Deep neural network for water/fat separation: Supervised training, unsupervised training, and no training

Unsupervised deep learning model for correcting Nyquist ghosts of single-shot spatiotemporal encoding

PURPOSE

To design an unsupervised deep learning (DL) model for correcting Nyquist ghosts of single-shot spatiotemporal encoding (SPEN) and evaluate the model for real MRI applications.

METHODS

The proposed method consists of three main components: (1) an unsupervised network that combines Residual Encoder and Restricted Subspace Mapping (RERSM-net) and is trained to generate a phase-difference map based on the even and odd SPEN images; (2) a spin physical forward model to obtain the corrected image with the learned phase difference map; and (3) cycle-consistency loss that is explored for training the RERSM-net.

RESULTS

The proposed RERSM-net could effectively generate smooth phase difference maps and correct Nyquist ghosts of single-shot SPEN. Both simulation and real in vivo MRI experiments demonstrated that our method outperforms the state-of-the-art SPEN Nyquist ghost correction method. Furthermore, the ablation experiments of generating phase-difference maps show the advantages of the proposed unsupervised model.

CONCLUSION

The proposed method can effectively correct Nyquist ghosts for the single-shot SPEN sequence.

Real-time water/fat imaging at 0.55T with spiral out-in-out-in sampling

PURPOSE

To develop an efficient and flexible water/fat separated real-time MRI (RT-MRI) method using spiral out-in-out-in (OIOI) sampling and balanced SSFP (bSSFP) at 0.55T.

METHODS

A bSSFP sequence with golden-angle spiral OIOI readout was developed, capturing three echoes to allow water/fat separation. A low-latency reconstruction that combines all echoes was available for online visualization. An offline reconstruction provided water and fat RT-MRI in two steps: (1) image reconstruction with spatiotemporally constrained reconstruction (STCR) and (2) water/fat separation with hierarchical iterative decomposition of water and fat with echo asymmetry and least-squares estimation (HIDEAL). In healthy volunteers, spiral OIOI was acquired in the wrist during a radial-to-ulnar deviation maneuver, in the heart without breath-hold and cardiac gating, and in the lower abdomen during free-breathing for visualizing small bowel motility.

RESULTS

We demonstrate successful water/fat separated RT-MRI for all tested applications. In the wrist, resulting images provided clear depiction of ligament gaps and their interactions during the radial-to-ulnar deviation maneuver. In the heart, water/fat RT-MRI depicted epicardial fat, provided improved delineation of epicardial coronary arteries, and provided high blood-myocardial contrast for ventricular function assessment. In the abdomen, water-only RT-MRI captured small bowel mobility clearly with improved water-fat contrast.

CONCLUSIONS

We have demonstrated a novel and flexible bSSFP spiral OIOI sequence at 0.55T that can provide water/fat separated RT-MRI with a variety of application-specific temporal resolution and spatial resolution requirements.

Comparative review of algorithms and methods for chemical-shift-encoded quantitative fat-water imaging

PURPOSE

To propose a standardized comparison between state-of-the-art open-source fat-water separation algorithms for proton density fat fraction (PDFF) andR 2 * $$ {R}_2^{\ast } $$ quantification using an open-source multi-language toolbox.

METHODS

Eight recent open-source fat-water separation algorithms were compared in silico, in vitro, and in vivo. Multi-echo data were synthesized with varying fat-fractions, B0 off-resonance, SNR and TEs. Experimental evaluation was conducted using calibrated fat-water phantoms acquired at 3T and multi-site open-source phantoms data. Algorithms' performances were observed on challenging in vivo datasets at 3T. Finally, reconstruction algorithms were investigated with different fat spectra to evaluate the importance of the fat model.

RESULTS

In silico and in vitro results proved most algorithms to be not sensitive to fat-water swaps andB 0 $$ {\mathrm{B}}_0 $$ offsets with five or more echoes. However, two methods remained inaccurate even with seven echoes and SNR = 50, and two other algorithms' precision depended on the echo spacing scheme (p < 0.05). The remaining four algorithms provided reliable performances with limits of agreement under 2% for PDFF and 6 s-1 forR 2 * $$ {R}_2^{\ast } $$ . The choice of fat spectrum model influenced quantification of PDFF mildly (<2% bias) and ofR 2 * $$ {R}_2^{\ast } $$ more severely, with errors up to 20 s-1 .

