Model Generalizability Investigation for GFCE-MRI Synthesis in NPC Radiotherapy Using Multi-Institutional Patient-Based Data Normalization

Recently, deep learning has been demonstrated to be feasible in eliminating the use of gadoliniumbased contrast agents (GBCAs) through synthesizing gadolinium-free contrast-enhanced MRI (GFCE-MRI) from contrast-free MRI sequences, providing the community with an alternative to get rid of GBCAs-associated safety issues in patients. Nevertheless, generalizability assessment of the GFCE-MRI model has been largely challenged by the high inter-institutional heterogeneity of MRI data, on top of the scarcity of multi-institutional data itself. Although various data normalization methods have been adopted to address the heterogeneity issue, it has been limited to single-institutional investigation and there is no standard normalization approach presently. In this study, we aimed at investigating generalizability of GFCE-MRI model using data from seven institutions by manipulating heterogeneity of MRI data under five popular normalization approaches. Three state-of-the-art neural networks were applied to map from T1-weighted and T2-weighted MRI to contrast-enhanced MRI (CE-MRI) for GFCE-MRI synthesis in patients with nasopharyngeal carcinoma. MRI data from three institutions were used separately to generate three uni-institution models and jointly for a tri-institution model. The five normalization methods were applied to normalize the data of each model. MRI data from the remaining four institutions served as external cohorts for model generalizability assessment. Quality of GFCE-MRI was quantitatively evaluated against ground-truth CE-MRI using mean absolute error (MAE) and peak signal-to-noise ratio(PSNR). Results showed that performance of all uni-institution models remarkably dropped on the external cohorts. By contrast, model trained using multi-institutional data with Z-Score normalization yielded the best model generalizability improvement.

Complementation-reinforced network for integrated reconstruction and segmentation of pulmonary gas MRI with high acceleration

BACKGROUND

Hyperpolarized (HP) gas MRI enables the clear visualization of lung structure and function. Clinically relevant biomarkers, such as ventilated defect percentage (VDP) derived from this modality can quantify lung ventilation function. However, long imaging time leads to image quality degradation and causes discomfort to the patients. Although accelerating MRI by undersampling k-space data is available, accurate reconstruction and segmentation of lung images are quite challenging at high acceleration factors.

PURPOSE

To simultaneously improve the performance of reconstruction and segmentation of pulmonary gas MRI at high acceleration factors by effectively utilizing the complementary information in different tasks.

METHODS

A complementation-reinforced network is proposed, which takes the undersampled images as input and outputs both the reconstructed images and the segmentation results of lung ventilation defects. The proposed network comprises a reconstruction branch and a segmentation branch. To effectively exploit the complementary information, several strategies are designed in the proposed network. Firstly, both branches adopt the encoder-decoder architecture, and their encoders are designed to share convolutional weights for facilitating knowledge transfer. Secondly, a designed feature-selecting block discriminately feeds shared features into decoders of both branches, which can adaptively pick suitable features for each task. Thirdly, the segmentation branch incorporates the lung mask obtained from the reconstructed images to enhance the accuracy of the segmentation results. Lastly, the proposed network is optimized by a tailored loss function that efficiently combines and balances these two tasks, in order to achieve mutual benefits.

RESULTS

Experimental results on the pulmonary HP 129 Xe MRI dataset (including 43 healthy subjects and 42 patients) show that the proposed network outperforms state-of-the-art methods at high acceleration factors (4, 5, and 6). The peak signal-to-noise ratio (PSNR), structural similarity (SSIM), and Dice score of the proposed network are enhanced to 30.89, 0.875, and 0.892, respectively. Additionally, the VDP obtained from the proposed network has good correlations with that obtained from fully sampled images (r = 0.984). At the highest acceleration factor of 6, the proposed network promotes PSNR, SSIM, and Dice score by 7.79%, 5.39%, and 9.52%, respectively, in comparison to the single-task models.

CONCLUSION

The proposed method effectively enhances the reconstruction and segmentation performance at high acceleration factors up to 6. It facilitates fast and high-quality lung imaging and segmentation, and provides valuable support in the clinical diagnosis of lung diseases.

An artificial intelligence algorithm for pulmonary embolism detection on polychromatic computed tomography: performance on virtual monochromatic images

OBJECTIVES

Virtual monochromatic images (VMI) are increasingly used in clinical practice as they improve contrast-to-noise ratio. However, due to their different appearances, the performance of artificial intelligence (AI) trained on conventional CT images may worsen. The goal of this study was to assess the performance of an established AI algorithm trained on conventional polychromatic computed tomography (CT) images (CPI) to detect pulmonary embolism (PE) on VMI.

METHODS

Paired 60 kiloelectron volt (keV) VMI and CPI of 114 consecutive patients suspected of PE, obtained with a detector-based spectral CT scanner, were retrospectively analyzed by an established AI algorithm. The CT pulmonary angiography (CTPA) were classified as positive or negative for PE on a per-patient level. The reference standard was established using a comprehensive method that combined the evaluation of the attending radiologist and three experienced cardiothoracic radiologists aided by two different detection tools. Sensitivity, specificity, positive and negative predictive values and likelihood ratios of the algorithm on VMI and CPI were compared.

RESULTS

The prevalence of PE according to the reference standard was 35.1% (40 patients). None of the diagnostic accuracy measures of the algorithm showed a significant difference between CPI and VMI. Sensitivity was 77.5% (95% confidence interval (CI) 64.6-90.4%) and 85.0% (73.9-96.1%) (p = 0.08) on CPI and VMI respectively and specificity 96.0% (91.4-100.0%) and 94.6% (89.4-99.7%) (p = 0.32).

CONCLUSIONS

Diagnostic performance of the AI algorithm that was trained on CPI did not drop on VMI, which is reassuring for its use in clinical practice.

CLINICAL RELEVANCE STATEMENT

A commercially available AI algorithm, trained on conventional polychromatic CTPA, could be safely used on virtual monochromatic images. This supports the sustainability of AI-aided detection of PE on CT despite ongoing technological advances in medical imaging, although monitoring in daily practice will remain important.

KEY POINTS

• Diagnostic accuracy of an AI algorithm trained on conventional polychromatic images to detect PE did not drop on virtual monochromatic images. • Our results are reassuring as innovations in hardware and reconstruction in CT are continuing, whilst commercial AI algorithms that are trained on older generation data enter healthcare.

Sensitivity, reliability and convergent validity of sequential dual-task measures of listening effort

OBJECTIVE

The aim of the current study was to assess the sensitivity, reliability and convergent validity of objective measures of listening effort collected in a sequential dual-task.

DESIGN

On each trial, participants viewed a set of digits and listened to a spoken sentence presented at one of a range of signal-to-noise ratios (SNR) and then typed the sentence-final word and recalled the digits. Listening effort measures included word response time, digit recall accuracy and digit response time. In Experiment 1, SNR on each trial was randomised. In Experiment 2, SNR varied in a blocked design, and in each block self-reported listening effort was also collected.

STUDY SAMPLES

Separate groups of 40 young adults participated in each experiment.

RESULTS

Effects of SNR were observed for all measures. Linear effects of SNR were generally observed even with word recognition accuracy factored out of the models. Among the objective measures, reliability was excellent, and repeated-measures correlations, though not between-subjects correlations, were nearly all significant.

CONCLUSION

The objective measures assessed appear to be sensitive and reliable indices of listening effort that are non-redundant with speech intelligibility and have strong within-participants convergent validity. Results support use of these measures in future studies of listening effort.

EEG-FMCNN: A fusion multi-branch 1D convolutional neural network for EEG-based motor imagery classification

Motor imagery (MI) electroencephalogram (EEG) signal is recognized as a promising paradigm for brain-computer interface (BCI) systems and has been extensively employed in various BCI applications, including assisting disabled individuals, controlling devices and environments, and enhancing human capabilities. The high-performance decoding capability of MI-EEG signals is a key issue that impacts the development of the industry. However, decoding MI-EEG signals is challenging due to the low signal-to-noise ratio and inter-subject variability. In response to the aforementioned core problems, this paper proposes a novel end-to-end network, a fusion multi-branch 1D convolutional neural network (EEG-FMCNN), to decode MI-EEG signals without pre-processing. The utilization of multi-branch 1D convolution not only exhibits a certain level of noise tolerance but also addresses the issue of inter-subject variability to some extent. This is attributed to the ability of multi-branch architectures to capture information from different frequency bands, enabling the establishment of optimal convolutional scales and depths. Furthermore, we incorporate 1D squeeze-and-excitation (SE) blocks and shortcut connections at appropriate locations to further enhance the generalization and robustness of the network. In the BCI Competition IV-2a dataset, our proposed model has obtained good experimental results, achieving accuracies of 78.82% and 68.41% for subject-dependent and subject-independent modes, respectively. In addition, extensive ablative experiments and fine-tuning experiments were conducted, resulting in a notable 7% improvement in the average performance of the network, which holds significant implications for the generalization and application of the network.

Effect of Breath Training on Image Quality of Chest Magnetic Resonance Free-breathing Sequence

BACKGROUND

Magnetic Resonance Imaging (MRI) plays a role in demonstrating substantial utility in lung lesion imaging, detection, diagnosis, and evaluation. Previous studies have found that free-breathing star VIBE sequences not only have high image quality but also have a high ability to detect and display nodules. However, in our routine clinical practice, we have encountered suboptimal image quality in the free-breathing sequences of certain patients.

OBJECTIVE

This study aims to assess the impact of breath training on the quality of chest magnetic resonance imaging obtained during free-breathing sequences.

