Four-Minute Bone SPECT Using Large-Field Cadmium-Zinc-Telluride Camera

Abstract

Purpose: The aim of this study was to determine the minimum acquisition time without decreasing lesion detectability of bone SPECT using a whole-body cadmium-zinc-telluride camera.

Methods: Patients referred for bone SPECT were retrospectively included. SPECT of 30 patients were reframed from native data (16 s/projection) to produce 10-, 5-, and 3-s/projection data sets. A "critical" acquisition time/projection was defined as that below which the SPECT quality becomes insufficient for interpretation, as determined by 3 reviewers using a 4-point scale (0 = quality insufficient for interpretation, 1 = average, 2 = good, 3 = excellent). Three reviewers (blinded to the acquisition time) evaluated SPECT data sets (n = 79), native and reframed with "critical" acquisition times, in a randomized order. A lesion was defined as any uptake considered pathological by a reviewer. Lesion detectability equivalence between native SPECT and reframed SPECT was assessed by calculating a coefficient (κ) for each reviewer.

Results: Image quality of the first sample (n = 30) was significantly and progressively less well graded for the reframed data sets by all reviewers. Only 1 patient was graded 0 by each reviewer for the 5-s/projection data set. For the 3-s/projection data set, 3 patients were graded 0. No patients were graded 0 for 10-s/projection data set. The minimal acquisition time, for each projection, was defined as 5 s/projection. The coefficient κ, between native and reframed, with critical acquisition time/projection SPECT was greater than 0.9 for each reviewer.

Conclusions: The more contrasted images of the cadmium-zinc-telluride camera allow performance of 5-s/projection SPECT without loss of lesion detectability. This suggests the possibility of performing whole-body SPECT in a reasonable time or reducing injected doses, especially in pediatric patients.