Edge-preserving denoising for intra-operative cone beam CT in endovascular aneurysm repair. (March 2017)

• Record Type: Journal Article
• Authors: Liu, Yi; Castro, Miguel; Lederlin, Mathieu; Shu, Huazhong; Kaladji, Adrien; Haigron, Pascal
• Is Part Of: Computerized medical imaging and graphics. Volume 56(2017)
• Page Start: 49
• Page End: 59
• Publication Date: 2017-03
• Subjects: Endovascular aneurysm repair -- Intra-operative CBCT -- Image denoising -- Penalized weighted least-squares -- Edge-preserving penalty; Diagnostic imaging -- Periodicals; Imaging systems in medicine -- Periodicals; Diagnosis, Radioscopic -- Data processing -- Periodicals; Diagnostic Imaging -- Periodicals; Imagerie pour le diagnostic -- Périodiques; Diagnostic imaging; Periodicals; Electronic journals; Electronic journals; 616.0754
• DOI: 10.1016/j.compmedimag.2017.01.004
• Languages: English
• ISSNs: 0895-6111

Abstract

Highlights: We propose an edge-preserving algorithm for denoising intra-operative low-dose CBCT images in endovascular aneurysm repair. The edge-preserving penalty based PWLS method relies on bilateral filter to reduce noise while improving image contrast. Experimental comparisons show that the proposed penalty outperforms the isotropic, anisotropic, TV and Huber penalties. Abstract: C-arm cone-beam computed tomography (CBCT) acquisition during endovascular aneurysm repair (EVAR) is an emergent technology with more and more applications. It offers real time imaging with a stationary patient and provides 3-D information to achieve guidance of intervention. However, there is growing concern on the overall radiation doses delivered to patients all along the endovascular management due to pre-, intra-, and post-operative X-ray imaging. Manufactures may have their low dose protocols to realize reduction of radiation dose, but CBCT with a low dose protocol has too many artifacts, particularly streak artifacts, and decreased contrast-to-noise ratio (CNR). To reduce noise and artifacts, a penalized weighted least-squares (PWLS) algorithm with an edge-preserving penalty is proposed. The proposed method is evaluated by quantitative parameters including a defined signal-to-noise ratio (SNR), CNR, and modulation transfer function (MTF) on clinical CBCT. Comparisons with PWLS algorithms with isotropic, TV, Huber, anisotropic penalties demonstrate that the proposed edge-preserving penalty performs well not only on edge preservation, but also on streak artifacts suppression, which may be crucial for observing guidewire and stentgraft in EVAR.