Deep-learning-based direct inversion for material decomposition

Prior image-based generative adversarial learning for multi-material decomposition in photon counting computed tomography

BACKGROUND

Photon counting detector computed tomography (PCD-CT) is a novel promising technique providing higher spatial resolution, lower radiation dose and greater energy spectrum differentiation, which create more possibilities to improve image quality. Multi-material decomposition is an attractive application for PCD-CT to identify complicated materials and provide accurate quantitative analysis. However, limited by the finite photon counting rate in each energy window of photon counting detector, the noise problem hinders the decomposition of high-quality basis material images.

METHODS

To address this issue, an end-to-end multi-material decomposition network based on prior images is proposed in this paper. First, the reconstructed images corresponding to the full spectrum with less noise are introduced as prior information to improve the overall signal-to-noise ratio of the data. Then, a generative adversarial network is designed to mine the relationship between reconstructed images and basis material images based on the information interaction of material decomposition. Furthermore, a weighted edge loss is introduced to adapt to the structural differences of different basis material images.

RESULTS

To verify the performance of the proposed method, simulation and real studies are carried out. In simulation study of structured fibro-glandular tissue model, the results show that the proposed method decreased the root mean square error by 67 % and 26 % on adipose, 66 % and 28 % on fibroglandular, 52 % and 8 % on calcification, compared to butterfly network and dual interactive Wasserstein generative adversarial network.

CONCLUSION

Experimentally, the proposed method shows certain advantages over other methods on noise suppression effect, detail retention ability and decomposition accuracy.

Deep learning-based material decomposition of iodine and calcium in mobile photon counting detector CT

Photon-counting detector (PCD)-based computed tomography (CT) offers several advantages over conventional energy-integrating detector-based CT. Among them, the ability to discriminate energy exhibits significant potential for clinical applications because it provides material-specific information. That is, material decomposition (MD) can be achieved through energy discrimination. In this study, deep learning-based material decomposition was performed using live animal data. We propose MD-Unet, which is a deep learning strategy for material decomposition based on an Unet architecture trained with data from three energy bins. To mitigate the data insufficiency, we developed a pretrained model incorporating various simulation data forms and augmentation strategies. Incorporating these approaches into model training results in enhanced precision in material decomposition, thereby enabling the identification of distinct materials at individual pixel locations. The trained network was applied to the acquired animal data to evaluate material decomposition results. Compared with conventional methods, the newly generated MD-Unet demonstrated more accurate material decomposition imaging. Moreover, the network demonstrated an improved material decomposition ability and significantly reduced noise. In addition, they can potentially offer an enhancement level similar to that of a typical contrast agent. This implies that it can acquire images of the same quality with fewer contrast agents administered to patients, thereby demonstrating its significant clinical value.

Image-domain material decomposition for dual-energy CT using unsupervised learning with data-fidelity loss

BACKGROUND

Dual-energy computed tomography (DECT) and material decomposition play vital roles in quantitative medical imaging. However, the decomposition process may suffer from significant noise amplification, leading to severely degraded image signal-to-noise ratios (SNRs). While existing iterative algorithms perform noise suppression using different image priors, these heuristic image priors cannot accurately represent the features of the target image manifold. Although deep learning-based decomposition methods have been reported, these methods are in the supervised-learning framework requiring paired data for training, which is not readily available in clinical settings.

PURPOSE

This work aims to develop an unsupervised-learning framework with data-measurement consistency for image-domain material decomposition in DECT.

METHODS

The proposed framework combines iterative decomposition and deep learning-based image prior in a generative adversarial network (GAN) architecture. In the generator module, a data-fidelity loss is introduced to enforce the measurement consistency in material decomposition. In the discriminator module, the discriminator is trained to differentiate the low-noise material-specific images from the high-noise images. In this scheme, paired images of DECT and ground-truth material-specific images are not required for the model training. Once trained, the generator can perform image-domain material decomposition with noise suppression in a single step.

