Spectral Analysis And Morphological Characteristics In The Differentiation Of Solitary Pulmonary Nodules: Preliminary Study

Part One The Value of Spectral CT Parameters in theDifferential Diagnosis of Solitary Pulmonary NodulesBackground and Purpose:To investigating the value of spectral imaging parameters in differentiatingmalignant solitary pulmonary nodule (SPN) from benign.Materials and Methods:Eighty five patients suspected with lung space-occupying lesions were includedand underwent CT spectral imaging during arterial phase (AP) and venous phase (VP)(Discovery HD750, GE healthcare).Record the CT value of lesions in40keV in monoenergetic images of AP andVP. Measuring the Iodine concentrations(IC) in lesions and aorta in the same levelfrom iodine-based material-decomposition material density CT images, thencalculating the normalized iodine concentration (NIC), the calculation formula is: NICles=ICles/ICao, which IClesand ICaorepresent the iodine content of the lesion andthe same layer of the aorta. The independent sample t test and the Kruskal-WallisTest were performed to compare quantitative parameters and subjective scores.Sensitivity and specificity were calculated according to the receiver operatingcharacteristic curve (ROC).Results:1. NICs, CT values in40keV in patients with malignant SPN were significantlyhigher than those in patients with benign SPN (t=6.011,3.124,4.603,2.874,P<0.001,=0.003,<0.001,=0.007). Mean NICs were22.21±1.16versus12.80±0.84and48.72±2.45versus32.41±3.02during the AP and VP.2. The area under ROC of NIC in AP is biggest (0.930,p＜0.001), when NICAP≥15.915was set as diagnostic threshold, the sensitivity and specificity were89.7%and80%. The area under ROC of NIC in VP is0.890(p＜0.001), when NICvp≥35.76was set as diagnostic threshold, the sensitivity and specificity were97%and75%respectively.Conclusion:The spectrum parameters can be used to identified benign and malignantpulmonary nodules. The standardized iodine concentration (NIC) in arterial andvenous phase had a high identification value in differentiating malignant nodulesfrom benign nodules. Part Two The Spectral Analysis and MorphologicalCharacteristics Valuation in the Differential Diagnosis ofSolitary Pulmonary NodulesBackground and Purpose:To investigate the value of spectral imaging parameters and morphologicalcharacteristics in differentiating malignant solitary pulmonary nodule (SPN) frombenign.Materials and Methods:Eighty five patients suspected with lung space-occupying lesions were includedand underwent CT spectral imaging during arterial phase (AP) and venous phase (VP)(Discovery HD750, GE healthcare).Record the CT value of lesions in40keV in monoenergetic images of AP andVP. Measuring the Iodine concentrations (IC) in lesions and aorta in the same levelfrom iodine-based material-decomposition material density CT images, thencalculating the normalized iodine concentration (NIC), the calculation formula is:NICles=ICles/ICao.Two experienced readers independently analyzed the image datasets accordingto a standard5-score method (based on the morphological characteristics of SPN):certain malignant (5),may be malignant (4),not sure (3),may be benign (2),certainbenign (1). Differentiate the malignant SPN from benign based on the use ofsubjective rating combination with the spectral parameters, according to the principleof consistency. When the subjective rating is3points, the conclusion is drawnaccording to the spectral parameters. The malignant SPNs`energy spectrumparameters is greater than the threshold, otherwise is benign. The independent samplet test and the Kruskal-Wallis Test were performed to compare quantitative parametersand subjective scores. Sensitivity and specificity were calculated according to thereceiver operating characteristic curve (ROC). Results:1. NICs, CT values in40keV in patients with malignant SPN were significantlyhigher than those in patients with benign SPN (t=6.011,3.124,4.603,2.874,P<0.001,=0.003,<0.001,=0.007). Mean NICs were22.21±1.16versus12.80±0.84and48.72±2.45versus32.41±3.02during the AP and VP. The subjectivescores of malignant is higher than the benign(4.22±0.11VS2.08±0.20,Z=5.793,P<0.001).2. The area under ROC of NIC in AP is biggest(0.930,P＜0.001), when NICAP≥15.915was set as diagnostic threshold, the sensitivity and specificity were89.7%and80%.3. Combining the subjective rating with NICAP≥15.915as diagnostic threshold, thesensitivity, specificity, accuracy, positive predictive value (PPV) and negativepredictive value (NPV) of diagnosis of SPN were98.14%,90.32%,95.29%,94.64%and96.55%.Conclusion:Use of spectral analysis and morphological characteristics alone candifferentiating the malignant from benign SPN, the combination use of both methodscan improve the efficiency of qualitative diagnosis of SPN. Dual energy spectral CT(DEsCT) provide spectrum and morphological information simultaneously andimprove the efficiency of qualitative diagnosis of SPN.