MR fingerprinting Deep RecOnstruction NEtwork (DRONE)

Deep Learning Staging of Liver Iron Content From Multiecho MR Images

BACKGROUND

MRI represents the most established liver iron content (LIC) evaluation approach by estimation of liver T2* value, but it is dependent on the choice of the measurement region and the software used for image analysis.

PURPOSE

To develop a deep-learning method for unsupervised classification of LIC from magnitude T2* multiecho MR images.

STUDY TYPE

Retrospective.

POPULATION/SUBJECTS

A total of 1069 thalassemia major patients enrolled in the core laboratory of the Myocardial Iron Overload in Thalassemia (MIOT) network, which were included in the training (80%) and test (20%) sets. Twenty patients from different MRI vendors included in the external test set.

FIELD STRENGTH/SEQUENCE

A5 T, T2* multiecho magnitude images.

ASSESSMENT

Four deep-learning convolutional neural networks (HippoNet-2D, HippoNet-3D, HippoNet-LSTM, and an ensemble network HippoNet-Ensemble) were used to achieve unsupervised staging of LIC using five classes (normal, borderline, middle, moderate, severe). The training set was employed to construct the deep-learning model. The performance of the LIC staging model was evaluated in the test set and in the external test set. The model's performances were assessed by evaluating the accuracy, sensitivity, and specificity with respect to the ground truth labels obtained by T2* measurements and by comparison with operator-induced variability originating from different region of interest (ROI) placements.

STATISTICAL TESTS

The network's performances were evaluated by single-class accuracy, specificity, and sensitivity and compared by one-way repeated measures analysis of variance (ANOVA) and one-way ANOVA.

RESULTS

HippoNet-Ensemble reached an accuracy significantly higher than the other networks, and a sensitivity and specificity higher than HippoNet-LSTM. Accuracy, sensitivity, and specificity values for the LIC stages were: normal: 0.96/0.93/0.97, borderline: 0.95/0.85/0.98, mild: 0.96/0.88/0.98, moderate: 0.95/0.89/0.97, severe: 0.97/0.95/0.98. Correctly staging of cases was in the range of 85%-95%, depending on the LIC class. Multiclass accuracy was 0.90 against 0.92 for the interobserver variability.

DATA CONCLUSION

The proposed HippoNet-Ensemble network can perform unsupervised LIC staging and achieves good prognostic performance.

EVIDENCE LEVEL

4 TECHNICAL EFFICACY: Stage 2.

Quantitative MRI in cardiometabolic disease: From conventional cardiac and liver tissue mapping techniques to multi-parametric approaches

Cardiometabolic disease refers to the spectrum of chronic conditions that include diabetes, hypertension, atheromatosis, non-alcoholic fatty liver disease, and their long-term impact on cardiovascular health. Histological studies have confirmed several modifications at the tissue level in cardiometabolic disease. Recently, quantitative MR methods have enabled non-invasive myocardial and liver tissue characterization. MR relaxation mapping techniques such as T₁, T_1ρ, T₂ and T₂* provide a pixel-by-pixel representation of the corresponding tissue specific relaxation times, which have been shown to correlate with fibrosis, altered tissue perfusion, oedema and iron levels. Proton density fat fraction mapping approaches allow measurement of lipid tissue in the organ of interest. Several studies have demonstrated their utility as early diagnostic biomarkers and their potential to bear prognostic implications. Conventionally, the quantification of these parameters by MRI relies on the acquisition of sequential scans, encoding and mapping only one parameter per scan. However, this methodology is time inefficient and suffers from the confounding effects of the relaxation parameters in each single map, limiting wider clinical and research applications. To address these limitations, several novel approaches have been proposed that encode multiple tissue parameters simultaneously, providing co-registered multiparametric information of the tissues of interest. This review aims to describe the multi-faceted myocardial and hepatic tissue alterations in cardiometabolic disease and to motivate the application of relaxometry and proton-density cardiac and liver tissue mapping techniques. Current approaches in myocardial and liver tissue characterization as well as latest technical developments in multiparametric quantitative MRI are included. Limitations and challenges of these novel approaches, and recommendations to facilitate clinical validation are also discussed.

SuperMAP: Deep ultrafast MR relaxometry with joint spatiotemporal undersampling

PURPOSE

To develop an ultrafast and robust MR parameter mapping network using deep learning.

THEORY AND METHODS

We design a deep learning framework called SuperMAP that directly converts a series of undersampled (both in k-space and parameter-space) parameter-weighted images into several quantitative maps, bypassing the conventional exponential fitting procedure. We also present a novel technique to simultaneously reconstruct T1rho and T2 relaxation maps within a single scan. Full data were acquired and retrospectively undersampled for training and testing using traditional and state-of-the-art techniques for comparison. Prospective data were also collected to evaluate the trained network. The performance of all methods is evaluated using the parameter qualification errors and other metrics in the segmented regions of interest.

RESULTS

SuperMAP achieved accurate T1rho and T2 mapping with high acceleration factors (R = 24 and R = 32). It exploited both spatial and temporal information and yielded low error (normalized mean square error of 2.7% at R = 24 and 2.8% at R = 32) and high resemblance (structural similarity of 97% at R = 24 and 96% at R = 32) to the gold standard. The network trained with retrospectively undersampled data also works well for the prospective data (with a slightly lower acceleration factor). SuperMAP is also superior to conventional methods.

CONCLUSION

Our results demonstrate the feasibility of generating superfast MR parameter maps through very few undersampled parameter-weighted images. SuperMAP can simultaneously generate T1rho and T2 relaxation maps in a short scan time.

CEST MR fingerprinting (CEST-MRF) for brain tumor quantification using EPI readout and deep learning reconstruction

PURPOSE

To develop a clinical CEST MR fingerprinting (CEST-MRF) method for brain tumor quantification using EPI acquisition and deep learning reconstruction.

METHODS

A CEST-MRF pulse sequence originally designed for animal imaging was modified to conform to hardware limits on clinical scanners while keeping scan time under 2 min. Quantitative MRF reconstruction was performed using a deep reconstruction network (DRONE) to yield the water relaxation and chemical exchange parameters. The feasibility of the six parameter DRONE reconstruction was tested in simulations using a digital brain phantom. A healthy subject was scanned with the CEST-MRF sequence, conventional MRF and CEST sequences for comparison. Reproducibility was assessed via test-retest experiments and the concordance correlation coefficient calculated for white matter and gray matter. The clinical utility of CEST-MRF was demonstrated on four patients with brain metastases in comparison to standard clinical imaging sequences. Tumors were segmented into edema, solid core, and necrotic core regions and the CEST-MRF values compared to the contra-lateral side.

