Quantitative analysis of xQuant reconstruction algorithm in SPECT/CT (original) (raw)

Radiation Physics and Chemistry, 2021

Abstract

Abstract Standard iterative Ordered Subset Expectation Maximisation (OSEM) reconstruction is well established in SPECT/CT, but despite its wide applications in image processing it comes with limitations in image noise and quality. A novel algorithm, xQuant (developed by Siemens Healthcare), uses Ordered Subset Conjugate Gradient Maximisation (OSCGM) which enables image quantification assessment such as standardised uptake value (SUV) measurements for reliable disease detection and evaluation of therapy response. As such, xQuant allows for dosimetry measurements, staging and management of diseases, analogous to the PET/CT modality for staging cancers and chemotherapy management. This study compares the accuracy of xQuant algorithm and current OSEM, by analysing image noise, SUV quantification and varying image reconstruction parameters. Standard clinical phantoms are used for comparison of both reconstruction algorithms: xQuant SUV accuracy with various Tc99m activity and varying scan times are performed for analysis. Results indicate that SUV measurements from xQuant are similar to expected SUV, regardless of selected reconstruction parameters, varying radiation activity and delayed scan times. Image noise assessment has shown that xQuant has lesser value of coefficient of variation (CoV) compared to standard OSEM, indicating xQuant's superior noise suppression without compromising image quality. The quantifiable superiority of xQuant reconstruction algorithm supersedes the basic iterative 3D OSEM reconstruction. It shows higher resolution qualitative assessment and provides consistent quantitative analysis. The reliability and superiority of xQuant enables clinical SPECT/CT quantification to detect disease and improve therapy management.

Teik Hin Tan hasn't uploaded this paper.

Let Teik Hin know you want this paper to be uploaded.

Ask for this paper to be uploaded.