Spectral diffusion: an algorithm for robust material decomposition of spectral CT data

Clark, Darin P.; Badea, Cristian T.

doi:10.1088/0031-9155/59/21/6445

Cited by 49 publications

(57 citation statements)

References 36 publications

(75 reference statements)

Supporting

Mentioning

Contrasting

Order By: Relevance

“…The range weights are computed as a function of the image intensity gradients, W K X t,e (Eqs. [13][14]. The K subscript of W denotes that the filtration domain is resampled with a spatially invariant, 2 nd order, classic kernel, denoted by K, prior to the computation of the intensity gradient [12].…”

Section: Regularmentioning

confidence: 99%

See 1 more Smart Citation

Joint regularization for spectro-temporal CT reconstruction

Clark

Badea

2016

SPIE Proceedings

View full text Add to dashboard Cite

X-ray CT is widely used, both clinically and preclinically, for fast, high-resolution, anatomic imaging; however, compelling opportunities exist to expand its use in functional imaging applications. For instance, spectral information combined with nanoparticle contrast agents enables quantification of tissue perfusion levels, while temporal information details cardiac and respiratory dynamics. In previous work, we proposed and demonstrated a projection acquisition and reconstruction strategy for 5D CT (3D + dual-energy + time) which recovered spectral and temporal information without substantially increasing radiation dose or sampling time relative to anatomic imaging protocols. The approach relied on the approximate separability of the temporal and spectral reconstruction sub-problems, which enabled substantial projection undersampling and effective regularization. Here, we extend this previous work to more general, nonseparable 5D CT reconstruction cases (3D + muti-energy + time) with applicability to K-edge imaging of exogenous contrast agents. We apply the newly proposed algorithm in phantom simulations using a realistic system and noise model for a photon counting x-ray detector with six energy thresholds. The MOBY mouse phantom used contains realistic concentrations of iodine, gold, and calcium in water. Relative to weighted least-squares reconstruction, the proposed 5D reconstruction algorithm improved reconstruction and material decomposition accuracy by 3-18 times. Furthermore, by exploiting joint, low rank image structure between time points and energies, ~80 HU of contrast associated with the Kedge of gold and ~35 HU of contrast associated with the blood pool and myocardium were recovered from more than 400 HU of noise.

show abstract

Section: Regularmentioning

confidence: 99%

“…The purpose of this resampling kernel is to improve denoising performance at edges in band-limited CT data. More details of our implementations of joint and 4D BF can be found in our previous work [1, 13,14].…”

Section: Regularmentioning

confidence: 99%

Joint regularization for spectro-temporal CT reconstruction

Clark

Badea

2016

SPIE Proceedings

View full text Add to dashboard Cite

show abstract

“…Noise measurements are performed algorithmically using high pass filtration and the median absolute deviation metric. 15 The range weights are computed as a function of the image intensity gradients:…”

Section: Joint Reconstruction Of Hybrid Ct Datamentioning

confidence: 99%

“…15 Notably, the range kernel is constructed jointly using the noisy data at each energy (Eq. 7) and is then used to filter each component energy (Eq.…”

