Simulation and Optimisation of Total Knee Replacement Imaging with Positron Emission Tomography/ Computed Tomography

Abstract

Abstract Rationale: Total knee replacement (TKR) surgeries are crucial for osteoarthritis management but pose challenges in assessing bone metabolism in positron emission tomography/computed tomography PET/CT due to metal artefacts. Addressing these challenges will improve the early diagnosis of TKR-related complications and refine treatment strategies. This PhD project aims to explore novel metal artefact correction methods to enhance TKR PET/CT image quality and bone turnover measurement precision in periprosthetic regions. Methods: The project adopts multiple approaches, starting with a systematic review of quantitative PET methods to assess their efficacy in measuring bone metabolism. Subsequently, the development of population-based TKR phantoms using 3D printing to create realistic imaging scenarios for experimental purposes. Additionally, testing iterative metal artefact reduction algorithm (iMAR) performance in reducing TKR CT image metal artefacts. Finally, compare the performance of the maximum likelihood activity and attenuation (MLAA) reconstruction algorithm with CT-based attenuation correction methods using time-of-flight TOF reconstruction combined with the iMAR algorithm to improve PET image quality and standardised uptake value (SUV) precision. Results: The systematic review shows high correlations between SUV, Hawkins, and Patlak methods for measuring bone metabolism. The use of 3D-printed TKR phantoms shows their potential to simulate human TKR, enabling comprehensive in vitro testing of metal artefact reduction methods applied to TKR PET/CT scans. The iMAR results highlight its effectiveness with varied image kernel sizes in improving CT image quality, especially when applied with a moderate image kernel size. A comparison of the attenuation correction maps shows that combining TOF with iMAR improves TKR PET/CT image quality and SUV measurement, while using the MLAA reconstruction improves the image quality but significantly underestimates SUV, especially with low counts scan. Conclusions: This project has taken significant steps in addressing metal artefact effects on TKR PET/CT scans, highlighting the need for further research to minimise these impacts.