64 Row Spiral Ct Perfusion Imaging In The Diagnosis Of Lung Cancer, Pneumonia And Tuberculosis Differential Value

Objective The purpose of this study was to determine the utility and feasibilityof64-slice spiral CT perfusion imaging in the differentiation of lung cancer,pneumonia and pulmonary tuberculosis.Materials and methods A retrospective analysis of our database of35patients,who had undergone64-slice spiral CT perfusion imaging, was performed.22patients with lung cancer,7patients with pulmonary tuberculosis and6patients withpneumonia, who were all pathologically verified, were examined on SiemensSOMATOM64-slice spiral CT for perfusion scan after a bolus injection of50mlcontrast at a rate of5ml/s. All the slice data acquired from the scan were processedusing syngo Body PCT software under the “Lung Tumor” option. Perfusionparameters including time-density curve (TDC), blood flow (BF), blood volume (BV)and permeability (PM) were recorded, and their correlation determined.Results (1) Lung cancer, pneumonia and pulmonary tuberculosis had differentTDC patterns. The TDC of lung cancer was categorized into four types: A, B, C and D.Type C predominated on patients with lung cancer (36%). Its curve ascendedgradually to peak with almost no subsequent descending thus sustaining at a higherlevel. The TDC of pulmonary tuberculosis could be divided into three types: A, B andC. Type A was the main one on patients with pulmonary tuberculosis (42%), showinga slow and flat pattern for both ascending and descending with the enhancingmagnitude less than10HU. The TDC of pneumonia can only be grouped into twotypes: A and B. Type A was the most commonly observed on pneumonia patients(67%) with the curve similar to that of type C in lung cancer while its enhancingextent was slightly higher.(2) The difference in BV was not statistically significantfor lung cancer, pneumonia and pulmonary tuberculosis (P>0.05). However, the PMand BF were significantly different between lung cancer and pneumonia or pulmonarytuberculosis (P<0.05). The PM and BF in lung cancer were significantly greater thanthose in pneumonia and pulmonary tuberculosis whereas no significant difference wasnoted between pneumonia and pulmonary tuberculosis (P>0.05).(3) The maximum Youden index was used to determine the optimum thresholds of all the parameters inROC analyses. Statistical analysis demonstrated a threshold value of16.36for PM,and11.30for BF to provide sensitivity and specificity of77.3%and100%, and72.7%and100%, respectively, in the differential diagnosis of lung cancer and pulmonarytuberculosis. A threshold value of13.99for PM, and14.82for BF provided sensitivityand specificity of81.8%and100%, and68.2%and100%, respectively, fordetermining lung cancer or pneumonia.Conclusion TDC and perfusion parameters derived from64-slice spiral CTperfusion imaging can be used to evaluate hemodynamic changes in lung lesions ofdifferent pathological types and a combination of the hemodynamic measurements ishelpful in the differentiation of lung cancer from pneumonia and pulmonarytuberculosis.