Section: Joint Reconstruction Of Hybrid Ct Datamentioning

confidence: 99%

Resolution-enhancing hybrid, spectral CT reconstruction

Clark

Badea

2016

SPIE Proceedings

View full text Add to dashboard Cite

Spectral x-ray imaging based on photon-counting x-ray detectors (PCXD) is an area of growing interest. By measuring the energy of x-ray photons, a spectral CT system can better differentiate elements using a single scan. However, the spatial resolution achievable with most PCXDs limits their application, particularly in preclinical CT imaging. Consequently, our group is developing a hybrid micro-CT scanner based on a high-resolution, energy-integrating (EID) detector and a lower-resolution, PCXD. To complement this system, we propose and demonstrate a hybrid, spectral CT reconstruction algorithm which robustly combines the spectral contrast of the PCXD with the spatial resolution of the EID. Specifically, the high-resolution, spectrally resolved data (X) is recovered as the sum of two matrices: one with low column rank (X L ) determined from the EID data and one with intensity gradient sparse columns (X S ) corresponding to the upsampled spectral contrast obtained from the PCXD data. We test the proposed algorithm in a feasibility study focused on molecular imaging of atherosclerotic plaque using activatable iodine and gold nanoparticles. The results show accurate estimation of material concentrations at increased spatial resolution for a voxel size ratio between the PCXD and the EID of 500 µm 3 :100 µm 3 . Specifically, regularized, iterative reconstruction of the MOBY mouse phantom around the K-edges of iodine (33.2 keV) and gold (80.7 keV) reduces the reconstruction error by more than a factor of three relative to least-squares, algebraic reconstruction. Likewise, the material decomposition accuracy into iodine, gold, calcium, and water improves by more than a factor of two.

show abstract

“…It can be determined in a two-step procedure, i.e., image reconstruction for spectral images and then material decomposition from these spectral images to material compositions [3], [6]–[16], or alternatively material-specific sinogram decomposition and then material reconstruction [4], [17]–[19]. Various iterative reconstruction models have been developed [20], with energy-by-energy reconstruction [3], [4], [9], [11], [17]–[19] and joint reconstruction [7], [10], [15], [16], such as total variation (TV) sparsity [14], [16], HYPR algorithm [8], tight frame sparsity [3], [11], bilateral filtration [12], [13], patch-based low-rank model [15], rank-and-sparsity decomposition model [7] and its tensor version [10]. In order to fully utilize the image similarity in the spectral dimension, the joint reconstruction is a natural formulation [7], [10], [15], [16].…”

Section: Introductionmentioning

confidence: 99%

TICMR: Total Image Constrained Material Reconstruction via Nonlocal Total Variation Regularization for Spectral CT

Liu

Ding

Molloi

et al. 2016

IEEE Trans. Med. Imaging

View full text Add to dashboard Cite

This work develops a material reconstruction method for spectral CT, namely Total Image Constrained Material Reconstruction (TICMR), to maximize the utility of projection data in terms of both spectral information and high signal-to-noise ratio (SNR). This is motivated by the following fact: when viewed as a spectrally-integrated measurement, the projection data can be used to reconstruct a total image without spectral information, which however has a relatively high SNR; when viewed as a spectrally-resolved measurement, the projection data can be utilized to reconstruct the material composition, which however has a relatively low SNR. The material reconstruction synergizes material decomposition and image reconstruction, i.e., the direct reconstruction of material compositions instead of a two-step procedure that first reconstructs images and then decomposes images. For material reconstruction with high SNR, we propose TICMR with nonlocal total variation (NLTV) regularization. That is, first we reconstruct a total image using spectrally-integrated measurement without spectral binning, and build the NLTV weights from this image that characterize nonlocal image features; then the NLTV weights are incorporated into a NLTV-based iterative material reconstruction scheme using spectrally-binned projection data, so that these weights serve as a high-SNR reference to regularize material reconstruction. Note that the nonlocal property of NLTV is essential for material reconstruction, since material compositions may have significant local intensity variations although their structural information is often similar. In terms of solution algorithm, TICMR is formulated as an iterative reconstruction method with the NLTV regularization, in which the nonlocal divergence is utilized based on the adjoint relationship. The alternating direction method of multipliers is developed to solve this sparsity optimization problem. The proposed TICMR method was validated using both simulated and experimental data. In comparison with FBP and total-variation-based iterative method, TICMR had improved image quality, e.g., contrast-to-noise ratio and spatial resolution.

show abstract

Spectral diffusion: an algorithm for robust material decomposition of spectral CT data

Cited by 49 publications

References 36 publications

Joint regularization for spectro-temporal CT reconstruction

Joint regularization for spectro-temporal CT reconstruction

Resolution-enhancing hybrid, spectral CT reconstruction

TICMR: Total Image Constrained Material Reconstruction via Nonlocal Total Variation Regularization for Spectral CT

Contact Info

Product

Resources

About