METHODS

A total of 68 patients with lung lesions, such as nodules or masses detected via Computed Tomography (CT) examination, were prospectively gathered. They were then randomly divided into two groups: an observation group and a control group. Standard preparation was performed for all patients in both groups before the examination. The observation group underwent 30 minutes of breath training prior to the MRI examination additionally, followed by the acquisition of MRI free-breathing sequence images. The signal intensity (SI) and standard deviation (SD) of the lesion and adjacent normal lung tissue were measured, and the image signal-to-noise ratio (SNR) and contrast signal-to-noise ratio (CNR) of the lesion were calculated for objective image quality evaluation. The subjective image quality of the two groups of images was also evaluated using a 5-point method.

RESULTS

MRI examinations were completed in both groups. Significantly better subjective image quality (edge and internal structure clarity, vascular clarity, breathing and cardiac artifacts, and overall image quality) was achieved in the observation group compared to the control group (P<0.05). In addition, higher SNR and CNR values for disease lesions were observed in the observation group compared to the control group (t=4.35, P<0.05; t=5.35, P<0.05).

CONCLUSION

It is concluded that the image quality of free-breathing sequences MRI can be improved through breath training before examination.

Solving large-scale MEG/EEG source localisation and functional connectivity problems simultaneously using state-space models

State-space models are widely employed across various research disciplines to study unobserved dynamics. Conventional estimation techniques, such as Kalman filtering and expectation maximisation, offer valuable insights but incur high computational costs in large-scale analyses. Sparse inverse covariance estimators can mitigate these costs, but at the expense of a trade-off between enforced sparsity and increased estimation bias, necessitating careful assessment in low signal-to-noise ratio (SNR) situations. To address these challenges, we propose a three-fold solution: (1) Introducing multiple penalised state-space (MPSS) models that leverage data-driven regularisation; (2) Developing novel algorithms derived from backpropagation, gradient descent, and alternating least squares to solve MPSS models; (3) Presenting a K-fold cross-validation extension for evaluating regularisation parameters. We validate this MPSS regularisation framework through lower and more complex simulations under varying SNR conditions, including a large-scale synthetic magneto- and electro-encephalography (MEG/EEG) data analysis. In addition, we apply MPSS models to concurrently solve brain source localisation and functional connectivity problems for real event-related MEG/EEG data, encompassing thousands of sources on the cortical surface. The proposed methodology overcomes the limitations of existing approaches, such as constraints to small-scale and region-of-interest analyses. Thus, it may enable a more accurate and detailed exploration of cognitive brain functions.

Fast high-quality MRI protocol of the lumbar spine with deep learning-based algorithm: an image quality and scanning time comparison with standard protocol

OBJECTIVE

The objective of this study is to prospectively compare quantitative and subjective image quality, scanning time, and diagnostic confidence between a new deep learning-based reconstruction(DLR) algorithm and standard MRI protocol of lumbar spine.

MATERIALS AND METHODS

Eighty healthy volunteers underwent 1.5T MRI examination of lumbar spine from September 2021 to May 2023. Protocol acquisition comprised sagittal T1- and T2-weighted fast spin echo and short-tau inversion recovery images and axial multislices T2-weighted fast spin echo images. All sequences were acquired with both DLR algorithm and standard protocols. Two radiologists, blinded to the reconstruction technique, performed quantitative and qualitative image quality analysis in consensus reading; diagnostic confidence was also assessed. Quantitative image quality analysis was assessed by calculating signal-to-noise ratio (SNR) and contrast-to-noise ratio (CNR). Qualitative image quality analysis and diagnostic confidence were assessed with a five-point Likert scale. Scanning times were also compared.

RESULTS

DLR SNR was higher in all sequences (all p<0.001). CNR of the DLR was superior to conventional dataset only for axial and sagittal T2-weighted fast spin echo images (p<0.001). Qualitative analysis showed DLR had higher overall quality in all sequences (all p<0.001), with an inter-rater agreement of 0.83 (0.78-0.86). DLR total protocol scanning time was lower compared to standard protocol (6:26 vs 12:59 min, p<0.001). Diagnostic confidence for DLR algorithm was not inferior to standard protocol.

CONCLUSION

DLR applied to 1.5T MRI is a feasible method for lumbar spine imaging providing morphologic sequences with higher image quality and similar diagnostic confidence compared with standard protocol, enabling a remarkable time saving (up to 50%).

Saving Contrast Media in Coronary CT Angiography with Photon-Counting Detector CT

RATIONALE AND OBJECTIVES

To determine the optimal virtual monoenergetic image (VMI) energy level and the potential of contrast-media (CM) reduction for coronary computed tomography angiography (CCTA) with photon-counting detector CT (PCD-CT).

MATERIALS AND METHODS

In this institutional review board-approved study, patients who underwent CCTA with dual-source PCD-CT with an identical scan protocol and radiation dose were included. In group 1, CCTA was performed with our standard CM protocol (volume: 72-85.2 mL, 370 mg iodine/mL). VMIs were reconstructed from 40 to 60 keV at 5 keV increments. Objective image quality (IQ) (vascular attenuation, image noise, and contrast-to-noise ratio [CNR]) was measured. Two blinded, independent readers rated subjective IQ (overall IQ, subjective image contrast, and subjective noise using a five-point discrete visual scale). Results of group 1 served to determine the best VMI level for CCTA. In group 2, CM volume was reduced by 20%, and in group 3 by another 20%.

RESULTS

A total of 100 patients were enrolled (45 females, mean age 54 ± 13 years). Inter-reader agreement was good-to-excellent for all comparisons (κ > 0.6). In group 1, the best VMI level regarding objective and subjective IQ was 45 keV, which was selected as the reference for groups 2 and 3. For group 2, mean vascular attenuation was 890 Hounsfield units (HU) and mean CNR was 26, with no differences compared to group 1, 45 keV for both objective and subjective IQ. For group 3, mean vascular attenuation was 676 HU and mean CNR was 21, and all patients were rated as diagnostic except one (severe motion artifacts).

CONCLUSION

Increased IQ of PCD-CT can be used for considerable CM volume reduction while still maintaining a diagnostic IQ of CCTA.

Accelerating acquisition of readout-segmented echo planar imaging (rs-EPI) with a simultaneous multislice (SMS) technique for diffusion-weighted (DW) breast MRI: Evaluation of image quality factors

PURPOSE

This study aims to compare the image quality, apparent diffusion coefficient (ADC), signal-to-noise ratio (SNR), contrast-to-noise ratio (CNR) values, and scan time between readout-segmented echo planar imaging (rs-EPI) and simultaneous multislice (SMS) rs-EPI sequences.

METHODS

A total of 80 consecutive women who underwent breast diffusion-weighted imaging (DWI) were included, and two rs-EPI DWI sequences with and without SMS were acquired and compared. Qualitative analysis involved three radiologists independently scoring image quality and radiologist preference. For quantitative comparison, the radiologists independently measured the ADC values in patients, while SNR, CNR, and ADC values were measured on a phantom.

RESULTS

The acquisition time was 5:47 min for rs-EPI and 3:20 min for SMS rs-EPI. In terms of image quality, scores were similar between rs-EPI and SMS rs-EPI sequences in the pooled data set, with the exception of skin-line distinction (p = 0.001) and background noise (p < 0.001). All radiologists considered SMS rs-EPI as equal or superior to rs-EPI in more than 70 % of cases. SMS rs-EPI demonstrated significantly higher ADC values than rs-EPI by all radiologists (p ≤ 0.002). For the phantom measurement, ADC (SMS: 1.26 ± 0.68 and RS: 1.26 ± 0.68, p = 0.198), SNR (SMS: 540.6 ± 342.1 and RS: 558.8 ± 523.2, p = 0.927), and CNR (SMS: 235.5 ± 38.9 and RS: 252.8 ± 108.0, p = 0.784) values did not significantly differ between the two sequences.

CONCLUSION

SMS rs-EPI exhibited comparable image quality and similar ADC, SNR, and CNR values to rs-EPI while reducing the scan time by 42%.

Optimization-based modeling of Lombard speech articulation: Supraglottal characteristics

This paper shows that a highly simplified model of speech production based on the optimization of articulatory effort versus intelligibility can account for some observed articulatory consequences of signal-to-noise ratio. Simulations of static vowels in the presence of various background noise levels show that the model predicts articulatory and acoustic modifications of the type observed in Lombard speech. These features were obtained only when the constraint applied to articulatory effort decreases as the level of background noise increases. These results support the hypothesis that Lombard speech is listener oriented and speakers adapt their articulation in noisy environments.

Influence mechanism of the temporal duration of laser irradiation on photoacoustic technique: a review

Significance

In the photoacoustic (PA) technique, the laser irradiation in the time domain (i.e., laser pulse duration) governs the characteristics of PA imaging-it plays a crucial role in the optical-acoustic interaction, the generation of PA signals, and the PA imaging performance.

Aim

We aim to provide a comprehensive analysis of the impact of laser pulse duration on various aspects of PA imaging, encompassing the signal-to-noise ratio, the spatial resolution of PA imaging, the acoustic frequency spectrum of the acoustic wave, the initiation of specific physical phenomena, and the photothermal-PA (PT-PA) interaction/conversion.

Approach

By surveying and reviewing the state-of-the-art investigations, we discuss the effects of laser pulse duration on the generation of PA signals in the context of biomedical PA imaging with respect to the aforementioned aspects.

Results

First, we discuss the impact of laser pulse duration on the PA signal amplitude and its correlation with the lateral resolution of PA imaging. Subsequently, the relationship between the axial resolution of PA imaging and the laser pulse duration is analyzed with consideration of the acoustic frequency spectrum. Furthermore, we examine the manipulation of the pulse duration to trigger physical phenomena and its relevant applications. In addition, we elaborate on the tuning of the pulse duration to manipulate the conversion process and ratio from the PT to PA effect.