RESULTS

DRONE reconstruction of the digital phantom yielded a normalized RMS error of ≤7% for all parameters. The CEST-MRF parameters were in good agreement with those from conventional MRF and CEST sequences and previous studies. The mean concordance correlation coefficient for all six parameters was 0.98 ± 0.01 in white matter and 0.98 ± 0.02 in gray matter. The CEST-MRF values in nearly all tumor regions were significantly different (P = 0.05) from each other and the contra-lateral side.

CONCLUSION

Combination of EPI readout and deep learning reconstruction enabled fast, accurate and reproducible CEST-MRF in brain tumors.

Comprehensive Clinical Evaluation of a Deep Learning-Accelerated, Single-Breath-Hold Abdominal HASTE at 1.5 T and 3 T

To evaluate the clinical performance of a deep learning-accelerated single-breath-hold half-Fourier acquisition single-shot turbo spin echo (HASTE_DL)-sequence for T2-weighted fat-suppressed MRI of the abdomen at 1.5 T and 3 T in comparison to standard T2-weighted fat-suppressed multi-shot turbo spin echo-sequence. A total of 320 patients who underwent a clinically indicated liver MRI at 1.5 T and 3 T between August 2020 and February 2021 were enrolled in this single-center, retrospective study. HASTE_DL and standard sequences were assessed regarding overall and organ-based image quality, noise, contrast, sharpness, artifacts, diagnostic confidence, as well as lesion detectability using a Likert scale ranging from 1 to 4 (4 = best). The number of visible lesions of each organ was counted and the largest diameter of the major lesion was measured. HASTE_DL showed excellent image quality (median 4, interquartile range 3-4), although BLADE (median 4, interquartile range 4-4) was rated significantly higher for overall and organ-based image quality of the adrenal gland (P < .001), contrast (P < 0.001), sharpness (P < 0.001), artifacts (P < 0.001), as well as diagnostic confidence (P < .001). No significant differences were found concerning noise (P = 0.886), organ-based image quality of the liver, pancreas, spleen, and kidneys (P = 0.120-0.366), number and measured diameter of the detected lesions (ICC = 0.972-1.0). Reduction of the aquisition time (TA) was at least 89% for 1.5 T images and 86% for 3 T images. HASTE_DL provided excellent image quality, good diagnostic confidence and lesion detection compared to a standard T2-sequences, allowing an eminent reduction of the acquisition time.

Region of interest-specific loss functions improve T2 quantification with ultrafast T2 mapping MRI sequences in knee, hip and lumbar spine

MRI T2 mapping sequences quantitatively assess tissue health and depict early degenerative changes in musculoskeletal (MSK) tissues like cartilage and intervertebral discs (IVDs) but require long acquisition times. In MSK imaging, small features in cartilage and IVDs are crucial for diagnoses and must be preserved when reconstructing accelerated data. To these ends, we propose region of interest-specific postprocessing of accelerated acquisitions: a recurrent UNet deep learning architecture that provides T2 maps in knee cartilage, hip cartilage, and lumbar spine IVDs from accelerated T2-prepared snapshot gradient-echo acquisitions, optimizing for cartilage and IVD performance with a multi-component loss function that most heavily penalizes errors in those regions. Quantification errors in knee and hip cartilage were under 10% and 9% from acceleration factors R = 2 through 10, respectively, with bias for both under 3 ms for most of R = 2 through 12. In IVDs, mean quantification errors were under 12% from R = 2 through 6. A Gray Level Co-Occurrence Matrix-based scheme showed knee and hip pipelines outperformed state-of-the-art models, retaining smooth textures for most R and sharper ones through moderate R. Our methodology yields robust T2 maps while offering new approaches for optimizing and evaluating reconstruction algorithms to facilitate better preservation of small, clinically relevant features.

3D magnetic resonance fingerprinting for rapid simultaneous T1, T2, and T1ρ volumetric mapping of human articular cartilage at 3 T

Quantitative MRI can detect early biochemical changes in cartilage; however, the conventional techniques only measure one parameter (e.g., T1 , T2 , and T1ρ ) at a time while also being comparatively slow. We implemented a 3D magnetic resonance fingerprinting (3D-MRF) technique for simultaneous, volumetric mapping of T1 , T2 , and T1ρ in knee articular cartilage in under 9 min. It is evaluated on 11 healthy volunteers (mean age: 53 ± 9 years), five mild knee osteoarthritis (OA) patients (Kellgren-Lawrence (KL) score: 2, mean age: 60 ± 4 years), and the National Institute of Standards and Technology (NIST)/International Society for Magnetic Resonance in Medicine (ISMRM) system phantom. Proton density image, and T1 , T2, T1ρ relaxation times, and B1 + were estimated in the NIST/ISMRM system phantom as well as in the human knee medial and lateral femur, medial and lateral tibia, and patellar cartilage. The repeatability and reproducibility of the proposed technique were assessed in the phantom using analysis of the Bland-Altman plots. The intrasubject repeatability was assessed with the coefficient of variation (CV) and root mean square CV (rmsCV). The Mann-Whitney U test was used to assess the difference between healthy subjects and mild knee OA patients. The Bland-Altman plots in the NIST/ISMRM phantom demonstrated an average difference of 0.001% ± 015%, 1.2% ± 7.1%, and 0.47% ± 3% between two scans from the same 3-T scanner (repeatability), and 0.002% ± 015%, 0.62% ± 10.5%, and 0.97% ± 14% between the scans acquired on two different 3-T scanners (reproducibility) for T1 , T2 , and T1ρ , respectively. The in vivo knee study showed excellent repeatability with rmsCV less than 1%, 2%, and 1% for T1 , T2 , and T1ρ , respectively. T1ρ relaxation time in the mild knee OA patients was significantly higher (p < 0.05) than in healthy subjects. The proposed 3D-MRF sequence is fast, reproducible, robust to B1 + inhomogeneity, and can simultaneously measure the T1 , T2 , T1ρ , and B1 + volumetric maps of the knee joint in a single scan within a clinically feasible scan time.