RESULTS

In the simulation studies of head and lung digital phantoms, the proposed method reduced the standard deviation (SD) in decomposed images by 97% and 91% from the values in direct inversion results. It also generated decomposed images with structural similarity index measures (SSIMs) greater than 0.95 against the ground truth. In the clinical head and lung patient studies, the proposed method suppressed the SD by 95% and 93% compared to the decomposed images of matrix inversion.

CONCLUSIONS

Since the invention of DECT, noise amplification during material decomposition has been one of the biggest challenges, impeding its quantitative use in clinical practice. The proposed method performs accurate material decomposition with efficient noise suppression. Furthermore, the proposed method is within an unsupervised-learning framework, which does not require paired data for model training and resolves the issue of lack of ground-truth data in clinical scenarios.

Multi-energy CT material decomposition using graph model improved CNN

In spectral CT imaging, the coefficient image of the basis material obtained by the material decomposition technique can estimate the tissue composition, and its accuracy directly affects the disease diagnosis. Although the precision of material decomposition is increased by employing convolutional neural networks (CNN), extracting the non-local features from the CT image is restricted using the traditional CNN convolution operator. A graph model built by multi-scale non-local self-similar patterns is introduced into multi-material decomposition (MMD). We proposed a novel MMD method based on graph edge-conditioned convolution U-net (GECCU-net) to enhance material image quality. The GECCU-net focuses on developing a multi-scale encoder. At the network coding stage, three paths are applied to capture comprehensive image features. The local and non-local feature aggregation (LNFA) blocks are designed to integrate the local and non-local features from different paths. The graph edge-conditioned convolution based on non-Euclidean space excavates the non-local features. A hybrid loss function is defined to accommodate multi-scale input images and avoid over-smoothing of results. The proposed network is compared quantitatively with base CNN models on the simulated and real datasets. The material images generated by GECCU-net have less noise and artifacts while retaining more information on tissue. The Structural SIMilarity (SSIM) of the obtained abdomen and chest water maps reaches 0.9976 and 0.9990, respectively, and the RMSE reduces to 0.1218 and 0.4903 g/cm3. The proposed method can improve MMD performance and has potential applications.

Systematic Review on Learning-based Spectral CT

Spectral computed tomography (CT) has recently emerged as an advanced version of medical CT and significantly improves conventional (single-energy) CT. Spectral CT has two main forms: dual-energy computed tomography (DECT) and photon-counting computed tomography (PCCT), which offer image improvement, material decomposition, and feature quantification relative to conventional CT. However, the inherent challenges of spectral CT, evidenced by data and image artifacts, remain a bottleneck for clinical applications. To address these problems, machine learning techniques have been widely applied to spectral CT. In this review, we present the state-of-the-art data-driven techniques for spectral CT.

Framework for dual-energy-like chest radiography image synthesis from single-energy computed tomography based on cycle-consistent generative adversarial network

BACKGROUND

Dual-energy (DE) chest radiography (CXR) enables the selective imaging of two relevant materials, namely, soft tissue and bone structures, to better characterize various chest pathologies (i.e., lung nodule, bony lesions, etc.) and potentially improve CXR-based diagnosis. Recently, deep-learning-based image synthesis techniques have attracted considerable attention as alternatives to existing DE methods (i.e., dual-exposure-based and sandwich-detector-based methods) because software-based bone-only and bone-suppression images in CXR could be useful.

PURPOSE

The objective of this study was to develop a new framework for DE-like CXR image synthesis from single-energy computed tomography (CT) based on a cycle-consistent generative adversarial network.

METHODS

The core techniques of the proposed framework are divided into three categories: (1) data configuration from the generation of pseudo CXR from single energy CT, (2) learning of the developed network architecture using pseudo CXR and pseudo-DE imaging using a single-energy CT, and (3) inference of the trained network on real single-energy CXR. We performed a visual inspection and comparative evaluation using various metrics and introduced a figure of image quality (FIQ) to consider the effects of our framework on the spatial resolution and noise in terms of a single index through various test cases.

