Effect of Increased Image Matrix on Lung Nodule Volumetry in Chest CT: A Phantom Study

Background/Aim: This phantom-based study evaluates the impact of increased computed tomography (CT) image matrix resolution on lung nodule segmentation and volumetry.

Materials and Methods: Using a SOMATOM x.ceed scanner, standardized phantom nodules of varying sizes (3-12 mm) and densities were assessed across matrix sizes (256-1,024) with both soft and sharp kernels.

Results: Higher matrix resolutions (768 and 1,024) significantly improved volumetric accuracy (p<0.05), while 512 showed no significant benefit over 256 (p=0.128). Accuracy increased with nodule size (p=0.029) and was greater with sharp kernel reconstructions (p<0.001). Point-and-click segmentation outperformed drag-based methods (p<0.001). Very low-density nodules (−800 HU) were not segmentable.

Conclusion: Enhanced matrix resolution and optimized reconstruction and interaction methods improve the precision of CT-based lung nodule analysis, with possible implications for early lung cancer detection in CT.