Artificial intelligence in cardiac magnetic resonance fingerprinting

Magnetic resonance fingerprinting (MRF) is a fast MRI-based technique that allows for multiparametric quantitative characterization of the tissues of interest in a single acquisition. In particular, it has gained attention in the field of cardiac imaging due to its ability to provide simultaneous and co-registered myocardial T₁ and T₂ mapping in a single breath-held cardiac MRF scan, in addition to other parameters. Initial results in small healthy subject groups and clinical studies have demonstrated the feasibility and potential of MRF imaging. Ongoing research is being conducted to improve the accuracy, efficiency, and robustness of cardiac MRF. However, these improvements usually increase the complexity of image reconstruction and dictionary generation and introduce the need for sequence optimization. Each of these steps increase the computational demand and processing time of MRF. The latest advances in artificial intelligence (AI), including progress in deep learning and the development of neural networks for MRI, now present an opportunity to efficiently address these issues. Artificial intelligence can be used to optimize candidate sequences and reduce the memory demand and computational time required for reconstruction and post-processing. Recently, proposed machine learning-based approaches have been shown to reduce dictionary generation and reconstruction times by several orders of magnitude. Such applications of AI should help to remove these bottlenecks and speed up cardiac MRF, improving its practical utility and allowing for its potential inclusion in clinical routine. This review aims to summarize the latest developments in artificial intelligence applied to cardiac MRF. Particularly, we focus on the application of machine learning at different steps of the MRF process, such as sequence optimization, dictionary generation and image reconstruction.

MR fingerprinting of the prostate

Multiparametric magnetic resonance imaging (mpMRI) has been adopted as the key tool for detection, localization, characterization, and risk stratification of patients suspected to have prostate cancer. Despite advantages over systematic biopsy, the interpretation of prostate mpMRI has limitations including a steep learning curve, leading to considerable interobserver variation. There is growing interest in clinical translation of quantitative imaging techniques for more objective lesion assessment. However, traditional mapping techniques are slow, precluding their use in the clinic. Magnetic resonance fingerprinting (MRF) is an efficient approach for quantitative maps of multiple tissue properties simultaneously. The T₁ and T₂ values obtained with MRF have been validated with phantom studies as well as in normal volunteers and patients. Studies have shown that MRF-derived T₁ and T₂ along with ADC values are all significant independent predictors in the differentiation between normal prostate tissue and prostate cancer, and hold promise in differentiating low and intermediate/high-grade cancers. This review seeks to introduce the basics of the prostate MRF technique, discuss the potential applications of prostate MRF for the characterization of prostate cancer, and describes ongoing areas of research.

Improved Balanced Steady-State Free Precession Based MR Fingerprinting with Deep Autoencoders

Magnetic Resonance (MR) Fingerprinting is an emerging transient-state imaging paradigm, which enables the quantization of multiple MR tissue parameters in a single experiment. Balanced steady-state free precession (bSSFP)-based MR Fingerprinting has excellent signal-to-noise characteristics and also allows for acquiring both tissue parameter maps and field inhomogeneity maps. However, field inhomogeneity often results in complex magnetization evolutions in bSSFP-based MR Fingerprinting, which creates significant challenges in image reconstruction. In this paper, we introduce a new method to address the image reconstruction problem. The proposed method incorporates a low-dimensional nonlinear manifold learned from an ensemble of magnetization evolutions using a deep autoencoder. It provides much better representation power than a low-dimensional linear subspace in capturing complex magnetization evolutions. We formulate the image reconstruction problem with this nonlinear model and solve the resulting optimization problem using an algorithm based on variable splitting and the alternating direction method of multipliers. We evaluate the performance of the proposed method using numerical experiments and demonstrate that it significantly outperforms the state-of-art method using a linear subspace model.

A Self-Supervised Deep Learning Reconstruction for Shortening the Breathhold and Acquisition Window in Cardiac Magnetic Resonance Fingerprinting

The aim of this study is to shorten the breathhold and diastolic acquisition window in cardiac magnetic resonance fingerprinting (MRF) for simultaneous T1, T2, and proton spin density (M0) mapping to improve scan efficiency and reduce motion artifacts. To this end, a novel reconstruction was developed that combines low-rank subspace modeling with a deep image prior, termed DIP-MRF. A system of neural networks is used to generate spatial basis images and quantitative tissue property maps, with training performed using only the undersampled k-space measurements from the current scan. This approach avoids difficulties with obtaining in vivo MRF training data, as training is performed de novo for each acquisition. Calculation of the forward model during training is accelerated by using GRAPPA operator gridding to shift spiral k-space data to Cartesian grid points, and by using a neural network to rapidly generate fingerprints in place of a Bloch equation simulation. DIP-MRF was evaluated in simulations and at 1.5 T in a standardized phantom, 18 healthy subjects, and 10 patients with suspected cardiomyopathy. In addition to conventional mapping, two cardiac MRF sequences were acquired, one with a 15-heartbeat(HB) breathhold and 254 ms acquisition window, and one with a 5HB breathhold and 150 ms acquisition window. In simulations, DIP-MRF yielded decreased nRMSE compared to dictionary matching and a sparse and locally low rank (SLLR-MRF) reconstruction. Strong correlation (R2 &gt; 0.999) with T1 and T2 reference values was observed in the phantom using the 5HB/150 ms scan with DIP-MRF. DIP-MRF provided better suppression of noise and aliasing artifacts in vivo, especially for the 5HB/150 ms scan, and lower intersubject and intrasubject variability compared to dictionary matching and SLLR-MRF. Furthermore, it yielded a better agreement between myocardial T1 and T2 from 15HB/254 ms and 5HB/150 ms MRF scans, with a bias of −9 ms for T1 and 2 ms for T2. In summary, this study introduces an extension of the deep image prior framework for cardiac MRF tissue property mapping, which does not require pre-training with in vivo scans, and has the potential to reduce motion artifacts by enabling a shortened breathhold and acquisition window.

An end-to-end AI-based framework for automated discovery of rapid CEST/MT MRI acquisition protocols and molecular parameter quantification (AutoCEST)

PURPOSE

To develop an automated machine-learning-based method for the discovery of rapid and quantitative chemical exchange saturation transfer (CEST) MR fingerprinting acquisition and reconstruction protocols.

METHODS

An MR physics-governed AI system was trained to generate optimized acquisition schedules and the corresponding quantitative reconstruction neural network. The system (termed AutoCEST) is composed of a CEST saturation block, a spin dynamics module, and a deep reconstruction network, all differentiable and jointly connected. The method was validated using a variety of chemical exchange phantoms and in vivo mouse brains at 9.4T.

RESULTS

The acquisition times for AutoCEST optimized schedules ranged from 35 to 71 s, with a quantitative image reconstruction time of only 29 ms. The resulting exchangeable proton concentration maps for the phantoms were in good agreement with the known solute concentrations for AutoCEST sequences (mean absolute error = 2.42 mM; Pearson's r=0.992 , p<0.0001 ), but not for an unoptimized sequence (mean absolute error = 65.19 mM; Pearson's r=-0.161 , p=0.522 ). Similarly, improved exchange rate agreement was observed between AutoCEST and quantification of exchange using saturation power (QUESP) methods (mean absolute error: 35.8 Hz, Pearson's r=0.971 , p<0.0001 ) compared to an unoptimized schedule and QUESP (mean absolute error = 58.2 Hz; Pearson's r=0.959 , p<0.0001 ). The AutoCEST in vivo mouse brain semi-solid proton volume fractions were lower in the cortex (12.77% ± 0.75%) compared to the white matter (19.80% ± 0.50%), as expected.