CONCLUSION

In promoting standardized comparisons of MRI-based fat and iron quantification using chemical-shift encoded multi-echo methods, this benchmark work has revealed some discrepancies between recent approaches for PDFF andR 2 * $$ {R}_2^{\ast } $$ mapping. Explicit choices and parameterization of the fat-water algorithm appear necessary for reproducibility. This open-source toolbox further enables the user to optimize acquisition parameters by predicting algorithms' margins of errors.

mcLARO: Multi-contrast learned acquisition and reconstruction optimization for simultaneous quantitative multi-parametric mapping

PURPOSE

To develop a method for rapid sub-millimeter T1 , T2 ,T 2 * $$ {\mathrm{T}}_2^{\ast } $$ , and QSM mapping in a single scan using multi-contrast learned acquisition and reconstruction optimization (mcLARO).

METHODS

A pulse sequence was developed by interleaving inversion recovery and T2 magnetization preparations and single-echo and multi-echo gradient echo acquisitions, which sensitized k-space data to T1 , T2 ,T 2 * $$ {\mathrm{T}}_2^{\ast } $$ , and magnetic susceptibility. The proposed mcLARO optimized both the multi-contrast k-space under-sampling pattern and image reconstruction based on image feature fusion using a deep learning framework. The proposed mcLARO method withR = 8 $$ R=8 $$ under-sampling was validated in a retrospective ablation study and compared with other deep learning reconstruction methods, including MoDL and Wave-MoDL, using fully sampled data as reference. Various under-sampling ratios in mcLARO were investigated. mcLARO was also evaluated in a prospective study using separately acquired conventionally sampled quantitative maps as reference standard.

RESULTS

The retrospective ablation study showed improved image sharpness of mcLARO compared to the baseline network without the multi-contrast sampling pattern optimization or image feature fusion module. The under-sampling ratio experiment showed that higher under-sampling ratios resulted in blurrier images and lower precision of quantitative values. The prospective study showed that small or negligible bias and narrow 95% limits of agreement on regional T1 , T2 ,T 2 * $$ {\mathrm{T}}_2^{\ast } $$ , and QSM values by mcLARO (5:39 mins) compared to reference scans (40:03 mins in total). mcLARO outperformed MoDL and Wave-MoDL in terms of image sharpness, noise suppression, and artifact removal.

CONCLUSION

mcLARO enabled fast sub-millimeter T1 , T2 ,T 2 * $$ {\mathrm{T}}_2^{\ast } $$ , and QSM mapping in a single scan.

Reducing the ambiguity of field inhomogeneity and chemical shift effect for fat-water separation by field factor

PURPOSE

To reduce the ambiguity between chemical shift and field inhomogeneity with flexible TE combinations by introducing a variable (field factor).

THEORY AND METHODS

The ambiguity between chemical shift and field inhomogeneity can be eliminated directly from the multiple in-phase images acquired at different TEs; however, it is only applicable to few echo combinations. In this study, we accommodated such an implementation in flexible TE combinations by introducing a new variable (field factor). The effects of the chemical shift were removed from the field inhomogeneity in the candidate solutions, thus reducing the ambiguity problem. To validate this concept, multi-echo MRI data acquired from various anatomies with different imaging parameters were tested. The derived fat and water images were compared with those of the state-of-the-art fat-water separation algorithms.

RESULTS

Robust fat-water separation was achieved with the accurate solution of field inhomogeneity, and no apparent fat-water swap was observed. In addition to the good performance, the proposed method is applicable to various fat-water separation applications, including different sequence types and flexible TE choices.

CONCLUSION

We propose an algorithm to reduce the ambiguity of chemical shift and field inhomogeneity and achieved robust fat-water separation in various applications.

