Analytical technique to recover the third dimension in planar imaging of inhaled aerosols: (1) impact on spatial quantification

Abstract

An analytical algorithm is described for converting planar scintigraphic images of aerosol distributions in the lungs to an equivalent three-dimensional (3D) representation. The recovery of volumetric information should benefit regional quantification. The technique has been validated using simulated planar images of eleven known aerosol distributions in ten realistic lungs. Global and regional 3D parameters, such as the total activity deposition (A), the penetration index (PI) and the relative penetration index (rPI), were quantified on the planar images and on their 3D representation. Random and systematic errors of the estimation were measured. Finally, the performance of planar imaging was compared with that of single-photon emission computed tomography (SPECT). SPECT images were simulated for the same aerosol distributions in the same subjects and quantified for A, PI, and rPI. The systematic errors in A, PI and rPI obtained from planar imaging were 8.9%, 64.8%, and 54.1%, respectively, using the two-dimensional (2D) analysis; they improved significantly to 4.4%, 19.0%, and 25.5% with the 3D analysis (p < 0.01). The corresponding values for SPECT were 5.2%, 9.8%, and 15.7%, significantly better for PI and rPI (p < 0.01). The random errors of A were similar for all techniques being about 5%; those of PI and rPI measurements were significantly higher for planar imaging (<or=14%) than SPECT (<or=8%). In conclusion, 3D spatial parameters can be derived from planar imaging that are significantly more accurate in characterizing different aerosol depositions than their 2D counterpart. However, the errors remain significantly higher than with SPECT.