Experimental evaluation of absolute quantification in 99m Tc-TRODAT-1 SPECT/CT brain dopamine transporter (DAT) studies

Abstract

Objective: To evaluate the quantitative accuracy of clinical brain dopamine transporters (DAT) investigations utilizing ^99m Tc-TRODAT-1 single-photon emission computed tomography (SPECT)/computed tomography (CT) in experimental and clinical settings.

Materials and methods: The study used an experimental phantom evaluation and a clinical dataset. Three-dimensional-ordered subsets expectation-maximization reconstructed the original and resampled datasets using attenuation correction, scatter correction, and resolution recovery. The reconstructed data were analyzed and reported as percentage difference, standardized uptake value reference (SUVr), and a coefficient of variation (CoV). The Taguchi method tested the impact of the three different parameters on signal-to-noise ratio (SNR) and SUVr, including number iteration, Poisson resampling, and phantom setup, with and without the plaster of Paris (POP). Six ^99m Tc-TRODAT-1 SPECT/CT scans were acquired in healthy subjects for verification purposes.

Results: The percentage activity difference between the phantom with and without POP is 20% and 5%, respectively. The SUVr reveals a 10% underestimate for both with and without POP. When it comes to the influence of Poisson resampling, the SUVr value for 75% Poisson resampling indicates 10% underestimation on both sides of the caudate and putamen area, with and without POP. When 25% of Poisson resampling is applied, the SUVr value is overestimated (±35%). In the Taguchi analysis, iteration numbers were the most dominant factor with the F-value of 9.41 and the contribution rate of 52.66% (p < 0.05) for SNR. In comparison, F-value of 9.1 for Poisson resampled with contribution rate of 58.91% (p < 0.05) for SUVr. Reducing counts by 25% from the original dataset resulted in a minimal bias in SUVr, compared to 50% and 75%.

Conclusion: The optimal absolute SPECT/CT quantification of brain DAT studies using ^99m Tc-TRODAT-1 appears achievable with at least 4i10s and SUVr as the surrogate parameter. In clinical investigations, it is possible to reduce the recommended administered dose by up to 25% while maintaining accurate measurement.

Keywords: 99mTc-TRODAT-1; DAT; SPECT/CT; SUVr; quantification.

FIGURE 1

Single‐photon emission computed tomography (SPECT)/computed tomography (CT) images for the original data (100%) with various levels of Poisson resampling (75, 50, and 25%): (a) final reconstructed images with 1i10s, (b) 2i10s, (c) 4i10s, (d) 6i10s, (e) 8i10s, (f) 10i10s, and (g) 20i10s

FIGURE 2

Percentage of activity difference for (a) caudate region, (b) putamen region, and (c) background region for phantom images dataset

FIGURE 3

Effect graph for percentage of activity difference in standard data reconstruction and Poisson resampling application for (a) right caudate region, (b) left caudate region, (c) right putamen region, (d) left putamen region, and (e) occipital

FIGURE 4

SUVr in original dataset and Poisson resampling dataset for (a) right caudate region, (b) left caudate region, (c) right putamen region, and (d) left putamen region for phantom dataset

FIGURE 5

CoV for (a) caudate region, (b) putamen, and (c) occipital for phantom dataset

FIGURE 6

^99mTc‐TRODAT‐1 computed tomography (SPECT) overlays on magnetic resonance imaging (MRI) images for various datasets from one of the study subjects. The SPECT images were reconstructed using 4i10s with Butterworth post‐filtering (cutoff value = 0.4 cm⁻¹, power value = 10) for better image display.

FIGURE 7

Percentage of mean SUVr difference for different Poisson resampling levels compared with the original dataset for (a) caudate, (b) putamen, and (c) striatum in six normal volunteers