RESULTS

Our results indicate that the proposed framework is effective and exhibits potential synthetic imaging ability for two relevant materials: soft tissue and bone structures. Its effectiveness was validated, and its ability to overcome the limitations associated with DE imaging techniques (e.g., increase in exposure dose owing to the requirement of two acquisitions, and emphasis on noise characteristics) via an artificial intelligence technique was presented.

CONCLUSIONS

The developed framework addresses X-ray dose issues in the field of radiation imaging and enables pseudo-DE imaging with single exposure.

Synergizing photon-counting CT with deep learning: potential enhancements in medical imaging

This review article highlights the potential of integrating photon-counting computed tomography (CT) and deep learning algorithms in medical imaging to enhance diagnostic accuracy, improve image quality, and reduce radiation exposure. The use of photon-counting CT provides superior image quality, reduced radiation dose, and material decomposition capabilities, while deep learning algorithms excel in automating image analysis and improving diagnostic accuracy. The integration of these technologies can lead to enhanced material decomposition and classification, spectral image analysis, predictive modeling for individualized medicine, workflow optimization, and radiation dose management. However, data requirements, computational resources, and regulatory and ethical concerns remain challenges that need to be addressed to fully realize the potential of this technology. The fusion of photon-counting CT and deep learning algorithms is poised to revolutionize medical imaging and transform patient care.

Systematic Review on Learning-based Spectral CT

Spectral computed tomography (CT) has recently emerged as an advanced version of medical CT and significantly improves conventional (single-energy) CT. Spectral CT has two main forms: dual-energy computed tomography (DECT) and photon-counting computed tomography (PCCT), which offer image improvement, material decomposition, and feature quantification relative to conventional CT. However, the inherent challenges of spectral CT, evidenced by data and image artifacts, remain a bottleneck for clinical applications. To address these problems, machine learning techniques have been widely applied to spectral CT. In this review, we present the state-of-the-art data-driven techniques for spectral CT.

Imaging of Multiple Myeloma: Present and Future

Multiple myeloma (MM) is the second most common adult hematologic malignancy, and early intervention increases survival in asymptomatic high-risk patients. Imaging is crucial for the diagnosis and follow-up of MM, as the detection of bone and bone marrow lesions often dictates the decision to start treatment. Low-dose whole-body computed tomography (CT) is the modality of choice for the initial assessment, and dual-energy CT is a developing technique with the potential for detecting non-lytic marrow infiltration and evaluating the response to treatment. Magnetic resonance imaging (MRI) is more sensitive and specific than 18F-fluorodeoxyglucose positron emission tomography/computed tomography (FDG-PET/CT) for the detection of small focal lesions and diffuse marrow infiltration. However, FDG-PET/CT is recommended as the modality of choice for follow-up. Recently, diffusion-weighted MRI has become a new technique for the quantitative assessment of disease burden and therapy response. Although not widespread, we address current proposals for structured reporting to promote standardization and diminish variations. This review provides an up-to-date overview of MM imaging, indications, advantages, limitations, and recommended reporting of each technique. We also cover the main differential diagnosis and pitfalls and discuss the ongoing controversies and future directions, such as PET-MRI and artificial intelligence.

Joint Reconstruction and Spectrum Refinement for Photon-Counting-Detector Spectral CT

Photon-counting detector CT (PCD-CT) is a revolutionary technology in decades in the field of CT. Its potential benefits in lowering noise, dose reduction, and material-specific imaging enable completely new clinical applications. Spectral reconstruction of basis material maps requires knowledge of the x-ray spectrum and the spectral response calibration of the detector. However, spectrum estimation errors caused by inaccurate energy threshold calibration will degrade the accuracy of the reconstructions. Existing spectrum estimation methods are not adequately modeled for bias in energy threshold position. Besides, directly solving a big number of variables of the pixel-wise effective spectra for PCD is an ill-conditioned problem so that stable solution is hardly achievable. In this paper, we assumed the effective spectra variation across the detector mainly comes from the calibration error in the energy threshold positions as well as the intrinsic threshold distribution. We propose a joint reconstruction and spectrum refinement algorithm (JoSR) that introduces an innovative spectrum model based on non-negative matrix factorization (NMF) to significantly reduce the dimension of unknowns so that makes the problem well-conditioned. The polychromatic spectral imaging model and the basis material decomposition method together form an optimization objective. The proximal regularized block coordinate descent algorithm is adopted to deal with the non-convex optimization problem to ensure convergence. Simulation studies and experiments on a laboratory PCD-CT system validated the proposed JoSR method. The results demonstrate its advantages on image quality and quantitative accuracy over other state-of-the-art methods in the field.