Conclusions

We contribute to the understanding of the physical mechanisms governing pulse-width-dependent PA techniques. By gaining insight into the mechanism behind the influence of the laser pulse, we can trigger the pulse-with-dependent physical phenomena for specific PA applications, enhance PA imaging performance in biomedical imaging scenarios, and modulate PT-PA conversion by tuning the pulse duration precisely.

Deuterium metabolic imaging for 3D mapping of glucose metabolism in humans with central nervous system lesions at 3T

PURPOSE

To explore the potential of 3T deuterium metabolic imaging (DMI) using a birdcage 2 H radiofrequency (RF) coil in both healthy volunteers and patients with central nervous system (CNS) lesions.

METHODS

A modified gradient filter, home-built 2 H volume RF coil, and spherical k-space sampling were employed in a three-dimensional chemical shift imaging acquisition to obtain high-quality whole-brain metabolic images of 2 H-labeled water and glucose metabolic products. These images were acquired in a healthy volunteer and three subjects with CNS lesions of varying pathologies. Hardware and pulse sequence experiments were also conducted to improve the signal-to-noise ratio of DMI at 3T.

RESULTS

The ability to quantify local glucose metabolism in correspondence to anatomical landmarks across patients with varying CNS lesions is demonstrated, and increased lactate is observed in one patient with the most active disease.

CONCLUSION

DMI offers the potential to examine metabolic activity in human subjects with CNS lesions with DMI at 3T, promising for the potential of the future clinical translation of this metabolic imaging technique.

Optimal CT windowing on low-monoenergetic images using a simplex algorithm-based approach for abdominal inflammatory processes

BACKGROUND

OBJECTIVES: To determine optimal window settings for conspicuity of abdominal inflammatory processes on 50 keV low-monoenergetic images derived from dual-energy spectral CT (DECT).

METHODS

A retrospective study of 30 patients with clinically proven pancreatitis (15/30) or pyelonephritis (15/30) with inflammatory lesions visible on DECT scans were selected to serve as reference populations. 50 keV low-monoenergetic images in the portal venous phase were iteratively evaluated by 6 abdominal radiologists in twenty-one different windows (7-350HU center; 120-580HU width), selected using a simplex optimization algorithm. Each reader graded the conspicuity of the parenchymal hypodense lesions and image background quality. Three-dimensional contour maps expressing the relationship between overall reader grade and window center and width were constructed and used to find the ideal window for inflammatory pancreatic and renal processes and the image background quality. Finally, 15 appendicitis cases were reviewed on optimal pancreas and kidney windows and the manufacturer recommended conventional abdominal window settings for conventional imaging.

RESULTS

Convergence to optimal windowing was achieved based upon a total of 3,780 reads (21 window settings × 6 readers × 15 cases for pancreas and kidney). Highest conspicuity grade (>4.5 ± 0.0) for pancreas inflammatory lesions was seen at 116HU/430HU, whereas hypodense pyelonephritis had highest conspicuity at 290HU/570HU. This rendered an ideal "compromise" window (>4 ± 0.2) of 150HU/450HU which differed substantially from conventional manufacturer recommended settings of 50HU/380HU (2.1 ± 1.0, p = 0.00001). Appendix mucosal enhancement was best visualized at manufacturer settings.

CONCLUSIONS

Optimal visualization of inflammatory processes in abdominal organs on 50 keV low-monoenergetic images may require tailored refinement of window settings.

Utilizing deep learning techniques to improve image quality and noise reduction in preclinical low-dose PET images in the sinogram domain

BACKGROUND

Low-dose positron emission tomography (LD-PET) imaging is commonly employed in preclinical research to minimize radiation exposure to animal subjects. However, LD-PET images often exhibit poor quality and high noise levels due to the low signal-to-noise ratio. Deep learning (DL) techniques such as generative adversarial networks (GANs) and convolutional neural network (CNN) have the capability to enhance the quality of images derived from noisy or low-quality PET data, which encodes critical information about radioactivity distribution in the body.

PURPOSE

Our objective was to optimize the image quality and reduce noise in preclinical PET images by utilizing the sinogram domain as input for DL models, resulting in improved image quality as compared to LD-PET images.

METHODS

A GAN and CNN model were utilized to predict high-dose (HD) preclinical PET sinograms from the corresponding LD preclinical PET sinograms. In order to generate the datasets, experiments were conducted on micro-phantoms, animal subjects (rats), and virtual simulations. The quality of DL-generated images was weighted by performing the following quantitative measures: structural similarity index measure (SSIM), root mean squared error (RMSE), peak signal-to-noise ratio (PSNR), signal-to-noise ratio (SNR), and contrast-to-noise ratio (CNR). Additionally, DL input and output were both subjected to a spatial resolution calculation of full width half maximum (FWHM) and full width tenth maximum (FWTM). DL outcomes were then compared with the conventional denoising algorithms such as non-local means (NLM), block-matching, and 3D filtering (BM3D).

RESULTS

The DL models effectively learned image features and produced high-quality images, as reflected in the quantitative metrics. Notably, the FWHM and FWTM values of DL PET images exhibited significantly improved accuracy compared to LD, NLM, and BM3D PET images, and just as precise as HD PET images. The MSE loss underscored the excellent performance of the models, indicating that the models performed well. To further improve the training, the generator loss (G loss) was increased to a value higher than the discriminator loss (D loss), thereby achieving convergence in the GAN model.

CONCLUSIONS

The sinograms generated by the GAN network closely resembled real HD preclinical PET sinograms and were more realistic than LD. There was a noticeable improvement in image quality and noise factor in the predicted HD images. Importantly, DL networks did not fully compromise the spatial resolution of the images.

Signal-to-noise ratio to assess magnitude, kinetics and impact on pharmacokinetics of the immune response to an adalimumab biosimilar

Background: The antidrug antibody (ADA) signal-to-noise (S/N) ratio was explored as a novel immunogenicity measure to evaluate the immune response of healthy subjects to a single dose of GP2017, an adalimumab biosimilar. Methodology/results: Bioanalytical methods used for the analysis of ADA S/N ratios and ADA titers were validated for sensitivity, precision and drug interference. ADA S/N ratios strongly correlated with ADA titers. Correlations between ADA area under the curve and ADAmax and pharmacokinetics (PK) were stronger for ADA S/N ratio than for ADA titers. Conclusion: ADA S/N ratio allowed for a more sensitive evaluation of the magnitude and kinetics of the immune response, was better correlated with adalimumab PK and was superior to ADA titers in assessing the impact of the immune response on PK.

Deep learning-based ultra-fast identification of Raman spectra with low signal-to-noise ratio

Ensuring the correct use of cell lines is crucial to obtaining reliable experimental results and avoiding unnecessary waste of resources. Raman spectroscopy has been confirmed to be able to identify cell lines, but the collection time is usually 10-30 s. In this study, we acquired Raman spectra of five cell lines with integration times of 0.1 and 8 s, respectively, and the average accuracy of using long-short memory neural network to identify the spectra of 0.1 s was 95%, and the average accuracy of identifying the spectra of 8 s was 99.8%. At the same time, we performed data enhancement of 0.1 s spectral data by real-valued non-volume preserving method, and the recognition average accuracy of long-short memory neural networks recognition of the enhanced spectral data was improved to 96.2%. With this method, we shorten the acquisition time of Raman spectra to 1/80 of the original one, which greatly improves the efficiency of cell identification.

Sub-aperture ultrafast volumetric ultrasound imaging for fully sampled dual-mode matrix array

Fully sampled dual-mode matrix array ultrasound transducer is capable of performing imaging and therapeutic ultrasound in three dimensions (3D). It is a promising tool for many clinical applications because of its precise multi-focus therapy with imaging guidance by itself. Our team previously designed a 256-element fully sampled dual-mode matrix array transducer, while its imaging quality needs to be further improved. In this work, we propose a high-contrast sub-aperture volumetric imaging strategy to improve the imaging quality of the dual-mode matrix array. We first analyzed the effect of various parameters of sub-aperture imaging on the imaging quality by Field II. Based on the optimized parameters, we compared the resolution and signal to noise ratio (SNR) of sub-aperture imaging with those of full aperture imaging on phantoms and rabbit brain. The experimental results showed the proposed sub-aperture imaging method could obtain a comparable resolution to full aperture imaging. Moreover, the average intensity of noise signal near the wire phantom decreased by about 5 dB and the SNR of tissue phantom image increased by 8 %. The proposed sub-aperture imaging method also enabled clearer and more accurate imaging of the rabbit brain. The obtained results indicate the proposed sub-aperture imaging is a promising method for practical use of a fully sampled dual-mode matrix array for volumetric ultrasound imaging.

APNet: Adaptive projection network for medical image denoising

BACKGROUND

In clinical medicine, low-dose radiographic image noise reduces the quality of the detected image features and may have a negative impact on disease diagnosis.

OBJECTIVE

In this study, Adaptive Projection Network (APNet) is proposed to reduce noise from low-dose medical images.

METHODS

APNet is developed based on an architecture of the U-shaped network to capture multi-scale data and achieve end-to-end image denoising. To adaptively calibrate important features during information transmission, a residual block of the dual attention method throughout the encoding and decoding phases is integrated. A non-local attention module to separate the noise and texture of the image details by using image adaptive projection during the feature fusion.

RESULTS

To verify the effectiveness of APNet, experiments on lung CT images with synthetic noise are performed, and the results demonstrate that the proposed approach outperforms recent methods in both quantitative index and visual quality. In addition, the denoising experiment on the dental CT image is also carried out and it verifies that the network has a certain generalization.

CONCLUSIONS

The proposed APNet is an effective method that can reduce image noise and preserve the required image details in low-dose radiographic images.

