[Paper Review] Multi-materials beam hardening artifacts correction for computed tomography (CT) based on X-ray spectrum estimation

Wei Zhao, Dengwang Li+4 more|arXiv (Cornell University)|Dec 6, 2018

Advanced X-ray and CT Imaging40 references4 citations

TL;DR

This paper proposes a fast, non-iterative beam hardening correction method for multi-material CT using X-ray spectrum estimation. By modeling polychromatic attenuation via reprojected template images and applying a spectrum-based correction term to raw projections or reconstructed images, the method effectively reduces cupping and streak artifacts while preserving Hounsfield Unit accuracy across numerical, phantom, and in vivo data.

ABSTRACT

Due to the energy-dependent nature of the attenuation coefficient and the polychromaticity of the X-ray source, beam hardening effect occurs when X-ray photons penetrate through an object, causing a nonlinear projection data. When a linear reconstruction algorithm, such as filtered backprojection, is applied to reconstruct the projection data, beam hardening artifacts which show as cupping and streaks are present in the CT image. The aim of this study was to develop a fast and accurate beam hardening correction method which can deal with beam hardening artifacts induced by multi-materials objects. Based on spectrum estimation, the nonlinear attenuation process of the X-ray projection was modeled by reprojecting a template image with the estimated polychromatic spectrum. The template images were obtained by segmenting the uncorrected into different components using a simple segmentation algorithm. Numerical simulations, experimental phantom data and animal data which were acquired on a modern diagnostic CT scanner (Discovery CT750 HD, GE Healthcare, WI, USA) and a modern C-Arm CT scanner (Artis Zee, Siemens Healthcare, Forchheim, Germany), respectively, were used to evaluate the proposed method. The results show the proposed method significantly reduced both cupping and streak artifacts, and successfully recovered the Hounsfield Units (HU) accuracy.

Motivation & Objective

• Address beam hardening artifacts in multi-material CT caused by polychromatic X-ray spectra and energy-dependent attenuation.
• Overcome limitations of linear reconstruction (e.g., FBP) that produce cupping and streak artifacts due to inconsistent data modeling.
• Develop a computationally efficient method that avoids iterative forward/backward projections common in existing correction techniques.
• Ensure robustness to uncertainties in attenuation coefficients and applicability across different CT systems and data types.
• Enable practical clinical deployment by minimizing computational overhead while maintaining high accuracy in Hounsfield Unit recovery.

Proposed method

• Estimate the polychromatic X-ray spectrum from projection data using a calibration-based approach.
• Segment the uncorrected CT image into material components (e.g., water, PMMA, aluminum) using a simple segmentation algorithm.
• Reproject segmented template images using the estimated spectrum to simulate polychromatic forward projection.
• Compute a correction term as the difference between monochromatic and polychromatic reprojection data, scaled appropriately.
• Apply the correction term to raw projection data (projection-domain) or reconstruct it as an artifact image (image-domain) for direct addition to the uncorrected image.
• Implement the method in both projection and image domains, enabling flexibility and compatibility with existing reconstruction pipelines.

Experimental results

Research questions

• RQ1 Can spectrum estimation enable accurate modeling of polychromatic attenuation in multi-material CT without iterative reconstruction?
• RQ2 How effective is the proposed method in reducing both first-order (cupping) and high-order (streaking) beam hardening artifacts?
• RQ3 To what extent is the method robust to errors in attenuation coefficient assignments and segmentation inaccuracies?
• RQ4 Can the method be efficiently implemented in real-time clinical settings using modern hardware acceleration?
• RQ5 How does the method perform across different CT scanners and data types (phantom, animal, clinical)?

Key findings

• The proposed method significantly reduced both cupping and streak artifacts in numerical simulations, phantom studies, and in vivo canine data acquired on a Discovery CT750 HD and Artis Zee C-Arm CT scanner.
• Hounsfield Unit (HU) accuracy was successfully recovered, with quantitative results showing improved HU consistency across different materials.
• The method demonstrated robustness to 10% deviations in attenuation coefficients from standard NIST values, indicating practical tolerance to measurement uncertainty.
• Despite the presence of scatter radiation in CBCT data, the method showed minimal degradation in image quality, likely due to effective scatter rejection by focused grids.
• The algorithm achieved a correction time of approximately 100 seconds for a 512×512×200 canine dataset using a single NVIDIA GeForce GTX 480 GPU, indicating feasibility for real-time applications.
• Segmentation inaccuracies (e.g., missing water-like materials) had minimal impact due to similar attenuation properties, and the method still fully removed artifacts in numerical phantoms.