Liver PDFF estimation using a multi-decoder water-fat separation neural network with a reduced number of echoes

OBJECTIVE

To accurately estimate liver PDFF from chemical shift-encoded (CSE) MRI using a deep learning (DL)-based Multi-Decoder Water-Fat separation Network (MDWF-Net), that operates over complex-valued CSE-MR images with only 3 echoes.

METHODS

The proposed MDWF-Net and a U-Net model were independently trained using the first 3 echoes of MRI data from 134 subjects, acquired with conventional 6-echoes abdomen protocol at 1.5 T. Resulting models were then evaluated using unseen CSE-MR images obtained from 14 subjects that were acquired with a 3-echoes CSE-MR pulse sequence with a shorter duration compared to the standard protocol. Resulting PDFF maps were qualitatively assessed by two radiologists, and quantitatively assessed at two corresponding liver ROIs, using Bland Altman and regression analysis for mean values, and ANOVA testing for standard deviation (STD) (significance level: .05). A 6-echo graph cut was considered ground truth.

RESULTS

Assessment of radiologists demonstrated that, unlike U-Net, MDWF-Net had a similar quality to the ground truth, despite it considered half of the information. Regarding PDFF mean values at ROIs, MDWF-Net showed a better agreement with ground truth (regression slope = 0.94, R2 = 0.97) than U-Net (regression slope = 0.86, R2 = 0.93). Moreover, ANOVA post hoc analysis of STDs showed a statistical difference between graph cuts and U-Net (p < .05), unlike MDWF-Net (p = .53).

CONCLUSION

MDWF-Net showed a liver PDFF accuracy comparable to the reference graph cut method, using only 3 echoes and thus allowing a reduction in the acquisition times.

CLINICAL RELEVANCE STATEMENT

We have prospectively validated that the use of a multi-decoder convolutional neural network to estimate liver proton density fat fraction allows a significant reduction in MR scan time by reducing the number of echoes required by 50%.

KEY POINTS

• Novel water-fat separation neural network allows for liver PDFF estimation by using multi-echo MR images with a reduced number of echoes. • Prospective single-center validation demonstrated that echo reduction leads to a significant shortening of the scan time, compared to standard 6-echo acquisition. • Qualitative and quantitative performance of the proposed method showed no significant differences in PDFF estimation with respect to the reference technique.

What Is Machine Learning, Artificial Neural Networks and Deep Learning?-Examples of Practical Applications in Medicine

Machine learning (ML), artificial neural networks (ANNs), and deep learning (DL) are all topics that fall under the heading of artificial intelligence (AI) and have gained popularity in recent years. ML involves the application of algorithms to automate decision-making processes using models that have not been manually programmed but have been trained on data. ANNs that are a part of ML aim to simulate the structure and function of the human brain. DL, on the other hand, uses multiple layers of interconnected neurons. This enables the processing and analysis of large and complex databases. In medicine, these techniques are being introduced to improve the speed and efficiency of disease diagnosis and treatment. Each of the AI techniques presented in the paper is supported with an example of a possible medical application. Given the rapid development of technology, the use of AI in medicine shows promising results in the context of patient care. It is particularly important to keep a close eye on this issue and conduct further research in order to fully explore the potential of ML, ANNs, and DL, and bring further applications into clinical use in the future.

Data adaptive regularization with reference tissue constraints for liver quantitative susceptibility mapping

PURPOSE

To improve repeatability and reproducibility across acquisition parameters and reduce bias in quantitative susceptibility mapping (QSM) of the liver, through development of an optimized regularized reconstruction algorithm for abdominal QSM.

METHODS

An optimized approach to estimation of magnetic susceptibility distribution is formulated as a constrained reconstruction problem that incorporates estimates of the input data reliability and anatomical priors available from chemical shift-encoded imaging. The proposed data-adaptive method was evaluated with respect to bias, repeatability, and reproducibility in a patient population with a wide range of liver iron concentration (LIC). The proposed method was compared to the previously proposed and validated approach in liver QSM for two multi-echo spoiled gradient-recalled echo protocols with different acquisition parameters at 3T. Linear regression was used for evaluation of QSM methods against a reference FDA-approvedR 2 $$ {R}_2 $$ -based LIC measure andR 2 ∗ $$ {R}_2^{\ast } $$ measurements; repeatability/reproducibility were assessed by Bland-Altman analysis.