Detector shifting and deep learning based ring artifact correction method for low-dose CT

BACKGROUND

In x-ray computed tomography (CT), the gain inconsistency of detector units leads to ring artifacts in the reconstructed images, seriously destroys the image structure, and is not conducive to image recognition. In addition, to reduce radiation dose and scanning time, especially photon counting CT, low-dose CT is required, so it is important to reduce the noise and suppress ring artifacts in low-dose CT images simultaneously.

PURPOSE

Deep learning is an effective method to suppress ring artifacts, but there are still residual artifacts in corrected images. And the feature recognition ability of the network for ring artifacts decreases due to the effect of noise in the low-dose CT images. In this paper, a method is proposed to achieve noise reduction and ring artifact removal simultaneously.

METHODS

To solve these problems, we propose a ring artifact correction method for low-dose CT based on detector shifting and deep learning in this paper. Firstly, at the CT scanning stage, the detector horizontally shifts randomly at each projection to alleviate the ring artifacts as front processing. Thus, the ring artifacts are transformed into dispersed noise in front processed images. Secondly, deep learning is used for dispersed noise and statistical noise reduction.

RESULTS

Both simulation and real data experiments are conducted to evaluate the proposed method. Compared to other methods, the results show that the proposed method in this paper has better effect on removing ring artifacts in the low-dose CT images. Specifically, the RMSEs and SSIMs of the two sets of simulated and experiment data are better compared to the raw images significantly.

CONCLUSIONS

The method proposed in this paper combines detector shifting and deep learning to remove ring artifacts and statistical noise simultaneously. The results show that the proposed method is able to get better performance.

Deep learning-based virtual noncalcium imaging in multiple myeloma using dual-energy CT

BACKGROUND

Dual-energy CT with virtual noncalcium (VNCa) images allows the evaluation of focal intramedullary bone marrow involvement in patients with multiple myeloma. However, current commercial VNCa techniques suffer from excessive image noise and artifacts due to material decomposition used in synthesizing VNCa images.

OBJECTIVES

In this work, we aim to improve VNCa image quality for the assessment of focal multiple myeloma, using an Artificial intelligence based Generalizable Algorithm for mulTi-Energy CT (AGATE) method.

MATERIALS AND METHODS

AGATE method used a custom dual-task convolutional neural network (CNN) that concurrently carries out material classification and quantification. The material classification task provided an auxiliary regularization to the material quantification task. CNN parameters were optimized using custom loss functions that involved cross-entropy, physics-informed constraints, structural redundancy in spectral and material images, and texture information in spectral images. For training data, CT phantoms (diameters 30 to 45 cm) with tissue-mimicking inserts were scanned on a third generation dual-source CT system. Scans were performed at routine dose and half of the routine dose. Small image patches (i.e., 40 × 40 pixels) of tissue-mimicking inserts with known basis material densities were extracted for training samples. Numerically simulated insert materials with various shapes increased diversity of training samples. Generalizability of AGATE was evaluated using CT images from phantoms and patients. In phantoms, material decomposition accuracy was estimated using mean-absolute-percent-error (MAPE), using physical inserts that were not used during the training. Noise power spectrum (NPS) and modulation transfer function (MTF) were compared across phantom sizes and radiation dose levels. Five patients with multiple myeloma underwent dual-energy CT, with VNCa images generated using a commercial method and AGATE. Two fellowship-trained musculoskeletal radiologists reviewed the VNCa images (commercial and AGATE) side-by-side using a dual-monitor display, blinded to VNCa type, rating the image quality for focal multiple myeloma lesion visualization using a 5-level Likert comparison scale (-2 = worse visualization and diagnostic confidence, -1 = worse visualization but equivalent diagnostic confidence, 0 = equivalent visualization and diagnostic confidence, 1 = improved visualization but equivalent diagnostic confidence, 2 = improved visualization and diagnostic confidence). A post hoc assignment of comparison ratings was performed to rank AGATE images in comparison to commercial ones.

