The Differential Diagnosis Between Benign And Malignant Cystictumors Of Pancreas By Spectral CT Imaging

Objectives: By differentiating the status of cystic tumors of pancreas (benign ormalignant) with Gemstone Spectral Imaging (GSI) technique and comparing thediagnosis outcome with postoperative pathological results, this study aims to evaluatethe application of Spectral CT imaging in the diagnosisbetween benign and malignant cystic tumors of pancreas and improve thepreoperative diagnosis of this disease. The results will provide valuable and reliableinformation for preoperative diagnosis, which is in favour of appropriate surgicaldecision.Study subjects:The gemstone CT (Discovery CT750HDCT，GE-Healthcare) wasconducted on42patients (26males and16females) with cystic tumors of pancreas.All patients were confirmed by surgical and histological biopsy examinations. Thescanning mode was plain scanning plus triple phase-enhanced scanning and GSIscanning was used in arterial phase.Materials and methods:Before examination, all patients are required to practicesuffocating to reduce the artifacts.88ml of contract agent (Omnipaque300)followed by10ml of saline were injected intravenously into middle elbow of patient ata rate of3.0ml/s. The arterial phase scanning and the portal phase scanning wereperformed for30s and60s after injection, respectively, together with a120s delay forbalance phase. The abdominal aortic CT value (Smart Prep technique) was applied toinitiate the arterial phase scanning, with monitoring threshold of100HU. The routinescanning procedure was carried out,30s after the arterial phase scanning is the portalphase scanning, followed by balance phase scanning after another60s. The raw datawere sent to adw4.5workshop for procession and analysis, with focus on the imagingfeatures of pancreatic lesions by plain and triple-phase enhanced scanning. Theassessments in differentiation of benign and malignant cystic tumors of pancreas withtraditional CT examination include:1). the density of tumor lesion in plain scanning(CT value);2). tumor size and morphology;3). tumor boundaries and surroundings;4). the existence of tumor calcification, bleeding or necrosis;5). tumor enhancementlevel in enhanced scanning;6). the pancreatic duct dilatation. The tumor spectrum attenuation curve was obtained using GSI software in arterial phase scanning. Bycomparison with normal pancreatic tissue, it can be concluded that the tumor isbenign or malignant.Results: There was statistical difference between the conclusion made fromtraditional method based on CT values and imaging features of several phasesscanning and the conclusion made by pathological examination.(P=0.025<0.05).However, there is no statistical difference regarding the conclusion made fromspectral CT imaging and pathological examination.(P=0.316>0.05).Conculsion:In arterial phase scanning, the spectrum attenuation curve of tumor lesiongot from GSI software can be applied for the differentiation of benign andmalignant cystic tumors of pancreas, by comparison with normal tissue. Theconsistency of the conclusion from spectral CT imaging and from postoperativepathological examination is higher than traditional CT scan, indicating that this newtechnique can provide more accurate information for diagnosis and help doctor makemore effective therapeutic strategy.