Iterative Reconstruction for Cardiopulmonary Computed Tomography

Abstract

Introduction

The number of computed tomography (CT) examinations has increased rapidly since CT became commercially available. This resulted in growing concerns regarding the risk of malignancies induced by application of medical ionizing radiation. Therefore, radiation dose reduction strategies are needed and being developed. This is especially important for patients who get many CT scans (e.g. chest CT for follow up of lung nodules), CT scans with high radiation doses (e.g. scans of the heart) and for children who are more sensitive to radiation. A recent dose reduction strategy focuses on optimization of image reconstruction. With current reconstruction methods, radiation dose reduction results in noisy images, which are susceptible to artefacts. Iterative reconstruction (IR) is an alternative image reconstruction method that allows for image quality improvement at similar radiation dose or imaging at lower radiation dose without compromising image quality compared to current methods. Improved computational power of CT reconstruction workstations has permitted the recent introduction of IR for CT.

Purpose

The aim of this thesis was to evaluate both potential applications of IR: improving image quality and reducing radiation dose of CT scans of the heart and the lungs. First, in-vitro and ex-vivo preclinical research was performed to globally assess the potentials of IR for both applications. Second, image acquisition of cardiopulmonary CT examinations was optimized in-vivo with IR. Finally, IR was applied to reduce the radiation dose in the clinical setting and explore novel indications for CT.

Results

With the current thesis, we showed that IR can be used to improve image quality in routine dose CT, to reduce the radiation dose without loss of image quality, or with a combination of both strategies. We conclude that radiation doses for CCTA, calcium scoring CT, contrast-enhanced chest CT and plain chest CT can be reduced substantially by using IR. For calcium scoring CT scans of the heart, we first performed a phantom study, then an ex-vivo study and finally a pilot study in 30 individuals evaluating the maximally achievable radiation dose reduction. Based on these studies we found a potential radiation dose reduction of 80%. Reducing the radiation dose of CT imaging also allows for broadening the diagnostic value of CT scans of the lungs. If the radiation dose issue would be resolved, CT imaging could replace plain radiography for certain indications to reduce misdiagnosis.

Conclusion

Radiation doses for CT scans of the heart and the lungs can be reduced substantially by using IR.