CONCLUSION

AutoCEST can automatically generate optimized CEST/MT acquisition protocols that can be rapidly reconstructed into quantitative exchange parameter maps.

Integrating mechanism-based modeling with biomedical imaging to build practical digital twins for clinical oncology

Digital twins employ mathematical and computational models to virtually represent a physical object (e.g., planes and human organs), predict the behavior of the object, and enable decision-making to optimize the future behavior of the object. While digital twins have been widely used in engineering for decades, their applications to oncology are only just emerging. Due to advances in experimental techniques quantitatively characterizing cancer, as well as advances in the mathematical and computational sciences, the notion of building and applying digital twins to understand tumor dynamics and personalize the care of cancer patients has been increasingly appreciated. In this review, we present the opportunities and challenges of applying digital twins in clinical oncology, with a particular focus on integrating medical imaging with mechanism-based, tissue-scale mathematical modeling. Specifically, we first introduce the general digital twin framework and then illustrate existing applications of image-guided digital twins in healthcare. Next, we detail both the imaging and modeling techniques that provide practical opportunities to build patient-specific digital twins for oncology. We then describe the current challenges and limitations in developing image-guided, mechanism-based digital twins for oncology along with potential solutions. We conclude by outlining five fundamental questions that can serve as a roadmap when designing and building a practical digital twin for oncology and attempt to provide answers for a specific application to brain cancer. We hope that this contribution provides motivation for the imaging science, oncology, and computational communities to develop practical digital twin technologies to improve the care of patients battling cancer.

An end-to-end AI-based framework for automated discovery of rapid CEST/MT MRI acquisition protocols and molecular parameter quantification (AutoCEST)

PURPOSE

To develop an automated machine-learning-based method for the discovery of rapid and quantitative chemical exchange saturation transfer (CEST) MR fingerprinting acquisition and reconstruction protocols.

METHODS

An MR physics-governed AI system was trained to generate optimized acquisition schedules and the corresponding quantitative reconstruction neural network. The system (termed AutoCEST) is composed of a CEST saturation block, a spin dynamics module, and a deep reconstruction network, all differentiable and jointly connected. The method was validated using a variety of chemical exchange phantoms and in vivo mouse brains at 9.4T.

RESULTS

The acquisition times for AutoCEST optimized schedules ranged from 35 to 71 s, with a quantitative image reconstruction time of only 29 ms. The resulting exchangeable proton concentration maps for the phantoms were in good agreement with the known solute concentrations for AutoCEST sequences (mean absolute error = 2.42 mM; Pearson's r = 0.992 , p < 0.0001 ), but not for an unoptimized sequence (mean absolute error = 65.19 mM; Pearson's r = - 0.161 , p = 0.522 ). Similarly, improved exchange rate agreement was observed between AutoCEST and quantification of exchange using saturation power (QUESP) methods (mean absolute error: 35.8 Hz, Pearson's r = 0.971 , p < 0.0001 ) compared to an unoptimized schedule and QUESP (mean absolute error = 58.2 Hz; Pearson's r = 0.959 , p < 0.0001 ). The AutoCEST in vivo mouse brain semi-solid proton volume fractions were lower in the cortex (12.77% ± 0.75%) compared to the white matter (19.80% ± 0.50%), as expected.

CONCLUSION

AutoCEST can automatically generate optimized CEST/MT acquisition protocols that can be rapidly reconstructed into quantitative exchange parameter maps.

Simultaneous bilateral T1 , T2 , and T1ρ relaxation mapping of the hip joint with magnetic resonance fingerprinting

Quantitative MRI can detect early biochemical changes in cartilage, but its bilateral use in clinical routines is challenging. The aim of this prospective study was to demonstrate the feasibility of magnetic resonance fingerprinting for bilateral simultaneous T₁ , T₂ , and T_1ρ mapping of the hip joint. The study population consisted of six healthy volunteers with no known trauma or pain in the hip. Monoexponential T₁ , T₂ , and T_1ρ relaxation components were assessed in femoral lateral, superolateral, and superomedial, and inferior, as well as acetabular, superolateral, and superomedial subregions in left and right hip cartilage. Aligned ranked nonparametric factorial analysis was used to assess the side's impact on the subregions. Kruskal-Wallis and Wilcoxon tests were used to compare subregions, and coefficient of variation to assess repeatability. Global averages of T₁ (676.0 ± 45.4 and 687.6 ± 44.5 ms), T₂ (22.5 ± 2.6 and 22.1 ± 2.5 ms), and T_1ρ (38.2 ± 5.5 and 38.2 ± 5.5 ms) were measured in the left and right hip, and articular cartilage, respectively. The Kruskal-Wallis test showed a significant difference between different subregions' relaxation times regardless of the hip side (p < 0.001 for T₁ , p = 0.012 for T₂ , and p < 0.001 for T_1ρ ). The Wilcoxon test showed that T₁ of femoral layers was significantly (p < 0.003) higher than that for acetabular cartilage. The experiments showed excellent repeatability with CV_rms of 1%, 2%, and 4% for T₁ , T₂ , and T_1ρ, respectively. It was concluded that bilateral T₁ , T₂ , and T_1ρ relaxation times, as well as B₁⁺ maps, can be acquired simultaneously from hip joints using the proposed MRF sequence.

Quantitative imaging of apoptosis following oncolytic virotherapy by magnetic resonance fingerprinting aided by deep learning

Non-invasive imaging methods for detecting intratumoural viral spread and host responses to oncolytic virotherapy are either slow, lack specificity or require the use of radioactive or metal-based contrast agents. Here we show that in mice with glioblastoma multiforme, the early apoptotic responses to oncolytic virotherapy (characterized by decreased cytosolic pH and reduced protein synthesis) can be rapidly detected via chemical-exchange-saturation-transfer magnetic resonance fingerprinting (CEST-MRF) aided by deep learning. By leveraging a deep neural network trained with simulated magnetic resonance fingerprints, CEST-MRF can generate quantitative maps of intratumoural pH and of protein and lipid concentrations by selectively labelling the exchangeable amide protons of endogenous proteins and the exchangeable macromolecule protons of lipids, without requiring exogenous contrast agents. We also show that in a healthy volunteer, CEST-MRF yielded molecular parameters that are in good agreement with values from the literature. Deep-learning-aided CEST-MRF may also be amenable to the characterization of host responses to other cancer therapies and to the detection of cardiac and neurological pathologies.