RESULTS

AGATE demonstrated consistent material quantification accuracy across phantom sizes and radiation dose levels, with MAPE ranging from 0.7% to 4.4% across all testing materials. Compared to commercial VNCa images, the AGATE-synthesized VNCa images yielded considerably lower image noise (50-77% noise reduction) without compromising noise texture or spatial resolution across different phantom sizes and two radiation doses. AGATE VNCa images had markedly reduced area under NPS curves and maintained NPS peak frequency (0.7 lp/cm to 1.0 lp/cm), with similar MTF curves (50% MTF at 3.0 lp/cm). In patients, AGATE demonstrated reduced image noise and artifacts with improved delineation of focal multiple myeloma lesions (all readers comparison scores indicating improved overall diagnostic image quality [scores 1 or 2]).

CONCLUSIONS

AGATE demonstrated reduced noise and artifacts in VNCa images and ability to improve visualization of bone marrow lesions for assessing multiple myeloma.

Small animal photon counting cone-beam CT on a preclinical radiation research platform to improve radiation dose calculation accuracy

Objective.Cone beam CT (CBCT) in preclinical small animal irradiation platforms provides essential information for image guidance and radiation dose calculation for experiment planning. This project developed a photon-counting detector (PCD)-based multi(3)-energy (ME-)CBCT on a small animal irradiator to improve the accuracy of material differentiation and hence dose calculation, and compared to conventional flat panel detector (FPD)-based CBCT.Approach.We constructed a mechanical structure to mount a PCD to an existing preclinical irradiator platform and built a data acquisition pipeline to acquire x-ray projection data with a 100 kVp x-ray beam using three different energy thresholds in a single gantry rotation. We implemented an energy threshold optimization scheme to determine optimal thresholds to balance signal-to-noise ratios (SNRs) among energy channels. Pixel-based detector response calibration was performed to remove ring artifacts in reconstructed CBCT images. Feldkamp-Davis-Kress method was employed to reconstruct CBCT images and a total-variance regularization-based optimization model was used to decompose CBCT images into bone and water material images. We compared dose calculation results using PCD-based ME-CBCT with that of FPD-based CBCT.Main results.The optimal nominal energy thresholds were determined as 26, 56, and 90 keV, under which SNRs in a selected region-of-interest in the water region were 6.11, 5.91 and 5.93 in the three energy channels, respectively. Compared with dose calculation results using FPD-based CBCT, using PCD-based ME-CBCT reduced the mean relative error from 49.5% to 16.4% in bone regions and from 7.5% to 6.9% in soft tissue regions.Significance.PCD-based ME-CBCT is beneficial in improving radiation dose calculation accuracy in experiment planning of preclinical small animal irradiation researches.

Feasibility study of three-material decomposition in dual-energy cone-beam CT imaging with deep learning

Objective. In this work, a dedicated end-to-end deep convolutional neural network, named as Triple-CBCT, is proposed to demonstrate the feasibility of reconstructing three different material distribution volumes from the dual-energy CBCT projection data. Approach. In Triple-CBCT, the features of the sinogram and the CT image are independently extracted and cascaded via a customized domain transform network module. This Triple-CBCT network was trained by numerically synthesized dual-energy CBCT data, and was tested with experimental dual-energy CBCT data of the Iodine-CaCl2 solution and pig leg specimen scanned on an in-house benchtop system. Main results. Results show that the information stored in both the sinogram and CT image domains can be used together to improve the decomposition quality of multiple materials (water, iodine, CaCl2 or bone) from the dual-energy projections. In addition, both the numerical and experimental results demonstrate that the Triple-CBCT is able to generate high-fidelity dual-energy CBCT basis images. Significance. An innovative end-to-end network that joints the sinogram and CT image domain information is developed to facilitate high quality automatic decomposition from the dual-energy CBCT scans.

