Pulmonary Nodule Evaluation With MDCT: The Factors Influencing Volume Quantification And Diagnostic Efficiency By Threshold Segmentation

ObjectiveTo evaluate in a phantom the effects of reconstruction kernel and section thickness on volumetric measurement of pulmonary nodule with different nodule size and density using multi-detector CT(MDCT). Test the reproducibility of volume measurements of pulmonary nodules. And to evaluate solitary pulmonary nodules(SPN) using dynamic contrast enhanced multi-slice spiral computed tomography(MDCT). Preliminary asses the value in distinguishing between malignant and benign/inflammatory nodule, compared with the PH method.Materials and MethodsThis phantom is 25 cm in diameter, a total of 16 contained spherical synthetic nodules with four diameter categories and five attenuation categories were studied. The phantom was scanned twice at intervals of two weeks with 64-detector MDCT. Eight reconstruction kernels and five section thicknesses were performed to obtain data for assessing accuracy and the reproducibility of volumetric measurement. Absolute percentage error (APE) was calculated to assess the accuracy of volumetric measurement.68 patients with SPN verified by pathology and clinic were analyzed, including 40 malignant nodules,20 benign nodules and 8 inflammatory nodules. Preenhanced and 8-phase dynamic enhanced MDCT scans were performed. The image were applyed threshold-120-120Hu. we segmented each solitary pulmonary nodule by CT value using colored 3D ROI. Record the precontrast and enhanced constituent ratio of each segment threshold-voxel index(VI),the average CT value of the whole nodule. Calculate and assesse the change of the VI of each threshold and peak heigh(PH). Patterns of time-density curve (T-DC) were assessed.Statistical analysis was performed by using one-way ANOVA, multivariate analysis of variance, multiple linear regression analysis, paired-samples t test and ROC curves.Results1. There was statistically significant(P<0.05)difference in the APE across the eight reconstruction kernels with the 0.625 mm and 1.25 mm section thicknesses. The chest kernel yielding the minimal error, and the bone kernel followed.2. Reconstruction section thickness, nodule size and nodule attenuation significantly (P<0.01) affected the APE. Both the reconstruction section thickness and nodule diameter had linear correlation with APE. 3. There was no significant variations (P>0.05) in the volumes obtainded by chest kernel with the section thicknesses of 0.625 mm and 1.25 mm between the twice scans.4. PH of the malignant and inflammatory nodules were higher than those of the benign nodules, while no statistical difference of the PH was found between the malignant and inflammatory nodules.5. The average absolute VI change of three groups of SPN (malignant, benign, inflammatory) had significant difference in the eight phases of dynamic enhancement(p<0.001).6. The optimal scanning time for differentiate between the malignant and benign nodules was 60s after contrast injection; The optimal scanning time for differentiate between the malignant and inflammatory nodules was 300s after contrast injection.7. To differentiate between the malignant and benign nodules:The threshold of PH was 22 Hu (equal to or higher than 22 Hu were considered as malignant) and sensitivity would be 82.5%, specificity 75.0%, and accuracy 80.0%. The threshold of the average absolute VI change at 60s after contrast injection was 5.9%(equal to or higher than 5.9% were considered as malignant), the sensitivity would be 82.5%, specificity 75.0%, and accuracy 80.0%.8. To differentiate between the malignant and inflammatory nodules:The threshold of PH was 36 Hu (equal to or less than 36Hu were considered as malignant)and sensitivity would be 70.0%, specificity 62.5%, and accuracy 68.7%. the average absolute VI change at 60s after contrast injection was 11.9%(equal to or less than 11.9% were considered as malignant), the sensitivity would be 67.5%, specificity 100.0%, and accuracy 72.9%.9. According to the date of the malignant and benign nodules/the malignant and inflammatory nodules, the ROC curves of the PH and the average absolute VI change were drawn. The areas under the curves of the average absolute VI chang (0.844/0.845)were larger than the PH(0.797/0.736).Conclusion1. All kernels except the lung kernel can be used for volumetric measurement. For nodules less than 5 mm, the section thickness should be 0.625 mm or less; For nodules 5 mm or larger, a reconstruction section thickness of 1.25 mm is recommended. The varying of nodule diameter and/or density also significantly affects volumetric APE. The reproducibility of MDCT-based volumetric measurements of pulmonary nodules is reliable.2. Dynamic enhanced scanning and CT value threshold segmentation is a feasible method in differentiating malignant nodule from benign and inflammatory ones.3. The diagnostic efficacy of the average absoluteâ…¥change was higher than PH in differential diagnosis of SPN.4. To differentiate between the malignant and benign nodules:the threshold of PH was 22 Hu; the threshold of the average absoluteâ…¥change at 60s after contrast injection was 5.9%. To differentiate between the malignant and inflammatory nodules: the threshold of PH was 36 Hu; the threshold of the average absoluteâ…¥change at 300s after contrast injection was 11.9%.