Learning-based optimization of acquisition schedule for magnetization transfer contrast MR fingerprinting

Magnetization transfer contrast MR fingerprinting (MTC-MRF) is a novel quantitative imaging method that simultaneously quantifies free bulk water and semisolid macromolecule parameters using pseudo-randomized scan parameters. To improve acquisition efficiency and reconstruction accuracy, the optimization of MRF sequence design has been of recent interest in the MRF field, but has been challenging due to the large number of degrees of freedom to be optimized in the sequence. Herein, we propose a framework for learning-based optimization of the acquisition schedule (LOAS), which optimizes RF saturation-encoded MRF acquisitions with a minimal number of scan parameters for tissue parameter determination. In a supervised learning framework, scan parameters were subsequently updated to minimize a predefined loss function that can directly represent tissue quantification errors. We evaluated the performance of the proposed approach with a numerical phantom and in in vivo experiments. For validation, MRF images were synthesized using the tissue parameters estimated from a fully connected neural network framework and compared with references. Our results showed that LOAS outperformed existing indirect optimization methods with regard to quantification accuracy and acquisition efficiency. The proposed LOAS method could be a powerful optimization tool in the design of MRF pulse sequences.

Understanding the Research Landscape of Deep Learning in Biomedical Science: Scientometric Analysis

BACKGROUND

Advances in biomedical research using deep learning techniques have generated a large volume of related literature. However, there is a lack of scientometric studies that provide a bird's-eye view of them. This absence has led to a partial and fragmented understanding of the field and its progress.

OBJECTIVE

This study aimed to gain a quantitative and qualitative understanding of the scientific domain by analyzing diverse bibliographic entities that represent the research landscape from multiple perspectives and levels of granularity.

METHODS

We searched and retrieved 978 deep learning studies in biomedicine from the PubMed database. A scientometric analysis was performed by analyzing the metadata, content of influential works, and cited references.

RESULTS

In the process, we identified the current leading fields, major research topics and techniques, knowledge diffusion, and research collaboration. There was a predominant focus on applying deep learning, especially convolutional neural networks, to radiology and medical imaging, whereas a few studies focused on protein or genome analysis. Radiology and medical imaging also appeared to be the most significant knowledge sources and an important field in knowledge diffusion, followed by computer science and electrical engineering. A coauthorship analysis revealed various collaborations among engineering-oriented and biomedicine-oriented clusters of disciplines.

CONCLUSIONS

This study investigated the landscape of deep learning research in biomedicine and confirmed its interdisciplinary nature. Although it has been successful, we believe that there is a need for diverse applications in certain areas to further boost the contributions of deep learning in addressing biomedical research problems. We expect the results of this study to help researchers and communities better align their present and future work.

Circumventing the curse of dimensionality in magnetic resonance fingerprinting through a deep learning approach

Magnetic resonance fingerprinting (MRF) is a rapidly developing approach for fast quantitative MRI. A typical drawback of dictionary-based MRF is an explosion of the dictionary size as a function of the number of reconstructed parameters, according to the "curse of dimensionality", which determines an explosion of resource requirements. Neural networks (NNs) have been proposed as a feasible alternative, but this approach is still in its infancy. In this work, we design a deep learning approach to MRF using a fully connected network (FCN). In the first part we investigate, by means of simulations, how the NN performance scales with the number of parameters to be retrieved in comparison with the standard dictionary approach. Four MRF sequences were considered: IR-FISP, bSSFP, IR-FISP-B₁ , and IR-bSSFP-B₁ , the latter two designed to be more specific for B1+ parameter encoding. Estimation accuracy, memory usage, and computational time required to perform the estimation task were considered to compare the scalability capabilities of the dictionary-based and the NN approaches. In the second part we study optimal training procedures by including different data augmentation and preprocessing strategies during training to achieve better accuracy and robustness to noise and undersampling artifacts. The study is conducted using the IR-FISP MRF sequence exploiting both simulations and in vivo acquisitions. Results demonstrate that the NN approach outperforms the dictionary-based approach in terms of scalability capabilities. Results also allow us to heuristically determine the optimal training strategy to make an FCN able to predict T₁ ,  T₂ , and M₀ maps that are in good agreement with those obtained with the original dictionary approach. k-SVD denoising is proposed and found to be critical as a preprocessing step to handle undersampled data.

Fast Quantitative Low-Field Magnetic Resonance Imaging With OPTIMUM-Optimized Magnetic Resonance Fingerprinting Using a Stationary Steady-State Cartesian Approach and Accelerated Acquisition Schedules

OBJECTIVE

The aim of the proposed work is to develop model-based, fast multiparametric magnetic resonance imaging (MRI) in field regimes where signal-to-noise ratio is poor, such as encountered at low-field and in low γ nuclei.

MATERIALS AND METHODS

A custom, optimized MRI pipeline was developed at low field (0.1 T) that relies on the magnetic resonance fingerprinting framework, called OPTIMUM. An optimization algorithm was used to select a short acquisition schedule (n = 18 images) that favors maximal discrimination across varying magnetic properties (T1, T2) and off-resonance effects while maintaining high transverse magnetization at the steady state. In the presented study, a stationary balanced steady-state approach was investigated that allows for Cartesian (used here) and non-Cartesian acquisition schemes. Images were collected in calibrated samples containing different concentrations of manganese(II) chloride (MnCl2) in deionized water and compared with gold standard techniques (ie, inversion recovery for T1, Carr-Purcell-Meiboom-Gill for T2). Images were then collected in vivo in the human hand and wrist.

RESULTS

OPTIMUM successfully provided sets of quantified maps (T1, T2, T2*, M0, ΔB0, B1+) in calibrated samples and in vivo in the human hand and wrist in 3 dimensions, in ~8.5 minutes, with a voxel resolution of [1.5 ×1.5 × 6.5] mm3. Relaxation parameters (T1, T2) scale linearly with [MnCl2] and are in good agreement with the calibrations performed for T1, with a consistent trend to underestimate T2.

CONCLUSION

We show that low-field MRI can benefit from innovative multiparametric approaches to gain speed and become realistic in clinical environments. For the first time, we report simultaneous, multiparametric imaging (6 quantitative maps) in 3 dimensions, in vivo in the human hand and wrist, obtained in just 8.5 minutes. It is sometimes overlooked that low magnetic fields provide higher dispersion of nuclear spin relaxation rates. Rapid quantification such as offered by OPTIMUM could be an enabling technology to explore new metrics and contrasts in point-of-care MRI diagnosis, making it an important step toward broad democratization.

