0%(17/20). The sensitivity of image fusion of US and CT in renal malignant tumors was97.1%(33/34), the specificity is95.0%(19/20).ConclusionThe image fusion is a new technology that allows the fusion of real-time US and preoperative CT data. It displays the synchronised pictures of both real-time US and CT in the same section of the body simultaneously. This study aims to compare the diagnostic value among US, CT and image fusion of US and CT, from which we can draw a conclusion that image fusion technology of US and CT can improve diagnostics efficiently.