Two-and-a-half order score-based model for solving 3D ill-posed inverse problems

Computed Tomography (CT) and Magnetic Resonance Imaging (MRI) are crucial technologies in the field of medical imaging. Score-based models demonstrated effectiveness in addressing different inverse problems encountered in the field of CT and MRI, such as sparse-view CT and fast MRI reconstruction. However, these models face challenges in achieving accurate three dimensional (3D) volumetric reconstruction. The existing score-based models predominantly concentrate on reconstructing two-dimensional (2D) data distributions, resulting in inconsistencies between adjacent slices in the reconstructed 3D volumetric images. To overcome this limitation, we propose a novel two-and-a-half order score-based model (TOSM). During the training phase, our TOSM learns data distributions in 2D space, simplifying the training process compared to working directly on 3D volumes. However, during the reconstruction phase, the TOSM utilizes complementary scores along three directions (sagittal, coronal, and transaxial) to achieve a more precise reconstruction. The development of TOSM is built on robust theoretical principles, ensuring its reliability and efficacy. Through extensive experimentation on large-scale sparse-view CT and fast MRI datasets, our method achieved state-of-the-art (SOTA) results in solving 3D ill-posed inverse problems, averaging a 1.56 dB peak signal-to-noise ratio (PSNR) improvement over existing sparse-view CT reconstruction methods across 29 views and 0.87 dB PSNR improvement over existing fast MRI reconstruction methods with × 2 acceleration. In summary, TOSM significantly addresses the issue of inconsistency in 3D ill-posed problems by modeling the distribution of 3D data rather than 2D distribution which has achieved remarkable results in both CT and MRI reconstruction tasks.

Evaluation of image quality and radiation dose in computed tomography urography following tube voltage optimisation

INTRODUCTION

Computed tomography urography (CTU) comprehensively evaluates the urinary tract. However, the procedure is associated with a high radiation dose due to multiple scan series and therefore requires optimisation. The study performed CTU protocol optimisation based on a reduction in tube voltage (kV) using quality assurance (QA) phantom and clinical images and evaluated image quality and radiation dose.

METHODS

The study was prospectively conducted on patients referred for CTU. The patients were grouped into A and B and were scanned with the standard protocol, a protocol used for the routine CTU at the CT centre before optimisation, and optimised protocol, a protocol with reduced kV respectively. The protocols were first tried on a quality assurance (QA) phantom before being applied to patients, and image quality was assessed based on signal-to-noise ratio (SNR) and contrast-to-noise ratio (CNR). In addition, the clinical images were assessed based on the visibility of the anatomical criteria for CT images by five observers with >5 years of experience. The data were analysed using both visual grading characteristic (VGC) curves and statistical package for social sciences (SPSS) version 22.0.

RESULTS

The dose was significantly lower in the optimised protocol with a 10 % reduction in both volume computed tomography dose index and (CTDIvol) and dose length product (DLP) for the phantom images, and a 26 % reduction in CTDIvol and 28 % in DLP for the clinical images. However, there was no significant difference in image quality noted between the standard and optimised protocols based on the quantitative and qualitative image quality evaluation using both the QA phantom and clinical images.

CONCLUSION

The findings revealed a significant dose reduction in the optimised protocol. Further, image quality in standard and optimised protocols did not differ significantly based on quantitative and qualitative methods.

IMPLICATION FOR PRACTICE

kV optimisation in contrast-enhanced procedures provides dose reduction and should be encouraged in the medical imaging departments.

Application of dual-energy CT with prospective ECG-gating in cardiac CT angiography for children: Radiation and contrast agent dose

OBJECTIVE

This research aimed to investigate the feasibility of utilizing dual-energy CT virtual monoenergetic images (VMI1) with prospective electrocardiogram (ECG2) gating for reducing radiation and contrast agent doses in pediatric patients with congenital heart disease (CHD3).

METHODS

There were 100 pediatric patients with CHD included in this study. Group A (n = 50) underwent dual-energy scanning with prospective ECG-gating, and group B (n = 50) underwent conventional scanning with retrospective ECG-gating. Comparative analysis of CT values of lumen, objective image quality assessment, subjective image quality evaluations, and diagnostic efficacy were performed.

RESULTS

CT values, image noise, signal-to-noise ratio (SNR4), and contrast-to-noise ratio (CNR5) were significantly affected by the VMI energy level, and they all increased with decreasing energy levels (P > 0.05). Combining subjective evaluation, the 45 keV VMI was considered the optimum image in group A. The 45 keV VMI exhibited higher CT values of lumen compared to conventional scanning images (P < 0.003 ∼ 0.836), but meanwhile, the image noise was also higher in the 45 keV VMI (P = 0.004). Differences between the two groups in SNR, CNR, and diagnostic accuracy were not statistically significant. Compared to group B, the 45 keV VMI showed fewer contrast-induced artifacts (P < 0.001) and higher image quality score (P = 0.037). Group A had a 64 % reduction in radiation dose and a 40 % decrease in iodine dose compared to group B.

CONCLUSION

The combination of dual-energy CT with prospective ECG-gating reduces radiation and iodine doses in pediatric patients with CHD. The 45 keV VMI can provide clinically acceptable image quality while declining contrast agent artifacts.

Electrode Montage Induced Changes in Air-Conducted Ocular Vestibular-Evoked Myogenic Potential

OBJECTIVES

Stimulus and recording parameters are pivotal for shaping the ocular vestibular-evoked myogenic potential (oVEMP). In the last decade, several attempts were made to identify the optimum electrode placement site to improve the oVEMP responses. A vast majority of these found larger response amplitudes for alternate electrode montages like belly-tendon (BT), chin-referenced (CR), and/or sternum-referenced montages than the clinically used infra-orbital montage. However, no study has yet compared all alternate electrode montages in a simultaneous recording paradigm to eliminate other confounding factors. Also, no study has compared all of them for their test-retest reliability, waveform morphology, and signal-to-noise ratio. Therefore, the decision on which among these electrode montages is best suited for oVEMP acquisition remains opaque. The present study aimed to investigate the effects of various electrode montages on oVEMP's response parameters and to determine the test-retest reliability of each of these in clinically healthy individuals using a simultaneous recording paradigm.

DESIGN

This study had a within-subject experimental design. Fifty-five young healthy adults (age range: 20-30 years) underwent contralateral oVEMP recording using infra-orbital, BT, chin-referenced, and sternum-referenced electrode montages simultaneously using a four-channel evoked potential system.

RESULTS

BT montage had a significantly shorter latency, larger amplitude, higher signal-to-noise ratio, and better morphology than other alternate montages ( p < 0.008). Further, all electrode montages of the current study showed fair/moderate to excellent test-retest reliability.

CONCLUSIONS

By virtue of producing significantly better response parameters than the other electrode montages, BT montage seems better suited to the recording of oVEMP than the known electrode montages thus far.

PET image denoising based on denoising diffusion probabilistic model

PURPOSE

Due to various physical degradation factors and limited counts received, PET image quality needs further improvements. The denoising diffusion probabilistic model (DDPM) was a distribution learning-based model, which tried to transform a normal distribution into a specific data distribution based on iterative refinements. In this work, we proposed and evaluated different DDPM-based methods for PET image denoising.

METHODS

Under the DDPM framework, one way to perform PET image denoising was to provide the PET image and/or the prior image as the input. Another way was to supply the prior image as the network input with the PET image included in the refinement steps, which could fit for scenarios of different noise levels. 150 brain [[Formula: see text]F]FDG datasets and 140 brain [[Formula: see text]F]MK-6240 (imaging neurofibrillary tangles deposition) datasets were utilized to evaluate the proposed DDPM-based methods.

RESULTS

Quantification showed that the DDPM-based frameworks with PET information included generated better results than the nonlocal mean, Unet and generative adversarial network (GAN)-based denoising methods. Adding additional MR prior in the model helped achieved better performance and further reduced the uncertainty during image denoising. Solely relying on MR prior while ignoring the PET information resulted in large bias. Regional and surface quantification showed that employing MR prior as the network input while embedding PET image as a data-consistency constraint during inference achieved the best performance.

CONCLUSION

DDPM-based PET image denoising is a flexible framework, which can efficiently utilize prior information and achieve better performance than the nonlocal mean, Unet and GAN-based denoising methods.

An organ-effective modulation for non-contrast chest computed tomography imaging: effect on image quality and thyroid exposure reduction

We investigate the efficacy of organ-effective modulation (OEM) technique for thyroid dose reduction among various body habitus and its impact on image quality in chest non-contrast computed tomography (CT). We prospectively enrolled 64 patients who underwent non-contrast chest CT from January to May 2022. The skin-absorbed radiation dose over the thyroid (Dthyroid) was obtained using a thermoluminescence dosemeter. Signal-to-noise ratio and image noise was also quantitatively assessed. In subjective analyses, two radiologists independently evaluated images based on a 5-point scale. The OEM group showed a markedly decrease in Dthyroid when compared with the non-OEM group (p < 0.05). No significant difference was observed regarding the image noise (p < 0.05), except for the ventral air space. The subjective scores of two radiologists showed no significant differences between the non-OEM and OEM groups. OEM can effectively reduce the radiation exposure of thyroid without compromising on image quality in non-contrast chest CT.