Rapid MR relaxometry using deep learning: An overview of current techniques and emerging trends

Quantitative mapping of MR tissue parameters such as the spin-lattice relaxation time (T₁ ), the spin-spin relaxation time (T₂ ), and the spin-lattice relaxation in the rotating frame (T_1ρ ), referred to as MR relaxometry in general, has demonstrated improved assessment in a wide range of clinical applications. Compared with conventional contrast-weighted (eg T₁ -, T₂ -, or T_1ρ -weighted) MRI, MR relaxometry provides increased sensitivity to pathologies and delivers important information that can be more specific to tissue composition and microenvironment. The rise of deep learning in the past several years has been revolutionizing many aspects of MRI research, including image reconstruction, image analysis, and disease diagnosis and prognosis. Although deep learning has also shown great potential for MR relaxometry and quantitative MRI in general, this research direction has been much less explored to date. The goal of this paper is to discuss the applications of deep learning for rapid MR relaxometry and to review emerging deep-learning-based techniques that can be applied to improve MR relaxometry in terms of imaging speed, image quality, and quantification robustness. The paper is comprised of an introduction and four more sections. Section 2 describes a summary of the imaging models of quantitative MR relaxometry. In Section 3, we review existing "classical" methods for accelerating MR relaxometry, including state-of-the-art spatiotemporal acceleration techniques, model-based reconstruction methods, and efficient parameter generation approaches. Section 4 then presents how deep learning can be used to improve MR relaxometry and how it is linked to conventional techniques. The final section concludes the review by discussing the promise and existing challenges of deep learning for rapid MR relaxometry and potential solutions to address these challenges.

Cramér-Rao bound-informed training of neural networks for quantitative MRI

PURPOSE

To improve the performance of neural networks for parameter estimation in quantitative MRI, in particular when the noise propagation varies throughout the space of biophysical parameters.

THEORY AND METHODS

A theoretically well-founded loss function is proposed that normalizes the squared error of each estimate with respective Cramér-Rao bound (CRB)-a theoretical lower bound for the variance of an unbiased estimator. This avoids a dominance of hard-to-estimate parameters and areas in parameter space, which are often of little interest. The normalization with corresponding CRB balances the large errors of fundamentally more noisy estimates and the small errors of fundamentally less noisy estimates, allowing the network to better learn to estimate the latter. Further, proposed loss function provides an absolute evaluation metric for performance: A network has an average loss of 1 if it is a maximally efficient unbiased estimator, which can be considered the ideal performance. The performance gain with proposed loss function is demonstrated at the example of an eight-parameter magnetization transfer model that is fitted to phantom and in vivo data.

RESULTS

Networks trained with proposed loss function perform close to optimal, that is, their loss converges to approximately 1, and their performance is superior to networks trained with the standard mean-squared error (MSE). The proposed loss function reduces the bias of the estimates compared to the MSE loss, and improves the match of the noise variance to the CRB. This performance gain translates to in vivo maps that align better with the literature.

CONCLUSION

Normalizing the squared error with the CRB during the training of neural networks improves their performance in estimating biophysical parameters.

Deep learning-based image reconstruction and post-processing methods in positron emission tomography for low-dose imaging and resolution enhancement

Image processing plays a crucial role in maximising diagnostic quality of positron emission tomography (PET) images. Recently, deep learning methods developed across many fields have shown tremendous potential when applied to medical image enhancement, resulting in a rich and rapidly advancing literature surrounding this subject. This review encapsulates methods for integrating deep learning into PET image reconstruction and post-processing for low-dose imaging and resolution enhancement. A brief introduction to conventional image processing techniques in PET is firstly presented. We then review methods which integrate deep learning into the image reconstruction framework as either deep learning-based regularisation or as a fully data-driven mapping from measured signal to images. Deep learning-based post-processing methods for low-dose imaging, temporal resolution enhancement and spatial resolution enhancement are also reviewed. Finally, the challenges associated with applying deep learning to enhance PET images in the clinical setting are discussed and future research directions to address these challenges are presented.

A review and experimental evaluation of deep learning methods for MRI reconstruction

Following the success of deep learning in a wide range of applications, neural network-based machine-learning techniques have received significant interest for accelerating magnetic resonance imaging (MRI) acquisition and reconstruction strategies. A number of ideas inspired by deep learning techniques for computer vision and image processing have been successfully applied to nonlinear image reconstruction in the spirit of compressed sensing for accelerated MRI. Given the rapidly growing nature of the field, it is imperative to consolidate and summarize the large number of deep learning methods that have been reported in the literature, to obtain a better understanding of the field in general. This article provides an overview of the recent developments in neural-network based approaches that have been proposed specifically for improving parallel imaging. A general background and introduction to parallel MRI is also given from a classical view of k-space based reconstruction methods. Image domain based techniques that introduce improved regularizers are covered along with k-space based methods which focus on better interpolation strategies using neural networks. While the field is rapidly evolving with plenty of papers published each year, in this review, we attempt to cover broad categories of methods that have shown good performance on publicly available data sets. Limitations and open problems are also discussed and recent efforts for producing open data sets and benchmarks for the community are examined.

Optimized multi-axis spiral projection MR fingerprinting with subspace reconstruction for rapid whole-brain high-isotropic-resolution quantitative imaging

PURPOSE

To improve image quality and accelerate the acquisition of 3D MR fingerprinting (MRF).

METHODS

Building on the multi-axis spiral-projection MRF technique, a subspace reconstruction with locally low-rank constraint and a modified spiral-projection spatiotemporal encoding scheme called tiny golden-angle shuffling were implemented for rapid whole-brain high-resolution quantitative mapping. Reconstruction parameters such as the locally low-rank regularization parameter and the subspace rank were tuned using retrospective in vivo data and simulated examinations. B₀ inhomogeneity correction using multifrequency interpolation was incorporated into the subspace reconstruction to further improve the image quality by mitigating blurring caused by off-resonance effect.

RESULTS

The proposed MRF acquisition and reconstruction framework yields high-quality 1-mm isotropic whole-brain quantitative maps in 2 min at better quality compared with 6-min acquisitions of prior approaches. The proposed method was validated to not induce bias in T₁ and T₂ mapping. High-quality whole-brain MRF data were also obtained at 0.66-mm isotropic resolution in 4 min using the proposed technique, where the increased resolution was shown to improve visualization of subtle brain structures.

CONCLUSIONS

The proposed tiny golden-angle shuffling, MRF with optimized spiral-projection trajectory and subspace reconstruction enables high-resolution quantitative mapping in ultrafast acquisition time.