RESULTS

The data-adaptive method produced susceptibility maps with higher subjective quality due to reduced shading artifacts. For both acquisition protocols, higher linear correlation with bothR 2 $$ {R}_2 $$ - andR 2 ∗ $$ {R}_2^{\ast } $$ -based measurements were observed for the data-adaptive method (r 2 = 0 . 74 / 0 . 69 $$ {r}^2=0.74/0.69 $$ forR 2 $$ {R}_2 $$ ,0 . 97 / 0 . 95 $$ 0.97/0.95 $$ forR 2 ∗ $$ {R}_2^{\ast } $$ ) than the standard method (r 2 = 0 . 60 / 0 . 66 $$ {r}^2=0.60/0.66 $$ and0 . 79 / 0 . 88 $$ 0.79/0.88 $$ ). For both protocols, the data-adaptive method enabled better test-retest repeatability (repeatability coefficients 0.19/0.30 ppm for the data-adaptive method, 0.38/0.47 ppm for the standard method) and reproducibility across protocols (reproducibility coefficient 0.28 vs. 0.53ppm) than the standard method.

CONCLUSIONS

The proposed data-adaptive QSM algorithm may enable quantification of LIC with improved repeatability/reproducibility across different acquisition parameters as 3T.

A low-rank deep image prior reconstruction for free-breathing ungated spiral functional CMR at 0.55 T and 1.5 T

OBJECTIVE

This study combines a deep image prior with low-rank subspace modeling to enable real-time (free-breathing and ungated) functional cardiac imaging on a commercial 0.55 T scanner.

MATERIALS AND METHODS

The proposed low-rank deep image prior (LR-DIP) uses two u-nets to generate spatial and temporal basis functions that are combined to yield dynamic images, with no need for additional training data. Simulations and scans in 13 healthy subjects were performed at 0.55 T and 1.5 T using a golden angle spiral bSSFP sequence with images reconstructed using [Formula: see text]-ESPIRiT, low-rank plus sparse (L + S) matrix completion, and LR-DIP. Cartesian breath-held ECG-gated cine images were acquired for reference at 1.5 T. Two cardiothoracic radiologists rated images on a 1-5 scale for various categories, and LV function measurements were compared.

RESULTS

LR-DIP yielded the lowest errors in simulations, especially at high acceleration factors (R [Formula: see text] 8). LR-DIP ejection fraction measurements agreed with 1.5 T reference values (mean bias - 0.3% at 0.55 T and - 0.2% at 1.5 T). Compared to reference images, LR-DIP images received similar ratings at 1.5 T (all categories above 3.9) and slightly lower at 0.55 T (above 3.4).

CONCLUSION

Feasibility of real-time functional cardiac imaging using a low-rank deep image prior reconstruction was demonstrated in healthy subjects on a commercial 0.55 T scanner.

QSM Throughout the Body

Magnetic materials in tissue, such as iron, calcium, or collagen, can be studied using quantitative susceptibility mapping (QSM). To date, QSM has been overwhelmingly applied in the brain, but is increasingly utilized outside the brain. QSM relies on the effect of tissue magnetic susceptibility sources on the MR signal phase obtained with gradient echo sequence. However, in the body, the chemical shift of fat present within the region of interest contributes to the MR signal phase as well. Therefore, correcting for the chemical shift effect by means of water-fat separation is essential for body QSM. By employing techniques to compensate for cardiac and respiratory motion artifacts, body QSM has been applied to study liver iron and fibrosis, heart chamber blood and placenta oxygenation, myocardial hemorrhage, atherosclerotic plaque, cartilage, bone, prostate, breast calcification, and kidney stone.

Uncertainty-aware physics-driven deep learning network for free-breathing liver fat and R2 * quantification using self-gated stack-of-radial MRI

PURPOSE

To develop a deep learning-based method for rapid liver proton-density fat fraction (PDFF) and R2 * quantification with built-in uncertainty estimation using self-gated free-breathing stack-of-radial MRI.