DERnet: a deep neural network for end-to-end reconstruction in magnetic particle imaging

Objective. Magnetic particle imaging (MPI) shows potential for contributing to biomedical research and clinical practice. However, MPI images are effectively affected by noise in the signal as its reconstruction is an ill-posed inverse problem. Thus, effective reconstruction method is required to reduce the impact of the noise while mapping signals to MPI images. Traditional methods rely on the hand-crafted data-consistency (DC) term and regularization term based on spatial priors to achieve noise-reducing and reconstruction. While these methods alleviate the ill-posedness and reduce noise effects, they may be difficult to fully capture spatial features.Approach. In this study, we propose a deep neural network for end-to-end reconstruction (DERnet) in MPI that emulates the DC term and regularization term using the feature mapping subnetwork and post-processing subnetwork, respectively, but in a data-driven manner. By doing so, DERnet can better capture signal and spatial features without relying on hand-crafted priors and strategies, thereby effectively reducing noise interference and achieving superior reconstruction quality.Main results. Our data-driven method outperforms the state-of-the-art algorithms with an improvement of 0.9-8.8 dB in terms of peak signal-to-noise ratio under various noise levels. The result demonstrates the advantages of our approach in suppressing noise interference. Furthermore, DERnet can be employed for measured data reconstruction with improved fidelity and reduced noise. In conclusion, our proposed method offers performance benefits in reducing noise interference and enhancing reconstruction quality by effectively capturing signal and spatial features.Significance. DERnet is a promising candidate method to improve MPI reconstruction performance and facilitate its more in-depth biomedical application.

Dual contrast attention-guided multi-frequency fusion for multi-contrast MRI super-resolution

Objective. Multi-contrast magnetic resonance (MR) imaging super-resolution (SR) reconstruction is an effective solution for acquiring high-resolution MR images. It utilizes anatomical information from auxiliary contrast images to improve the quality of the target contrast images. However, existing studies have simply explored the relationships between auxiliary contrast and target contrast images but did not fully consider different anatomical information contained in multi-contrast images, resulting in texture details and artifacts unrelated to the target contrast images.Approach. To address these issues, we propose a dual contrast attention-guided multi-frequency fusion (DCAMF) network to reconstruct SR MR images from low-resolution MR images, which adaptively captures relevant anatomical information and processes the texture details and low-frequency information from multi-contrast images in parallel. Specifically, after the feature extraction, a feature selection module based on a dual contrast attention mechanism is proposed to focus on the texture details of the auxiliary contrast images and the low-frequency features of the target contrast images. Then, based on the characteristics of the selected features, a high- and low-frequency fusion decoder is constructed to fuse these features. In addition, a texture-enhancing module is embedded in the high-frequency fusion decoder, to highlight and refine the texture details of the auxiliary contrast and target contrast images. Finally, the high- and low-frequency fusion process is constrained by integrating a deeply-supervised mechanism into the DCAMF network.Main results. The experimental results show that the DCAMF outperforms other state-of-the-art methods. The peak signal-to-noise ratio and structural similarity of DCAMF are 39.02 dB and 0.9771 on the IXI dataset and 37.59 dB and 0.9770 on the BraTS2018 dataset, respectively. The image recovery is further validated in segmentation tasks.Significance. Our proposed SR model can enhance the quality of MR images. The results of the SR study provide a reliable basis for clinical diagnosis and subsequent image-guided treatment.

Clutter filtering of angular domain data for contrast-free ultrafast microvascular imaging

Objective. Contrast-free microvascular imaging is clinically valuable for the assessment of physiological status and the early diagnosis of diseases. Effective clutter filtering is essential for microvascular visualization without contrast enhancement. Singular value decomposition (SVD)-based spatiotemporal filter has been widely used to suppress clutter. However, clinical real-time imaging relies on short ensembles (dozens of frames), which limits the implementation of SVD filtering due to the large error of eigen-correlated estimations and high dependence on optimal threshold when used in such ensembles.Approach. To address the above challenges of imaging in short ensembles, two optimized filters of angular domain data are proposed in this paper: grouped angle SVD (GA-SVD) and angular-coherence-based higher-order SVD (AC-HOSVD). GA-SVD applies SVD to the concatenation of all angular data to improve clutter rejection performance in short ensembles, while AC-HOSVD applies HOSVD to the angular data tensor and utilizes angular coherence in addition to spatial and temporal features for filtering. Feasible threshold selection strategies in each feature space are provided. The clutter rejection performance of the proposed filters and SVD was evaluated with Doppler phantom andin vivostudies at different cases. Moreover, the robustness of the filters was explored under wrong singular value threshold estimation, and their computational complexity was studied.Main results. Qualitative and quantitative results indicated that GA-SVD and AC-HOSVD can effectively improve clutter rejection performance in short ensembles, especially AC-HOSVD. Notably, the proposed methods using 20 frames had similar image quality to SVD using 100 frames.In vivostudies showed that compared to SVD, GA-SVD increased the signal-to-noise-ratio (SNR) by 6.03 dB on average, and AC-HOSVD increased the SNR by 8.93 dB on average. Furthermore, AC-HOSVD remained better power Doppler image quality under non-optimal thresholds, followed by GA-SVD.Significance. The proposed filters can greatly enhance contrast-free microvascular visualization in short ensembles and have potential for different clinical translations due to the performance differences.

Adaptive frequency-domain filtering for neural signal preprocessing

Electrical interference from various sources is a common issue for experimental extracellular electrophysiology recordings collected using multi-electrode array neural recording systems. This interference deteriorates the signal-to-noise ratio (SNR) of the raw electrophysiology signals and hampers the accuracy of data post-processing using techniques such as spike-sorting. Traditional signal processing methods to digitally remove electrical interference during post-processing include bandpass filtering to limit the signal to the relevant spectral range of the biological data, e.g., the spikes band (300 Hz - 7 kHz), targeted notch filtering to remove power line interference from standard alternating current mains electricity and common reference removal to minimize noise common to all electrodes. These methods require a priori knowledge of the frequency of the interfering signal source to address the unique electromagnetic interference environment of each experimental setup. We discuss an adaptive method for automatically removing narrow-band electrical interference through a spectral peak detection and removal (SPDR) step that can be applied during post-processing of the recorded data, based on the intuition that tall, narrowband signals localized in the signal spectrum correspond to interference, rather than the activity of neurons. A spectral peak prominence (SPP) threshold is used to detect these peaks in the frequency domain, which will then be removed via notch filtering. We applied this method to simulated waveforms and also experimental electrophysiology data collected from cerebral organoids to demonstrate its effectiveness for removing unwanted interference without significantly distorting the neural signals. We discuss that proper selection of the SPP threshold is required to avoid over-filtering, which can result in distortion of the electrophysiology data. We also compare the firing-rate activity in the filtered electrophysiology with fluorescence calcium imaging, a secondary cellular activity marker, to quantify signal distortion and provide bounds on SNR-based optimization of the SPP threshold. The adaptive filtering technique demonstrated in this paper is a powerful method that can automatically detect and remove interband interference in recorded neural signals, potentially enabling data collection in more naturalistic settings where external interference signals are difficult to eliminate.

Learnable PM diffusion coefficients and reformative coordinate attention network for low dose CT denoising

Objective.Various deep learning methods have recently been used for low dose CT (LDCT) denoising. Aggressive denoising may destroy the edge and fine anatomical structures of CT images. Therefore a key issue in LDCT denoising tasks is the difficulty of balancing noise/artifact suppression and edge/structure preservation.Approach.We proposed an LDCT denoising network based on the encoder-decoder structure, namely the Learnable PM diffusion coefficient and efficient attention network (PMA-Net). First, using the powerful feature modeling capability of partial differential equations, we constructed a multiple learnable edge module to generate precise edge information, incorporating the anisotropic image processing idea of Perona-Malik (PM) model into the neural network. Second, a multiscale reformative coordinate attention module was designed to extract multiscale information. Non-overlapping dilated convolution capturing abundant contextual content was combined with coordinate attention which could embed the spatial location information of important features into the channel attention map. Finally, we imposed additional constraints on the edge information using edge-enhanced multiscale perceptual loss to avoid structure loss and over-smoothing.Main results.Experiments are conducted on simulated and real datasets. The quantitative and qualitative results show that the proposed method has better performance in suppressing noise/artifacts and preserving edges/structures.Significance.This work proposes a novel edge feature extraction method that unfolds partial differential equation into neural networks, which contributes to the interpretability and clinical application value of neural network.

Low-Field-Strength Body MRI: Challenges and Opportunities at 0.55 T

Advances in MRI technology have led to the development of low-field-strength (hereafter, "low-field") (0.55 T) MRI systems with lower weight, fewer shielding requirements, and lower cost than those of traditional (1.5-3 T) systems. The trade-offs of lower signal-to-noise ratio (SNR) at 0.55 T are partially offset by patient safety and potential comfort advantages (eg, lower specific absorption rate and a more cost-effective larger bore diameter) and physical advantages (eg, decreased T2* decay, shorter T1 relaxation times). Image reconstruction advances leveraging developing technologies (such as deep learning-based denoising) can be paired with traditional techniques (such as increasing the number of signal averages) to improve SNR. The overall image quality produced by low-field MRI systems, although perhaps somewhat inferior to 1.5-3 T MRI systems in terms of SNR, is nevertheless diagnostic for a broad variety of body imaging applications. Effective low-field body MRI requires (a) an understanding of the trade-offs resulting from lower field strengths, (b) an approach to modifying routine sequences to overcome SNR challenges, and (c) a workflow for carefully selecting appropriate patients. The authors describe the rationale, opportunities, and challenges of low-field body MRI; discuss important considerations for low-field imaging with common body MRI sequences; and delineate a variety of use cases for low-field body MRI. The authors also include lessons learned from their preliminary experience with a new low-field MRI system at a tertiary care center. Finally, they explore the future of low-field MRI, summarizing current limitations and potential future developments that may enhance the clinical adoption of this technology. ©RSNA, 2023 Supplemental material is available for this article. Quiz questions for this article are available through the Online Learning Center. See the invited commentary by Venkatesh in this issue.