Material decomposition from photon-counting CT using a convolutional neural network and energy-integrating CT training labels

Objective.Photon-counting CT (PCCT) has better dose efficiency and spectral resolution than energy-integrating CT, which is advantageous for material decomposition. Unfortunately, the accuracy of PCCT-based material decomposition is limited due to spectral distortions in the photon-counting detector (PCD).Approach.In this work, we demonstrate a deep learning (DL) approach that compensates for spectral distortions in the PCD and improves accuracy in material decomposition by using decomposition maps provided by high-dose multi-energy-integrating detector (EID) data as training labels. We use a 3D U-net architecture and compare networks with PCD filtered back projection (FBP) reconstruction (FBP2Decomp), PCD iterative reconstruction (Iter2Decomp), and PCD decomposition (Decomp2Decomp) as the input.Main results.We found that our Iter2Decomp approach performs best, but DL outperforms matrix inversion decomposition regardless of the input. Compared to PCD matrix inversion decomposition, Iter2Decomp gives 27.50% lower root mean squared error (RMSE) in the iodine (I) map and 59.87% lower RMSE in the photoelectric effect (PE) map. In addition, it increases the structural similarity (SSIM) by 1.92%, 6.05%, and 9.33% in the I, Compton scattering (CS), and PE maps, respectively. When taking measurements from iodine and calcium vials, Iter2Decomp provides excellent agreement with multi-EID decomposition. One limitation is some blurring caused by our DL approach, with a decrease from 1.98 line pairs/mm at 50% modulation transfer function (MTF) with PCD matrix inversion decomposition to 1.75 line pairs/mm at 50% MTF when using Iter2Decomp.Significance.Overall, this work demonstrates that our DL approach with high-dose multi-EID derived decomposition labels is effective at generating more accurate material maps from PCD data. More accurate preclinical spectral PCCT imaging such as this could serve for developing nanoparticles that show promise in the field of theranostics (therapy and diagnostics).

Improving coronary artery imaging in single source CT with cardiac motion correction using attention and spatial transformer based neural networks

Motion artifact is a major challenge in cardiac CT which hampers accurate delineation of key anatomic (e.g. coronary lumen) and pathological features (e.g. stenosis). Conventional motion correction techniques are limited on patients with high / irregular heart rate, due to simplified modeling of CT systems and cardiac motion. Emerging deep learning based cardiac motion correction techniques have demonstrated the potential of further quality improvement. Yet, many methods require CT projection data or advanced motion simulation tools that are not readily available. We aim to develop an image-domain motion-correction method, using convolutional neural network (CNN) integrated with customized attention and spatial transformer techniques. Forty cardiac CT exams acquired from a clinical dual-source CT system were retrospectively collected to generate training (n=26) and testing (n=14) sets. Dual-source data uniquely allow image reconstruction with different temporal resolutions from the same patient scan. Slow temporal resolution (140ms; equivalent to single-source CT (SSCT) half scan) and fast temporal resolution (75ms; dual source) images were reconstructed to generate paired samples of motion-corrupted and reference images. The combinations of 2 training-inference strategies and 3 CNNs were evaluated: strategy #1 - whole-heart images in training / inference; strategy #2 - vessel patches in training / inference; CNN #1 - attention only; CNN #2 - spatial-transformer (STN) only; CNN #3 - attention & STN synergy. Testing data showed that CNN #3 with strategy #2 provided relatively better performance: improving vessel delineation, increasing structural similarity index from 0.85 to 0.91, and reducing mean CT number error of lumen by 71.0%. Our method could improve the image quality in cardiac exams with SSCT.