METHODS

This work developed an uncertainty-aware physics-driven deep learning network (UP-Net) to (1) suppress radial streaking artifacts because of undersampling after self-gating, (2) calculate accurate quantitative maps, and (3) provide pixel-wise uncertainty maps. UP-Net incorporated a phase augmentation strategy, generative adversarial network architecture, and an MRI physics loss term based on a fat-water and R2 * signal model. UP-Net was trained and tested using free-breathing multi-echo stack-of-radial MRI data from 105 subjects. UP-Net uncertainty scores were calibrated in a validation dataset and used to predict quantification errors for liver PDFF and R2 * in a testing dataset.

RESULTS

Compared with images reconstructed using compressed sensing (CS), UP-Net achieved structural similarity index >0.87 and normalized root mean squared error <0.18. Compared with reference quantitative maps generated using CS and graph-cut (GC) algorithms, UP-Net achieved low mean differences (MD) for liver PDFF (-0.36%) and R2 * (-0.37 s-1 ). Compared with breath-holding Cartesian MRI results, UP-Net achieved low MD for liver PDFF (0.53%) and R2 * (6.75 s-1 ). UP-Net uncertainty scores predicted absolute liver PDFF and R2 * errors with low MD of 0.27% and 0.12 s-1 compared to CS + GC results. The computational time for UP-Net was 79 ms/slice, whereas CS + GC required 3.2 min/slice.

CONCLUSION

UP-Net rapidly calculates accurate liver PDFF and R2 * maps from self-gated free-breathing stack-of-radial MRI. The pixel-wise uncertainty maps from UP-Net predict quantification errors in the liver.

GRASPNET: Fast spatiotemporal deep learning reconstruction of golden-angle radial data for free-breathing dynamic contrast-enhanced magnetic resonance imaging

The purpose of the current study was to develop a deep learning technique called Golden-angle RAdial Sparse Parallel Network (GRASPnet) for fast reconstruction of dynamic contrast-enhanced 4D MRI acquired with golden-angle radial k-space trajectories. GRASPnet operates in the image-time space and does not use explicit data consistency to minimize the reconstruction time. Three different network architectures were developed: (1) GRASPnet-2D: 2D convolutional kernels (x,y) and coil and contrast dimensions collapsed into a single combined dimension; (2) GRASPnet-3D: 3D kernels (x,y,t); and (3) GRASPnet-2D + time: two 3D kernels to first exploit spatial correlations (x,y,1) followed by temporal correlations (1,1,t). The networks were trained using iterative GRASP reconstruction as the reference. Free-breathing 3D abdominal imaging with contrast injection was performed on 33 patients with liver lesions using a T1-weighted golden-angle stack-of-stars pulse sequence. Ten datasets were used for testing. The three GRASPnet architectures were compared with iterative GRASP results using quantitative and qualitative analysis, including impressions from two body radiologists. The three GRASPnet techniques reduced the reconstruction time to about 13 s with similar results with respect to iterative GRASP. Among the GRASPnet techniques, GRASPnet-2D + time compared favorably in the quantitative analysis. Spatiotemporal deep learning enables reconstruction of dynamic 4D contrast-enhanced images in a few seconds, which would facilitate translation to clinical practice of compressed sensing methods that are currently limited by long reconstruction times.

LARO: Learned acquisition and reconstruction optimization to accelerate quantitative susceptibility mapping

Quantitative susceptibility mapping (QSM) involves acquisition and reconstruction of a series of images at multi-echo time points to estimate tissue field, which prolongs scan time and requires specific reconstruction technique. In this paper, we present our new framework, called Learned Acquisition and Reconstruction Optimization (LARO), which aims to accelerate the multi-echo gradient echo (mGRE) pulse sequence for QSM. Our approach involves optimizing a Cartesian multi-echo k-space sampling pattern with a deep reconstruction network. Next, this optimized sampling pattern was implemented in an mGRE sequence using Cartesian fan-beam k-space segmenting and ordering for prospective scans. Furthermore, we propose to insert a recurrent temporal feature fusion module into the reconstruction network to capture signal redundancies along echo time. Our ablation studies show that both the optimized sampling pattern and proposed reconstruction strategy help improve the quality of the multi-echo image reconstructions. Generalization experiments show that LARO is robust on the test data with new pathologies and different sequence parameters. Our code is available at https://github.com/Jinwei1209/LARO-QSM.git.

