The impact of different reconstruction parameters on quantitative ^99mTc-DPD SPECT/CT values in the assessment of cardiac transthyretin amyloidosis

To assess in a phantom and in a clinical study the influence of different reconstruction parameters on quantitative SPECT/CT values in the assessment of cardiac transthyretin amyloidosis (ATTR-CA).

A hybrid SPECT/CT camera with a proprietary software for SPECT/CT-based quantification of myocardial uptake of ^99mTc-DPD was used. Images were reconstructed with 6 different protocols, differing in iterations, subset and Gaussian filter. Quantitative values were tested in phantom and clinical studies across different reconstruction protocols. Values were automatically calculated both as kBq/ml and as maximum, mean and peak standardized uptake value (SUV).

The standard parameters provided by the manufacturer (reconstruction 1) yielded higher accuracy in quantifying, with measuring 97.1% of the expected activity in the phantom. Reconstructions with higher Gaussian filter caused a systematic underestimation of quantified values of 27.2% (p < 0.01). Results were replicated in the clinical study consisting of 155 patients with suspected ATTR-CA, wherein changing the number of iterations and subsets was not associated with a statistically significant difference in quantitative values compared to reconstruction 1, while a higher Gaussian filter caused inaccurate quantification with up to 24% of difference measured.

Different reconstruction parameters can impact quantitative values on ^99mTc-DPD SPECT/CT. Therefore, parameters should be maintained consistently across different acquisitions and different centres.