[Image analysis in multislice spiral CT of the lung with MPR and MIP reconstructions]

Radiologe. 1999 Nov;39(11):952-7. doi: 10.1007/s001170050587.
[Article in German]

Abstract

Purpose: To test, whether axial, coronal and sagittal MIP and MPR reconstructions of diagnostic quality can be obtained from 1-mm collimation MSCT data of the chest for the evaluation of thoracic anatomy and pathology.

Materials and methods: 1-mm collimation MSCT scans were obtained with a pitch of 6 in an acrylic phantom and in 20 patients. Axial images were reconstructed with 0.6-mm increment. Multiplanar reformations (MPRs) and sliding thin-slab maximum intensity projections (STS-MIPs) were reconstructed in axial, coronal and sagittal planes. Images were printed in lung windows and evaluated by three readers by using a standardized evaluation scheme.

Results: Overall, both methods allowed good visualization of anatomic structures. MIP was superior for visualization of the pulmonary arteries (p < 0.05) while central and peripheral bronchi and the lung parenchyma were better depicted on multiplanar reconstructions. A confident diagnosis of thoracic pathology was feasible using both modalities, however MIPs appeared less usefull for evaluation of gross parenchymal abnormalities, such as pneumonic infiltrates or fibrotic changes. No significant difference in the degree of motion artifacts were detected between both modalities.

Conclusion: MSCT data sets are ideally suited for generating MPR and MIP reconstructions. While MIPs are superior for the evaluation of thoracic vessels, MPR is advantageous for visualizing central and peripheral bronchi and the pulmonary parenchyma.

MeSH terms

  • Humans
  • Image Processing, Computer-Assisted / methods*
  • Image Processing, Computer-Assisted / statistics & numerical data
  • Logistic Models
  • Lung / diagnostic imaging*
  • Lung Diseases / diagnostic imaging
  • Phantoms, Imaging / statistics & numerical data
  • Radiography, Thoracic / methods
  • Radiography, Thoracic / statistics & numerical data
  • Tomography, X-Ray Computed / methods*
  • Tomography, X-Ray Computed / statistics & numerical data