Artifact-free fat-water separation in Dixon MRI using deep learning

Chemical-shift encoded MRI (CSE-MRI) is a widely used technique for the study of body composition and metabolic disorders, where derived fat and water signals enable the quantification of adipose tissue and muscle. The UK Biobank is acquiring whole-body Dixon MRI (a specific implementation of CSE-MRI) for over 100,000 participants. Current processing methods associated with large whole-body volumes are time intensive and prone to artifacts during fat-water separation performed by the scanner, making quantitative analysis challenging. The most common artifacts are fat-water swaps, where the labels are inverted at the voxel level. It is common for researchers to discard swapped data (generally around 10%), which is wasteful and may lead to unintended biases. Given the large number of whole-body Dixon MRI acquisitions in the UK Biobank, thousands of swaps are expected to be present in the fat and water volumes from image reconstruction performed on the scanner. If they go undetected, errors will propagate into processes such as organ segmentation, and dilute the results in population-based analyses. There is a clear need for a robust method to accurately separate fat and water volumes in big data collections like the UK Biobank. We formulate fat-water separation as a style transfer problem, where swap-free fat and water volumes are predicted from the acquired Dixon MRI data using a conditional generative adversarial network, and introduce a new loss function for the generator model. Our method is able to predict highly accurate fat and water volumes free from artifacts in the UK Biobank. We show that our model separates fat and water volumes using either single input (in-phase only) or dual input (in-phase and opposed-phase) data, with the latter producing superior results. Our proposed method enables faster and more accurate downstream analysis of body composition from Dixon MRI in population studies by eliminating the need for visual inspection or discarding data due to fat-water swaps.

Supplementary Information

The online version contains supplementary material available at 10.1186/s40537-022-00677-1.

Deep learning-based quantitative susceptibility mapping (QSM) in the presence of fat using synthetically generated multi-echo phase training data

PURPOSE

To enable a fast and automatic deep learning-based QSM reconstruction of tissues with diverse chemical shifts, relevant to most regions outside the brain.

METHODS

A UNET was trained to reconstruct susceptibility maps using synthetically generated, unwrapped, multi-echo phase data as input. The RMS error with respect to synthetic validation data was computed. The method was tested on two in vivo knee and two pelvis data sets. Comparisons were made to a conventional fat-water separation pipeline by applying a commonly used graph-cut algorithm, both without and with an extended mask for background field removal (FWS-CONV-QSM and FWS-MASK-CONV-QSM, respectively). Several regions of interest were segmented and compared. Furthermore, the approach was tested on a prostate cancer patient receiving low-dose-rate brachytherapy, to detect and localize the seeds by MRI.

RESULTS

The RMS error was 0.292 ppm with FWS-CONV-QSM and 0.123 ppm for the UNET approach. Susceptibility maps were reconstructed much faster (< 10 s) and completely automatically (no background masking needed) by the UNET compared with the other applied techniques (5 min 51 s and 22 min 44 s for CONV-QSM and FWS-MASK-CONV-QSM, respectively. Background artifacts, fat-water swaps, and hypointense artifacts between I-125 seeds of a patient receiving low-dose brachytherapy in the prostate were largely reduced in the UNET approach.

CONCLUSIONS

Deep learning-based QSM reconstruction, trained solely with synthetic data, is well-suited to rapidly reconstructing high-quality susceptibility maps in the presence of fat without needing masking for background field removal.

Subsecond accurate myelin water fraction reconstruction from FAST-T2 data with 3D UNET

PURPOSE

To develop a 3D UNET convolutional neural network for rapid extraction of myelin water fraction (MWF) maps from six-echo fast acquisition with spiral trajectory and T₂ -prep data and to evaluate its accuracy in comparison with multilayer perceptron (MLP) network.