Super-resolved q-space learning of diffusion MRI

BACKGROUND

Diffusion magnetic resonance imaging (dMRI) provides a powerful tool to non-invasively investigate neural structures in the living human brain. Nevertheless, its reconstruction performance on neural structures relies on the number of diffusion gradients in the q-space. High-angular (HA) dMRI requires a long scan time, limiting its use in clinical practice, whereas directly reducing the number of diffusion gradients would lead to the underestimation of neural structures.

PURPOSE

We propose a deep compressive sensing-based q-space learning (DCS-qL) approach to estimate HA dMRI from low-angular dMRI.

METHODS

In DCS-qL, we design the deep network architecture by unfolding the proximal gradient descent procedure that addresses the compressive sense problem. In addition, we exploit a lifting scheme to design a network structure with reversible transform properties. For implementation, we apply a self-supervised regression to enhance the signal-to-noise ratio of diffusion data. Then, we utilize a semantic information-guided patch-based mapping strategy for feature extraction, which introduces multiple network branches to handle patches with different tissue labels.

RESULTS

Experimental results show that the proposed approach can yield a promising performance on the tasks of reconstructed HA dMRI images, microstructural indices of neurite orientation dispersion and density imaging, fiber orientation distribution, and fiber bundle estimation.

CONCLUSIONS

The proposed method achieves more accurate neural structures than competing approaches.

Bio-friendly long-term subcellular dynamic recording by self-supervised image enhancement microscopy

Fluorescence microscopy has become an indispensable tool for revealing the dynamic regulation of cells and organelles. However, stochastic noise inherently restricts optical interrogation quality and exacerbates observation fidelity when balancing the joint demands of high frame rate, long-term recording and low phototoxicity. Here we propose DeepSeMi, a self-supervised-learning-based denoising framework capable of increasing signal-to-noise ratio by over 12 dB across various conditions. With the introduction of newly designed eccentric blind-spot convolution filters, DeepSeMi effectively denoises images with no loss of spatiotemporal resolution. In combination with confocal microscopy, DeepSeMi allows for recording organelle interactions in four colors at high frame rates across tens of thousands of frames, monitoring migrasomes and retractosomes over a half day, and imaging ultra-phototoxicity-sensitive Dictyostelium cells over thousands of frames. Through comprehensive validations across various samples and instruments, we prove DeepSeMi to be a versatile and biocompatible tool for breaking the shot-noise limit.

A Paired Phase and Magnitude Reconstruction for Advanced Diffusion-Weighted Imaging

OBJECTIVE

Multi-shot interleaved echo planer imaging (Ms-iEPI) can obtain diffusion-weighted images (DWI) with high spatial resolution and low distortion, but suffers from ghost artifacts introduced by phase variations between shots. In this work, we aim at solving the ms-iEPI DWI reconstructions under inter-shot motions and ultra-high b-values.

METHODS

An iteratively joint estimation model with paired phase and magnitude priors is proposed to regularize the reconstruction (PAIR). The former prior is low-rankness in the k-space domain. The latter explores similar edges among multi-b-value and multi-direction DWI with weighted total variation in the image domain. The weighted total variation transfers edge information from the high SNR images (b-value = 0) to DWI reconstructions, achieving simultaneously noise suppression and image edges preservation.

RESULTS

Results on simulated and in vivo data show that PAIR can remove inter-shot motion artifacts very well (8 shots) and suppress the noise under the ultra-high b-value (4000 s/mm2) significantly.

CONCLUSION

The joint estimation model PAIR with complementary priors has a good performance on challenging reconstructions under inter-shot motions and a low signal-to-noise ratio.

SIGNIFICANCE

PAIR has potential in advanced clinical DWI applications and microstructure research.

Improvement of LED-based photoacoustic imaging using lag-coherence factor (LCF) beamforming

BACKGROUND

Owing to its portability, affordability, and energy-efficiency, LED-based photoacoustic (PA) imaging is increasingly becoming popular when compared to its laser-based alternative, mainly for superficial vascular imaging applications. However, this technique suffers from low SNR and thereby limited imaging depth. As a result, visual image quality of LED-based PA imaging is not optimal, especially in sub-surface vascular imaging applications.

PURPOSE

Combination of linear ultrasound (US) probes and LED arrays are the most common implementation in LED-based PA imaging, which is currently being explored for different clinical imaging applications. Traditional delay-and-sum (DAS) is the most common beamforming algorithm in linear array-based PA detection. Side-lobes and reconstruction-related artifacts make the DAS performance unsatisfactory and poor for a clinical-implementation. In this work, we explored a new weighting-based image processing technique for LED-based PAs to yield improved image quality when compared to the traditional methods.

METHODS

We are proposing a lag-coherence factor (LCF), which is fundamentally based on the combination of the spatial auto-correlation of the detected PA signals. In LCF, the numerator contains lag-delay-multiply-and-sum (DMAS) beamformer instead of a conventional DAS beamformer. A spatial auto-correlation operation is performed between the detected US array signals before using DMAS beamformer. We evaluated the new method on both tissue-mimicking phantom (2D) and human volunteer imaging (3D) data acquired using a commercial LED-based PA imaging system.

RESULTS

Our novel correlation-based weighting technique showed LED-based PA image quality improvement when it is combined with conventional DAS beamformer. Both phantom and human volunteer imaging results gave a direct confirmation that by introducing LCF, image quality was improved and this method could reduce side-lobes and artifacts when compared to the DAS and coherence-factor (CF) approaches. Signal-to-noise ratio, generalized contrast-to-noise ratio, contrast ratio and spatial resolution were evaluated and compared with conventional beamformers to assess the reconstruction performance in a quantitative way. Results show that our approach offered image quality enhancement with an average signal-to-noise ratio and spatial resolution improvement of around 20% and 25% respectively, when compared with conventional CF based DAS algorithm.

CONCLUSIONS

Our results demonstrate that the proposed LCF based algorithm performs better than the conventional DAS and CF algorithms by improving signal-to-noise ratio and spatial resolution. Therefore, our new weighting technique could be a promising tool to improve the performance of LED-based PA imaging and thus accelerate its clinical translation.

Enhancing the brain MRI at ultra-high field systems using a meta-array structure

BACKGROUND

The main advantage of ultra-high field (UHF) magnetic resonance neuroimaging is theincreased signal-to-noise ratio (SNR) compared with lower field strength imaging. However, the wavelength effect associated with UHF MRI results in radiofrequency (RF) inhomogeneity, compromising whole brain coverage for many commercial coils. Approaches to resolving this issue of transmit field inhomogeneity include the design of parallel transmit systems (PTx), RF pulse design, and applying passive RF shimming such as high dielectric materials. However, these methods have some drawbacks such as unstable material parameters of dielectric pads, high-cost, and complexity of PTx systems. Metasurfaces are artificial structures with a unique platform that can control the propagation of the electromagnetic (EM) waves, and they are very promising for engineering EM device. Implementation of meta-arrays enhancing MRI has been explored previously in several studies.

PURPOSE

The aim of this study was to assess the effect of new meta-array technology on enhancing the brain MRI at 7T. A meta-array based on a hybrid structure consisting of an array of broadside-coupled split-ring resonators and high-permittivity materials was designed to work at the Larmor frequency of a 7 Tesla (7T) MRI scanner. When placed behind the head and neck, this construct improves the SNR in the region of the cerebellum,brainstem and the inferior aspect of the temporal lobes.

METHODS

Numerical electromagnetic simulations were performed to optimize the meta-array design parameters and determine the RF circuit configuration. The resultant transmit-efficiency and signal sensitivity improvements were experimentally analyzed in phantoms followed by healthy volunteers using a 7T whole-body MRI scanner equipped with a standard one-channel transmit, 32-channel receive head coil. Efficacy was evaluated through acquisition with and without the meta-array using two basic sequences: gradient-recalled-echo (GRE) and turbo-spin-echo (TSE).

RESULTS

Experimental phantom analysis confirmed two-fold improvement in the transmit efficiency and 1.4-fold improvement in the signal sensitivity in the target region. In vivo GRE and TSE images with the meta-array in place showed enhanced visualization in inferior regions of the brain, especially of the cerebellum, brainstem, and cervical spinal cord.

CONCLUSION

Addition of the meta-array to commonly used MRI coils can enhance SNR to extend the anatomical coverage of the coil and improve overall MRI coil performance. This enhancement in SNR can be leveraged to obtain a higher resolution image over the same time slot or faster acquisition can be achieved with same resolution. Using this technique could improve the performance of existing commercial coils at 7T for whole brain and other applications.

Deeply Accelerated Arterial Spin Labeling Perfusion MRI for Measuring Cerebral Blood Flow and Arterial Transit Time

Cerebral blood flow (CBF) indicates both vascular integrity and brain function. Regional CBF can be non-invasively measured with arterial spin labeling (ASL) perfusion MRI. By repeating the same ASL MRI sequence several times, each with a different post-labeling delay (PLD), another important neurovascular index, the arterial transit time (ATT) can be estimated by fitting the acquired ASL signal to a kinetic model. This process however faces two challenges: one is the multiplicatively prolonged scan time, making it impractically for clinical use due to the escalated risk of motions; the other is the reduced signal-to-noise-ratio (SNR) in the long PLD scans due to the T1 decay of the labeled spins. Increasing SNR needs more repetitions which will further increase the total scan time. Currently, there lacks a way to accurately estimate ATT from a parsimonious number of PLDs. In this paper, we proposed a deep learning-based algorithm to reduce the number of PLDs and to accurately estimate ATT and CBF. Two separate deep networks were trained: one is designed to estimate CBF and ATT from ASL data with a single PLD; the other is to estimate CBF and ATT from ASL data with two PLDs. The models were trained and tested using the large Human Connectome Project multiple-PLD ASL MRI. Performance of the DL-based approach was compared to the traditional full dataset-based data fitting approach. Our results showed that ATT and CBF can be reliably estimated using deep networks even with one PLD.