Dual-Contrast Biphasic Liver Imaging With Iodine and Gadolinium Using Photon-Counting Detector Computed Tomography: An Exploratory Animal Study

PURPOSE

The aims of this study were to develop a single-scan dual-contrast protocol for biphasic liver imaging with 2 intravenous contrast agents (iodine and gadolinium) and to evaluate its effectiveness in an exploratory swine study using a photon-counting detector computed tomography (PCD-CT) system.

MATERIALS AND METHODS

A dual-contrast CT protocol was developed for PCD-CT to simultaneously acquire 2 phases of liver contrast enhancement, with the late arterial phase enhanced by 1 contrast agent (iodine-based) and the portal venous phase enhanced by the other (gadolinium-based). A gadolinium contrast bolus (gadobutrol: 64 mL, 8 mL/s) and an iodine contrast bolus (iohexol: 40 mL, 5 mL/s) were intravenously injected in the femoral vein of a healthy domestic swine, with the second injection initiated after 17 seconds from the beginning of the first injection; PCD-CT image acquisition was performed 12 seconds after the beginning of the iodine contrast injection. A convolutional neural network (CNN)-based denoising technique was applied to PCD-CT images to overcome the inherent noise magnification issue in iodine/gadolinium decomposition task. Iodine and gadolinium material maps were generated using a 3-material decomposition method in image space. A set of contrast samples (mixed iodine and gadolinium) was attached to the swine belly; quantitative accuracy of material decomposition in these inserts between measured and true concentrations was calculated using root mean square error. An abdominal radiologist qualitatively evaluated the delineation of arterial and venous vasculatures in the swine liver using iodine and gadolinium maps obtained using the dual-contrast PCD-CT protocol.

RESULTS

The iodine and gadolinium samples attached to the swine were quantified with root mean square error values of 0.75 mg/mL for iodine and 0.45 mg/mL for gadolinium from the contrast material maps derived from the denoised PCD-CT images. Hepatic arteries containing iodine and veins containing gadolinium in the swine liver could be clearly visualized. Compared with the original images, better distinctions between 2 liver phases were achieved using CNN denoising, with approximately 60% to 80% noise reduction in contrast material maps acquired with the denoised PCD-CT images compared with the original images.

CONCLUSIONS

Simultaneous biphasic liver imaging in a single multienergy PCD-CT acquisition using a dual-contrast (iodine and gadolinium) injection protocol and CNN denoising was demonstrated in a swine study, where the enhanced hepatic arteries (containing iodine) and the enhanced hepatic veins (containing gadolinium) could be clearly visualized and delineated in the swine liver.

Deep learning enabled ultra-fast-pitch acquisition in clinical X-ray computed tomography

OBJECTIVE

In X-raycomputed tomography (CT), many important clinical applications may benefit from a fast acquisition speed. The helical scan is the most widely used acquisition mode in clinical CT, where a fast helical pitch can improve the acquisition speed. However, on a typical single-source helical CT (SSCT) system, the helical pitch p typically cannot exceed 1.5; otherwise, reconstruction artifacts will result from data insufficiency. The purpose of this work is to develop a deep convolutional neural network (CNN) to correct for artifacts caused by an ultra-fast pitch, which can enable faster acquisition speed than what is currently achievable.

METHODS

A customized CNN (denoted as ultra-fast-pitch network (UFP-net)) was developed to restore the underlying anatomical structure from the artifact-corrupted post-reconstruction data acquired from SSCT with ultra-fast pitch (i.e., p ≥ 2). UFP-net employed residual learning to capture the features of image artifacts. UFP-net further deployed in-house-customized functional blocks with spatial-domain local operators and frequency-domain non-local operators, to explore multi-scale feature representation. Images of contrast-enhanced patient exams (n = 83) with routine pitch setting (i.e., p < 1) were retrospectively collected, which were used as training and testing datasets. This patient cohort involved CT exams over different scan ranges of anatomy (chest, abdomen, and pelvis) and CT systems (Siemens Definition, Definition Flash, Definition AS+, Siemens Healthcare, Inc.), and the corresponding base CT scanning protocols used consistent settings of major scan parameters (e.g., collimation and pitch). Forward projection of the original images was calculated to synthesize helical CT scans with one regular pitch setting (p = 1) and two ultra-fast-pitch setting (p = 2 and 3). All patient images were reconstructed using the standard filtered-back-projection (FBP) algorithm. A customized multi-stage training scheme was developed to incrementally optimize the parameters of UFP-net, using ultra-fast-pitch images as network inputs and regular pitch images as labels. Visual inspection was conducted to evaluate image quality. Structural similarity index (SSIM) and relative root-mean-square error (rRMSE) were used as quantitative quality metrics.