METHODS

The MWF maps were extracted from 138 patients with multiple sclerosis using an iterative three-pool nonlinear least-squares algorithm (NLLS) without and with spatial regularization (srNLLS), which were used as ground-truth labels to train, validate, and test UNET and MLP networks as a means to accelerate data fitting. Network testing was performed in 63 patients with multiple sclerosis and a numerically simulated brain phantom at SNR of 200, 100 and 50.

RESULTS

Simulations showed that UNET reduced the MWF mean absolute error by 30.1% to 56.4% and 16.8% to 53.6% over the whole brain and by 41.2% to 54.4% and 21.4% to 49.4% over the lesions for predicting srNLLS and NLLS MWF, respectively, compared to MLP, with better performance at lower SNRs. UNET also outperformed MLP for predicting srNLLS MWF in the in vivo multiple-sclerosis brain data, reducing mean absolute error over the whole brain by 61.9% and over the lesions by 67.5%. However, MLP yielded 41.1% and 51.7% lower mean absolute error for predicting in vivo NLLS MWF over the whole brain and the lesions, respectively, compared with UNET. The whole-brain MWF processing time using a GPU was 0.64 seconds for UNET and 0.74 seconds for MLP.

CONCLUSION

Subsecond whole-brain MWF extraction from fast acquisition with spiral trajectory and T₂ -prep data using UNET is feasible and provides better accuracy than MLP for predicting MWF output of srNLLS algorithm.

Computed Tomography Image Features under Deep Learning Algorithm Applied in Staging Diagnosis of Bladder Cancer and Detection on Ceramide Glycosylation

The research is aimed at investigating computed tomography (CT) image based on deep learning algorithm and the application value of ceramide glycosylation in diagnosing bladder cancer. The images of ordinary CT detection were improved. In this study, 60 bladder cancer patients were selected and performed with ordinary CT detection, and the detection results were processed by CT based on deep learning algorithms and compared with pathological diagnosis. In addition, Western Blot technology was used to detect the expression of glucose ceramide synthase (GCS) in the cell membrane of tumor tissues and normal tissues of bladder. The comparison results found that, in simple CT clinical staging, the coincidence rates of T1 stage, T2a stage, T2b stage, T3 stage, and T4 stage were 28.56%, 62.51%, 78.94%, 84.61%, and 74.99%, respectively; and the total coincidence rate of CT clinical staging was 63.32%, which was greatly different from the clinical staging of pathological diagnosis (P < 0.05). In the clinical staging of algorithm-based CT test results, the coincidence rates of T1 stage and T2a stage were 50.01% and 91.65%, respectively; and those of T2b stage, T3 stage, and T4 stage were 100.00%; and the total coincidence rate was 96.69%, which was not obviously different from the clinical staging of pathological diagnosis (P > 0.05). Therefore, it could be concluded that the algorithm-based CT detection results were more accurate, and the use of CT scans based on deep learning algorithms in the preoperative staging and clinical treatment of bladder cancer showed reliable guiding significance and clinical value. In addition, it was found that the expression level of GCS in normal bladder tissues was much lower than that in bladder cancer tissues. This indicated that the changes in GCS were closely related to the development and prognosis of bladder cancer. Therefore, it was believed that GCS may be an effective target for the treatment of bladder cancer in the future, and further research was needed for specific conditions.

Generation of quantification maps and weighted images from synthetic magnetic resonance imaging using deep learning network

OBJECTIVE

The generation of quantification maps and weighted images in synthetic MRI techniques is based on complex fitting equations. This process requires longer image generation times. The objective of this study is to evaluate the feasibility of deep learning method for fast reconstruction of synthetic MRI.