MR fingerprinting for semisolid magnetization transfer and chemical exchange saturation transfer quantification

Chemical exchange saturation transfer (CEST) MRI has positioned itself as a promising contrast mechanism, capable of providing molecular information at sufficient resolution and amplified sensitivity. However, it has not yet become a routinely employed clinical technique, due to a variety of confounding factors affecting its contrast-weighted image interpretation and the inherently long scan time. CEST MR fingerprinting (MRF) is a novel approach for addressing these challenges, allowing simultaneous quantitation of several proton exchange parameters using rapid acquisition schemes. Recently, a number of deep-learning algorithms have been developed to further boost the performance and speed of CEST and semi-solid macromolecule magnetization transfer (MT) MRF. This review article describes the fundamental theory behind semisolid MT/CEST-MRF and its main applications. It then details supervised and unsupervised learning approaches for MRF image reconstruction and describes artificial intelligence (AI)-based pipelines for protocol optimization. Finally, practical considerations are discussed, and future perspectives are given, accompanied by basic demonstration code and data.

Acceleration of Magnetic Resonance Fingerprinting Reconstruction Using Denoising and Self-Attention Pyramidal Convolutional Neural Network

Magnetic resonance fingerprinting (MRF) based on echo-planar imaging (EPI) enables whole-brain imaging to rapidly obtain T1 and T2* relaxation time maps. Reconstructing parametric maps from the MRF scanned baselines by the inner-product method is computationally expensive. We aimed to accelerate the reconstruction of parametric maps for MRF-EPI by using a deep learning model. The proposed approach uses a two-stage model that first eliminates noise and then regresses the parametric maps. Parametric maps obtained by dictionary matching were used as a reference and compared with the prediction results of the two-stage model. MRF-EPI scans were collected from 32 subjects. The signal-to-noise ratio increased significantly after the noise removal by the denoising model. For prediction with scans in the testing dataset, the mean absolute percentage errors between the standard and the final two-stage model were 3.1%, 3.2%, and 1.9% for T1, and 2.6%, 2.3%, and 2.8% for T2* in gray matter, white matter, and lesion locations, respectively. Our proposed two-stage deep learning model can effectively remove noise and accurately reconstruct MRF-EPI parametric maps, increasing the speed of reconstruction and reducing the storage space required by dictionaries.

Tailored Magnetic Resonance Fingerprinting for simultaneous non-synthetic and quantitative imaging: A repeatability study

PURPOSE

The goals of this study include: (a) generating tailored magnetic resonance fingerprinting (TMRF) based non-synthetic imaging; (b) assessing the repeatability of TMRF and deep learning-based mapping of in vitro ISMRM/NIST phantom and in vivo brain data of healthy human subjects.

METHODS

We have acquired qualitative images obtained from the vendor-supplied gold standard, MRF (synthetic), and TMRF (non-synthetic) on one representative healthy human brain. We also acquired thirty datasets on the ISMRM/NIST phantom for the in vitro repeatability study on a GE Discovery 3T MR750w scanner using the TMRF sequence. We compared T₁ and T₂ maps generated from thirty ISMRM/NIST phantom datasets to the spin-echo (SE) based gold standard (GS) method as part of the in vitro repeatability study. R-squared coefficient of determination in a simple linear regression and Bland-Altman analysis were computed for thirty datasets of ISMRM/NIST phantom to assess the accuracy of in vitro quantitative TMRF data. The repeatability of T₁ and T₂ estimates by TMRF was evaluated by calculating the standard deviation (SD) divided by the average of thirty datasets for each sphere, respectively. We acquired ten volunteers for the in vivo repeatability study on the same scanner using the same TMRF sequence. These volunteers were imaged five times with two runs per repetition, resulting in one hundred in vivo datasets. Five contrasts, T₁ and T₂ maps of ten human volunteers acquired over five repetitions, were evaluated in the in vivo repeatability study. We computed the intraclass correlation coefficient (ICC) of the signal-to-noise ratio (SNR), signal intensities, T₁ and T₂ relaxation times in white matter (WM) and gray matter (GM).

RESULTS

The synthetic images generated from MRF show partial volume and flow artifacts compared to non-synthetic images obtained from TMRF images and the gold standard. in vitro studies show that TMRF estimates have less than 5% variations except sphere 14 in the T₂ array (6.36%). TMRF and spin-echo relaxometry measurements were strongly correlated; R² values were 0.9958 and 0.9789 for T₁ and T₂ estimates, respectively. Based on the ICC values, SNR, mean intensity values, and relaxation times of WM and GM for the in vivo studies were consistent. T₁ and T₂ values of WM and GM were similar to previously published values. The mean ± SD of T₁ and T₂ for WM for ten subjects and five repeats are 992 ± 41 ms and 99 ± 6 ms, while the corresponding values for T₁ and T₂ for GM are 1598 ± 73 ms and 152 ± 14 ms.

CONCLUSION

TMRF and deep learning-based reconstruction produce repeatable, non-synthetic multi-contrast images and parametric maps simultaneously. This article is protected by copyright. All rights reserved.

Accelerated cardiac T1 mapping in four heartbeats with inline MyoMapNet: a deep learning-based T1 estimation approach

PURPOSE

To develop and evaluate MyoMapNet, a rapid myocardial T1 mapping approach that uses fully connected neural networks (FCNN) to estimate T1 values from four T1-weighted images collected after a single inversion pulse in four heartbeats (Look-Locker, LL4).

METHOD

We implemented an FCNN for MyoMapNet to estimate T1 values from a reduced number of T1-weighted images and corresponding inversion-recovery times. We studied MyoMapNet performance when trained using native, post-contrast T1, or a combination of both. We also explored the effects of number of T1-weighted images (four and five) for native T1. After rigorous training using in-vivo modified Look-Locker inversion recovery (MOLLI) T1 mapping data of 607 patients, MyoMapNet performance was evaluated using MOLLI T1 data from 61 patients by discarding the additional T1-weighted images. Subsequently, we implemented a prototype MyoMapNet and LL4 on a 3 T scanner. LL4 was used to collect T1 mapping data in 27 subjects with inline T1 map reconstruction by MyoMapNet. The resulting T1 values were compared to MOLLI.

RESULTS

MyoMapNet trained using a combination of native and post-contrast T1-weighted images had excellent native and post-contrast T1 accuracy compared to MOLLI. The FCNN model using four T1-weighted images yields similar performance compared to five T1-weighted images, suggesting that four T1 weighted images may be sufficient. The inline implementation of LL4 and MyoMapNet enables successful acquisition and reconstruction of T1 maps on the scanner. Native and post-contrast myocardium T1 by MOLLI and MyoMapNet was 1170 ± 55 ms vs. 1183 ± 57 ms (P = 0.03), and 645 ± 26 ms vs. 630 ± 30 ms (P = 0.60), and native and post-contrast blood T1 was 1820 ± 29 ms vs. 1854 ± 34 ms (P = 0.14), and 508 ± 9 ms vs. 514 ± 15 ms (P = 0.02), respectively.