RESULTS

The UFP-net dramatically improved image quality over standard FBP at both ultra-fast-pitch settings. At p = 2, UFP-net yielded higher mean SSIM (> 0.98) with lower mean rRMSE (< 2.9%), compared to FBP (mean SSIM < 0.93; mean rRMSE > 9.1%). Quantitative metrics at p = 3: UFP-net-mean SSIM [0.86, 0.94] and mean rRMSE [5.0%, 8.2%]; FBP-mean SSIM [0.36, 0.61] and mean rRMSE [36.0%, 58.6%].

CONCLUSION

The proposed UFP-net has the potential to enable ultra-fast data acquisition in clinical CT without sacrificing image quality. This method has demonstrated reasonable generalizability over different body parts when the corresponding CT exams involved consistent base scan parameters.

Deep-learning-based direct synthesis of low-energy virtual monoenergetic images with multi-energy CT

Purpose: We developed a deep learning method to reduce noise and beam-hardening artifact in virtual monoenergetic image (VMI) at low x-ray energy levels. Approach: An encoder-decoder type convolutional neural network was implemented with customized inception modules and in-house-designed training loss (denoted as Incept-net), to directly estimate VMI from multi-energy CT images. Images of an abdomen-sized water phantom with varying insert materials were acquired from a research photon-counting-detector CT. The Incept-net was trained with image patches ( 64 × 64    pixels ) extracted from the phantom data, as well as synthesized, random-shaped numerical insert materials. The whole CT images ( 512 × 512    pixels ) with the remaining real insert materials that were unseen in network training were used for testing. Seven contrast-enhanced abdominal CT exams were used for preliminary evaluation of Incept-net generalizability over anatomical background. Mean absolute percentage error (MAPE) was used to evaluate CT number accuracy. Results: Compared to commercial VMI software, Incept-net largely suppressed beam-hardening artifact and reduced noise (53%) in phantom study. Incept-net presented comparable CT number accuracy at higher-density ( P -value [0.0625, 0.999]) and improved it at lower-density inserts ( P - value = 0.0313 ) with overall MAPE: Incept-net [2.9%, 4.6%]; commercial-VMI [6.7%, 10.9%]. In patient images, Incept-net suppressed beam-hardening artifact and reduced noise (up to 50%, P - value = 0.0156 ). Conclusion: In this preliminary study, Incept-net presented the potential to improve low-energy VMI quality.

Weakly-supervised progressive denoising with unpaired CT images

Although low-dose CT imaging has attracted a great interest due to its reduced radiation risk to the patients, it suffers from severe and complex noise. Recent fully-supervised methods have shown impressive performances on CT denoising task. However, they require a huge amount of paired normal-dose and low-dose CT images, which is generally unavailable in real clinical practice. To address this problem, we propose a weakly-supervised denoising framework that generates paired original and noisier CT images from unpaired CT images using a physics-based noise model. Our denoising framework also includes a progressive denoising module that bypasses the challenges of mapping from low-dose to normal-dose CT images directly via progressively compensating the small noise gap. To quantitatively evaluate diagnostic image quality, we present the noise power spectrum and signal detection accuracy, which are well correlated with the visual inspection. The experimental results demonstrate that our method achieves remarkable performances, even superior to fully-supervised CT denoising with respect to the signal detectability. Moreover, our framework increases the flexibility in data collection, allowing us to utilize any unpaired data at any dose levels.