APPROACH

A total of 44 healthy subjects were recruited and random divided into a training set (30 subjects) and a testing set (14 subjects). A multiple-dynamic, multiple-echo (MDME) sequence was used to acquire synthetic MRI images. Quantification maps (T1, T2, and proton density (PD) maps) and weighted (T1W, T2W, and T2W FLAIR) images were created with MAGiC software and then used as the ground truth images in the deep learning (DL) model. An improved multichannel U-Net structure network was trained to generate quantification maps and weighted images from raw synthetic MRI imaging data (8 module images). Quantitative evaluation was performed on quantification maps. Quantitative evaluation metrics, as well as qualitative evaluation were used in weighted image evaluation. Nonparametric Wilcoxon signed-rank tests were performed in this study.

MAIN RESULTS

The results of quantitative evaluation show that the error between the generated quantification images and the reference images is small. For weighted images, no significant difference in overall image quality or SNR was identified between DL images and synthetic images. Notably, the DL images achieved improved image contrast with T2W images, and fewer artifacts were present on DL images than synthetic images acquired by T2W FLAIR.

SIGNIFICANCE

The DL algorithm provides a promising method for image generation in synthetic MRI techniques, in which every step of the calculation can be optimized and faster, thereby simplifying the workflow of synthetic MRI techniques.

QQ-NET - using deep learning to solve quantitative susceptibility mapping and quantitative blood oxygen level dependent magnitude (QSM+qBOLD or QQ) based oxygen extraction fraction (OEF) mapping

PURPOSE

To improve accuracy and speed of quantitative susceptibility mapping plus quantitative blood oxygen level-dependent magnitude (QSM+qBOLD or QQ) -based oxygen extraction fraction (OEF) mapping using a deep neural network (QQ-NET).

METHODS

The 3D multi-echo gradient echo images were acquired in 34 ischemic stroke patients and 4 healthy subjects. Arterial spin labeling and diffusion weighted imaging (DWI) were also performed in the patients. NET was developed to solve the QQ model inversion problem based on Unet. QQ-based OEF maps were reconstructed with previously introduced temporal clustering, tissue composition, and total variation (CCTV) and NET. The results were compared in simulation, ischemic stroke patients, and healthy subjects using a two-sample Kolmogorov-Smirnov test.

RESULTS

In the simulation, QQ-NET provided more accurate and precise OEF maps than QQ-CCTV with 150 times faster reconstruction speed. In the subacute stroke patients, OEF from QQ-NET had greater contrast-to-noise ratio (CNR) between DWI-defined lesions and their unaffected contralateral normal tissue than with QQ-CCTV: 1.9 ± 1.3 vs 6.6 ± 10.7 (p = 0.03). In healthy subjects, both QQ-CCTV and QQ-NET provided uniform OEF maps.

CONCLUSION

QQ-NET improves the accuracy of QQ-based OEF with faster reconstruction.

Multi-muscle deep learning segmentation to automate the quantification of muscle fat infiltration in cervical spine conditions

Muscle fat infiltration (MFI) has been widely reported across cervical spine disorders. The quantification of MFI requires time-consuming and rater-dependent manual segmentation techniques. A convolutional neural network (CNN) model was trained to segment seven cervical spine muscle groups (left and right muscles segmented separately, 14 muscles total) from Dixon MRI scans (n = 17, 17 scans < 2 weeks post motor vehicle collision (MVC), and 17 scans 12 months post MVC). The CNN MFI measures demonstrated high test reliability and accuracy in an independent testing dataset (n = 18, 9 scans < 2 weeks post MVC, and 9 scans 12 months post MVC). Using the CNN in 84 participants with scans < 2 weeks post MVC (61 females, 23 males, age = 34.2 ± 10.7 years) differences in MFI between the muscle groups and relationships between MFI and sex, age, and body mass index (BMI) were explored. Averaging across all muscles, females had significantly higher MFI than males (p = 0.026). The deep cervical muscles demonstrated significantly greater MFI than the more superficial muscles (p < 0.001), and only MFI within the deep cervical muscles was moderately correlated to age (r > 0.300, p ≤ 0.001). CNN's allow for the accurate and rapid, quantitative assessment of the composition of the architecturally complex muscles traversing the cervical spine. Acknowledging the wider reports of MFI in cervical spine disorders and the time required to manually segment the individual muscles, this CNN may have diagnostic, prognostic, and predictive value in disorders of the cervical spine.