Breast intervention device for low-field MRI with a customized unilateral coil

With the incidence of breast cancer rising to the top among female malignant tumors, magnetic resonance images guided breast biopsy intervention and minimally invasive treatment have developed as a clinically practical research issue. High field studies have shown the diagnostic value of breast MRI, but the examination costs greatly exceed those of competing conventional mammography. In this case, low-field MRI cannot merely provide typical MRI contrast, but also significantly reduce the cost of diagnosis and treatment for breast cancer patients. This work describes a unilateral breast coil and prototype intervention device, which provides a customized solution for low-field MRI-guided breast intervention. Results demonstrate that the low-field MRI breast intervention device facilitates medical intervention procedures. And the designed positioning device can locate the target lesion within 2-3 mm accuracy. Phantom tests with the customized unilateral coil indicate that the open loops perform as well as the 4-channel commercial closed breast coil, presenting a relatively good SNR (signal-to-noise ratio) and uniformity characteristics. MR scanning images of the volunteer breast using the breast intervention coil also show high SNR, which lays a foundation for further implementation of image-guided breast interventional minimally invasive surgery with the low-field MRI system.

Practical utility of a head-mounted gaze-directed beamforming system

Assistive auditory devices that enhance signal-to-noise ratio must follow the user's changing attention; errors could lead to the desired source being suppressed as noise. A method for measuring the practical benefit of attention-following speech enhancement is described and used to show a benefit for gaze-directed beamforming over natural binaural hearing. First, participants watched a recorded video conference call between two people with six additional interfering voices in different directions. The directions of the target voices corresponded to the spatial layout of their video streams. A simulated beamformer was yoked to the participant's gaze direction using an eye tracker. For the control condition, all eight voices were spatially distributed in a simulation of unaided binaural hearing. Participants completed questionnaires on the content of the conversation, scoring twice as high in the questionnaires for the beamforming condition. Sentence-by-sentence intelligibility was then measured using new participants who viewed the same audiovisual stimulus for each isolated sentence. Participants recognized twice as many words in the beamforming condition. The results demonstrate the potential practical benefit of gaze-directed beamforming for hearing aids and illustrate how detailed intelligibility data can be retrieved from an experiment that involves behavioral engagement in an ongoing listening task.

Applicability Evaluation of Full-Reference Image Quality Assessment Methods for Computed Tomography Images

Image quality assessments (IQA) are an important task for providing appropriate medical care. Full-reference IQA (FR-IQA) methods, such as peak signal-to-noise ratio (PSNR) and structural similarity (SSIM), are often used to evaluate imaging conditions, reconstruction conditions, and image processing algorithms, including noise reduction and super-resolution technology. However, these IQA methods may be inapplicable for medical images because they were designed for natural images. Therefore, this study aimed to investigate the correlation between objective assessment by some FR-IQA methods and human subjective assessment for computed tomography (CT) images. For evaluation, 210 distorted images were created from six original images using two types of degradation: noise and blur. We employed nine widely used FR-IQA methods for natural images: PSNR, SSIM, feature similarity (FSIM), information fidelity criterion (IFC), visual information fidelity (VIF), noise quality measure (NQM), visual signal-to-noise ratio (VSNR), multi-scale SSIM (MSSSIM), and information content-weighted SSIM (IWSSIM). Six observers performed subjective assessments using the double stimulus continuous quality scale (DSCQS) method. The performance of IQA methods was quantified using Pearson's linear correlation coefficient (PLCC), Spearman rank order correlation coefficient (SROCC), and root-mean-square error (RMSE). Nine FR-IQA methods developed for natural images were all strongly correlated with the subjective assessment (PLCC and SROCC > 0.8), indicating that these methods can apply to CT images. Particularly, VIF had the best values for all three items, PLCC, SROCC, and RMSE. These results suggest that VIF provides the most accurate alternative measure to subjective assessments for CT images.

Noise reduction by multiple path neural network using Attention mechanisms with an emphasis on robustness against Errors: A pilot study on brain Diffusion-Weighted images

PURPOSE

In deep learning-based noise reduction, larger networks offer advanced and complex functionality by utilizing its greater degree of freedom, but come with increased unpredictability, raising the potential risk of unforeseen errors. Here, we introduce a novel denoising model for diffusion-weighted images that intentionally limits the network output freedom by incorporating multiple pathways with varying degrees of freedom, with the aim of minimizing the chance of unintended alterations to the input. The purpose of this pilot study is to assess the model's ability to perform effective denoising under the constraints.

METHODS

Images from 10 healthy volunteers were used. Key innovations in our model development include: (1) neural network architecture that separated the function for calculating the specific output values from the function for adjusting the calculation for each pixel and (2) training that optimised the network based on both image and secondary obtained diffusion tensor. The generated images were compared with the original ones by measuring the deviation from ground truth images (averaged across eight acquisitions).

RESULTS

The generated images demonstrated closer alignment with the ground truth images, both visually and statistically (Q < 0.05), compared to the original images. Furthermore, the advantage of the generated images over the original images was also found in the secondary obtained quantitative parameter maps with significance (Q < 0.05).

CONCLUSION

The usefulness of the proposed method was suggested because it was successful in improving both the quality of the generated images and accuracy of the major diffusion parameter maps under the given restrictions.

Improving portable low-field MRI image quality through image-to-image translation using paired low- and high-field images

Low-field portable magnetic resonance imaging (MRI) scanners are more accessible, cost-effective, sustainable with lower carbon emissions than superconducting high-field MRI scanners. However, the images produced have relatively poor image quality, lower signal-to-noise ratio, and limited spatial resolution. This study develops and investigates an image-to-image translation deep learning model, LoHiResGAN, to enhance the quality of low-field (64mT) MRI scans and generate synthetic high-field (3T) MRI scans. We employed a paired dataset comprising T1- and T2-weighted MRI sequences from the 64mT and 3T and compared the performance of the LoHiResGAN model with other state-of-the-art models, including GANs, CycleGAN, U-Net, and cGAN. Our proposed method demonstrates superior performance in terms of image quality metrics, such as normalized root-mean-squared error, structural similarity index measure, peak signal-to-noise ratio, and perception-based image quality evaluator. Additionally, we evaluated the accuracy of brain morphometry measurements for 33 brain regions across the original 3T, 64mT, and synthetic 3T images. The results indicate that the synthetic 3T images created using our proposed LoHiResGAN model significantly improve the image quality of low-field MRI data compared to other methods (GANs, CycleGAN, U-Net, cGAN) and provide more consistent brain morphometry measurements across various brain regions in reference to 3T. Synthetic images generated by our method demonstrated high quality both quantitatively and qualitatively. However, additional research, involving diverse datasets and clinical validation, is necessary to fully understand its applicability for clinical diagnostics, especially in settings where high-field MRI scanners are less accessible.

Next-generation MRI scanner designed for ultra-high-resolution human brain imaging at 7 Tesla

To increase granularity in human neuroimaging science, we designed and built a next-generation 7 Tesla magnetic resonance imaging scanner to reach ultra-high resolution by implementing several advances in hardware. To improve spatial encoding and increase the image signal-to-noise ratio, we developed a head-only asymmetric gradient coil (200 mT m-1, 900 T m-1s-1) with an additional third layer of windings. We integrated a 128-channel receiver system with 64- and 96-channel receiver coil arrays to boost signal in the cerebral cortex while reducing g-factor noise to enable higher accelerations. A 16-channel transmit system reduced power deposition and improved image uniformity. The scanner routinely performs functional imaging studies at 0.35-0.45 mm isotropic spatial resolution to reveal cortical layer functional activity, achieves high angular resolution in diffusion imaging and reduces acquisition time for both functional and structural imaging.

Impact of virtual monoenergetic levels on coronary plaque volume components using photon-counting computed tomography

OBJECTIVES

Virtual monoenergetic images (VMIs) from photon-counting CT (PCCT) may change quantitative coronary plaque volumes. We aimed to assess how plaque component volumes change with respect to VMIs.

METHODS

Coronary CT angiography (CTA) images were acquired using a dual-source PCCT and VMIs were reconstructed between 40 and 180 keV in 10-keV increments. Polychromatic images at 120 kVp (T3D) were used as reference. Quantitative plaque analysis was performed on T3D images and segmentation masks were copied to VMI reconstructions. Calcified plaque (CP; > 350 Hounsfield units, HU), non-calcified plaque (NCP; 30 to 350 HU), and low-attenuation NCP (LAP; - 100 to 30 HU) volumes were calculated using fixed thresholds.

RESULTS

We analyzed 51 plaques from 51 patients (67% male, mean age 65 ± 12 years). Average attenuation and contrast-to-noise ratio (CNR) decreased significantly with increasing keV levels, with similar values observed between T3D and 70 keV images (299 ± 209 vs. 303 ± 225 HU, p = 0.15 for mean HU; 15.5 ± 3.7 vs. 15.8 ± 3.5, p = 0.32 for CNR). Mean NCP volume was comparable between T3D and 100-180-keV reconstructions. There was a monotonic decrease in mean CP volume, with a significant difference between all VMIs and T3D (p < 0.05). LAP volume increased with increasing keV levels and all VMIs showed a significant difference compared to T3D, except for 50 keV (28.0 ± 30.8 mm3 and 28.6 ± 30.1 mm3, respectively, p = 0.63).

CONCLUSIONS

Estimated coronary plaque volumes significantly differ between VMIs. Normalization protocols are needed to have comparable results between future studies, especially for LAP volume which is currently defined using a fixed HU threshold.

CLINICAL RELEVANCE STATEMENT

Different virtual monoenergetic images from photon-counting CT alter attenuation values and therefore corresponding plaque component volumes. New clinical standards and protocols are required to determine the optimal thresholds to derive plaque volumes from photon-counting CT.

KEY POINTS

• Utilizing different VMI energy levels from photon-counting CT for the analysis of coronary artery plaques leads to substantial changes in attenuation values and corresponding plaque component volumes. • Low-energy images (40-70 keV) improved contrast-to-noise ratio, however also increased image noise. • Normalization protocols are needed to have comparable results between future studies, especially for low-attenuation plaque volume which is currently defined using a fixed HU threshold.

Toward cell nuclei precision between OCT and H&E images translation using signal-to-noise ratio cycle-consistency

Medical image-to-image translation is often difficult and of limited effectiveness due to the differences in image acquisition mechanisms and the diverse structure of biological tissues. This work presents an unpaired image translation model between in-vivo optical coherence tomography (OCT) and ex-vivo Hematoxylin and eosin (H&E) stained images without the need for image stacking, registration, post-processing, and annotation. The model can generate high-quality and highly accurate virtual medical images, and is robust and bidirectional. Our framework introduces random noise to (1) blur redundant features, (2) defend against self-adversarial attacks, (3) stabilize inverse conversion, and (4) mitigate the impact of OCT speckles. We also demonstrate that our model can be pre-trained and then fine-tuned using images from different OCT systems in just a few epochs. Qualitative and quantitative comparisons with traditional image-to-image translation models show the robustness of our proposed signal-to-noise ratio (SNR) cycle-consistency method.

Denoising and uncertainty estimation in parameter mapping with approximate Bayesian deep image priors

PURPOSE

To mitigate the problem of noisy parameter maps with high uncertainties by casting parameter mapping as a denoising task based on Deep Image Priors.

METHODS

We extend the concept of denoising with Deep Image Prior (DIP) into parameter mapping by treating the output of an image-generating network as a parametrization of tissue parameter maps. The method implicitly denoises the parameter mapping process by filtering low-level image features with an untrained convolutional neural network (CNN). Our implementation includes uncertainty estimation from Bernoulli approximate variational inference, implemented with MC dropout, which provides model uncertainty in each voxel of the denoised parameter maps. The method is modular, so the specifics of different applications (e.g., T1 mapping) separate into application-specific signal equation blocks. We evaluate the method on variable flip angle T1 mapping, multi-echo T2 mapping, and apparent diffusion coefficient mapping.

RESULTS

We found that deep image prior adapts successfully to several applications in parameter mapping. In all evaluations, the method produces noise-reduced parameter maps with decreased uncertainty compared to conventional methods. The downsides of the proposed method are the long computational time and the introduction of some bias from the denoising prior.

CONCLUSION

DIP successfully denoise the parameter mapping process and applies to several applications with limited hyperparameter tuning. Further, it is easy to implement since DIP methods do not use network training data. Although time-consuming, uncertainty information from MC dropout makes the method more robust and provides useful information when properly calibrated.

End-to-end memory-efficient reconstruction for cone beam CT

BACKGROUND

Cone beam computed tomography (CBCT) plays an important role in many medical fields nowadays. Unfortunately, the potential of this imaging modality is hampered by lower image quality compared to the conventional CT, and producing accurate reconstructions remains challenging. A lot of recent research has been directed towards reconstruction methods relying on deep learning, which have shown great promise for various imaging modalities. However, practical application of deep learning to CBCT reconstruction is complicated by several issues, such as exceedingly high memory costs of deep learning methods when working with fully 3D data. Additionally, deep learning methods proposed in the literature are often trained and evaluated only on data from a specific region of interest, thus raising concerns about possible lack of generalization to other regions.

PURPOSE

In this work, we aim to address these limitations and propose LIRE: a learned invertible primal-dual iterative scheme for CBCT reconstruction.

METHODS

LIRE is a learned invertible primal-dual iterative scheme for CBCT reconstruction, wherein we employ a U-Net architecture in each primal block and a residual convolutional neural network (CNN) architecture in each dual block. Memory requirements of the network are substantially reduced while preserving its expressive power through a combination of invertible residual primal-dual blocks and patch-wise computations inside each of the blocks during both forward and backward pass. These techniques enable us to train on data with isotropic 2 mm voxel spacing, clinically-relevant projection count and detector panel resolution on current hardware with 24 GB video random access memory (VRAM).

RESULTS

Two LIRE models for small and for large field-of-view (FoV) setting were trained and validated on a set of 260 + 22 thorax CT scans and tested using a set of 142 thorax CT scans plus an out-of-distribution dataset of 79 head and neck CT scans. For both settings, our method surpasses the classical methods and the deep learning baselines on both test sets. On the thorax CT set, our method achieves peak signal-to-noise ratio (PSNR) of 33.84 ± 2.28 for the small FoV setting and 35.14 ± 2.69 for the large FoV setting; U-Net baseline achieves PSNR of 33.08 ± 1.75 and 34.29 ± 2.71 respectively. On the head and neck CT set, our method achieves PSNR of 39.35 ± 1.75 for the small FoV setting and 41.21 ± 1.41 for the large FoV setting; U-Net baseline achieves PSNR of 33.08 ± 1.75 and 34.29 ± 2.71 respectively. Additionally, we demonstrate that LIRE can be finetuned to reconstruct high-resolution CBCT data with the same geometry but 1 mm voxel spacing and higher detector panel resolution, where it outperforms the U-Net baseline as well.

CONCLUSIONS

Learned invertible primal-dual schemes with additional memory optimizations can be trained to reconstruct CBCT volumes directly from the projection data with clinically-relevant geometry and resolution. Such methods can offer better reconstruction quality and generalization compared to classical deep learning baselines.

Adaptive Trend Filtering for ECG Denoising and Delineation

Standard recordings of electrocardiograhic signals are contaminated by a large variety of noises and interferences, which impair their analysis and the further related diagnosis. In this article, we propose a method, based on compressive sensing techniques, to remove the main noise artifacts and to locate the main features of the pulses in the electrocardiogram (ECG). The motivation is to use trend filtering with a varying proximal parameter, in order to sequentially capture the peaks of the ECG, which have different functional regularities. The practical implementation is based on an adaptive version of the alternating direction method of multiplier (ADMM) algorithm. We present results obtained on simulated signals and on real data illustrating the validity of this approach, showing that results in peak localization are very good in both cases and comparable to state of the art approaches.

DC-SiamNet: Deep contrastive Siamese network for self-supervised MRI reconstruction

Reconstruction methods based on deep learning have greatly shortened the data acquisition time of magnetic resonance imaging (MRI). However, these methods typically utilize massive fully sampled data for supervised training, restricting their application in certain clinical scenarios and posing challenges to the reconstruction effect when high-quality MR images are unavailable. Recently, self-supervised methods have been developed that only undersampled MRI images participate in the network training. Nevertheless, due to the lack of complete referable MR image data, self-supervised reconstruction is prone to produce incorrect structure contents, such as unnatural texture details and over-smoothed tissue sites. To solve this problem, we propose a self-supervised Deep Contrastive Siamese Network (DC-SiamNet) for fast MR imaging. First, DC-SiamNet performs the reconstruction with a Siamese unrolled structure and obtains visual representations in different iterative phases. Particularly, an attention-weighted average pooling module is employed at the bottleneck layer of the U-shape regularization unit, which can effectively aggregate valuable local information of the underlying feature map in the generated representation vector. Then, a novel hybrid loss function is designed to drive the self-supervised reconstruction and contrastive learning simultaneously by forcing the output consistency across different branches in the frequency domain, the image domain, and the latent space. The proposed method is extensively evaluated with different sampling patterns on the IXI brain dataset and the MRINet knee dataset. Experimental results show that DC-SiamNet can achieve 0.93 in structural similarity and 33.984 dB in peak signal-to-noise ratio on the IXI brain dataset under 8x acceleration. It has better reconstruction accuracy than other methods, and the performance is close to the corresponding model trained with full supervision, especially when the sampling rate is low. In addition, generalization experiments verify that our method has a strong cross-domain reconstruction ability for different contrast brain images.

Optimization of wrist tendon detection in virtual monochromatic images using dual energy-computed tomography

OBJECTIVES

To evaluate the depiction of wrist tendons in virtual monochromatic images (VMIs) during a dual-energy CT (DE-CT) with the VMI image of conventional equivalent to 120 kVp.

MATERIALS AND METHODS

Using Catphan600 and phantom analysis software for CT evaluation, measurements of VMI in a DE-CT were performed corresponding to the tube voltages of single-energy CT at 120 kVp. Using a Discovery CT750 HD CT scanner (GE Healthcare) with DE-CT technology, 73 patients were scanned. We calculated the CT number, image noise, visual score, and contrast noise ratio (CNR) at the extensor pollicis tendon, extensor digitorum tendon, and flexor tendon in 11 VMIs from the DE-CT and VMI image of conventional equivalent to 120 kVp. The results from the optimal VMIs were then compared with that of the VMI image of the conventional equivalent to 120 kVp.

RESULTS

The highest CT number and CNR for the tendon were for the 140 keV VMI in the DE-CT compared to the other energy levels. There were significantly higher CT numbers, CNR values, and visual scores for each tendon at 140 keV VMI with the DE-CT (p < 0.01) compared with a VMI image of conventional equivalent to 120 kVp.

CONCLUSION

Energy level of the VMIs during DE-CT for the best wrist tendon delineation was 140 keV. This value of 140 keV for the DE-CT was significantly higher than the CT number and CNR for the extensor pollicis, extensor digitorum, and flexor tendon compared with a VMI image of conventional equivalent to 120 kVp.