CONCLUSION

A FCNN, trained using MOLLI data, can estimate T1 values from only four T1-weighted images. MyoMapNet enables myocardial T1 mapping in four heartbeats with similar accuracy as MOLLI with inline map reconstruction.

Accelerate gas diffusion-weighted MRI for lung morphometry with deep learning

OBJECTIVES

Multiple b-value gas diffusion-weighted MRI (DW-MRI) enables non-invasive and quantitative assessment of lung morphometry, but its long acquisition time is not well-tolerated by patients. We aimed to accelerate multiple b-value gas DW-MRI for lung morphometry using deep learning.

METHODS

A deep cascade of residual dense network (DC-RDN) was developed to reconstruct high-quality DW images from highly undersampled k-space data. Hyperpolarized ¹²⁹Xe lung ventilation images were acquired from 101 participants and were retrospectively collected to generate synthetic DW-MRI data to train the DC-RDN. Afterwards, the performance of the DC-RDN was evaluated on retrospectively and prospectively undersampled multiple b-value ¹²⁹Xe MRI datasets.

RESULTS

Each slice with size of 64 × 64 × 5 could be reconstructed within 7.2 ms. For the retrospective test data, the DC-RDN showed significant improvement on all quantitative metrics compared with the conventional reconstruction methods (p < 0.05). The apparent diffusion coefficient (ADC) and morphometry parameters were not significantly different between the fully sampled and DC-RDN reconstructed images (p > 0.05). For the prospectively accelerated acquisition, the required breath-holding time was reduced from 17.8 to 4.7 s with an acceleration factor of 4. Meanwhile, the prospectively reconstructed results showed good agreement with the fully sampled images, with a mean difference of -0.72% and -0.74% regarding global mean ADC and mean linear intercept (L_m) values.

CONCLUSIONS

DC-RDN is effective in accelerating multiple b-value gas DW-MRI while maintaining accurate estimation of lung microstructural morphometry, facilitating the clinical potential of studying lung diseases with hyperpolarized DW-MRI.

KEY POINTS

• The deep cascade of residual dense network allowed fast and high-quality reconstruction of multiple b-value gas diffusion-weighted MRI at an acceleration factor of 4. • The apparent diffusion coefficient and morphometry parameters were not significantly different between the fully sampled images and the reconstructed results (p > 0.05). • The required breath-holding time was reduced from 17.8 to 4.7 s and each slice with size of 64 × 64 × 5 could be reconstructed within 7.2 ms.

Accelerated reconstruction of dictionary-based T2 relaxation maps based on dictionary compression and gradient descent search algorithms

Background Quantitative T₂-relaxation-based contrast maps have shown to be highly beneficial for clinical diagnosis and follow-up. The generation of quantitative maps, however, is impaired by long acquisition times, and time-consuming post-processing schemes. The EMC platform is a dictionary-based technique, which involves simulating theoretical signal curves for different physical and experimental values, followed by matching the experimentally acquired signals to the set simulated ones. Purpose Although the EMC technique has shown to produce accurate T₂ maps, it involves computationally intensive post-processing procedures. In this work we present an approach for accelerating the reconstruction of T₂ relaxation maps. Methods This work presents two alternative post-processing approaches for accelerating the reconstruction of EMC-based T₂ relaxation maps. These are (a) Dictionary compression using principal component analysis (PCA) and (b) gradient-descent search algorithm. Additional acceleration was achieved by finding the optimal MATLAB C++ compiler. The utility of the two suggested approaches was examined by calculating the relative error, produced by each technique. Results Gradient descent method was in perfect agreement with the ground truth exhaustive search matching process. PCA based acceleration produced root mean square error (RMSE) of up to 4% compared to exhaustive matching process. Overall acceleration of x16 was achieved using gradient descent in addition to x7 acceleration by choosing the optimal MATLAB C++ compiler. Conclusions Postprocessing of EMC-based T₂ relaxation maps can be accelerated without impairing the accuracy of the ensuing T₂ values.

Simultaneous T1 , T2 , and T1ρ cardiac magnetic resonance fingerprinting for contrast agent-free myocardial tissue characterization

PURPOSE

To develop a simultaneous T₁ , T₂ , and T_1ρ cardiac magnetic resonance fingerprinting (MRF) approach to enable comprehensive contrast agent-free myocardial tissue characterization in a single breath-hold scan.

METHODS

A 2D gradient-echo electrocardiogram-triggered cardiac MRF sequence with low flip angles, varying magnetization preparation, and spiral trajectory was acquired at 1.5 T to encode T₁ , T₂ , and T_1⍴ simultaneously. The MRF images were reconstructed using low-rank inversion, regularized with a multicontrast patch-based higher-order reconstruction. Parametric maps were generated and matched in the singular value domain to extended phase graph-based dictionaries. The proposed approach was tested in phantoms and 10 healthy subjects and compared against conventional methods in terms of coefficients of determination and best fits for the phantom study, and in terms of Bland-Altman agreement, average values and coefficient of variation of T₁ , T₂ , and T_1⍴ for the healthy subjects study.

RESULTS

The T₁ , T₂ , and T_1⍴ MRF values showed excellent correlation with conventional spin-echo and clinical mapping methods in phantom studies (r² > 0.97). Measured MRF values in myocardial tissue (mean ± SD) were 1133 ± 33 ms, 38.8 ± 3.5 ms, and 52.0 ± 4.0 ms for T₁ , T₂ and T1_⍴ , respectively, against 1053 ± 47 ms, 50.4 ± 3.9 ms, and 55.9 ± 3.3 ms for T₁ modified Look-Locker inversion imaging, T₂ gradient and spin echo, and T_1⍴ turbo field echo, respectively.

CONCLUSION

A cardiac MRF approach for simultaneous quantification of myocardial T₁ , T₂ , and T_1ρ in a single breath-hold MR scan of about 16 seconds has been proposed. The approach has been investigated in phantoms and healthy subjects showing good agreement with reference spin echo measurements and conventional clinical maps.

Encoding capability prediction of acquisition schedules in CEST MR fingerprinting for pH quantification

PURPOSE

To identify a reliable metric for predicting the encoding capability of CEST MR fingerprinting acquisition schedules for pH quantification, which may facilitate CEST MR fingerprinting protocol optimization.

METHODS

Numerical simulations and Cr phantom MRI experiments were conducted at 3 Tesla under representative CEST MR fingerprinting sampling scenarios, including the pseudorandomization of imaging parameters (e.g., saturation power B₁ , saturation frequency offset, saturation time, and relaxation time), and variation of the maximum saturation power B_1max , B₁ number, and sampling pattern. The CEST effect at 2 ppm was measured using asymmetry analysis and matched to a predefined dictionary to determine the pH. The pH quantification error was assessed using RMSE. Three metrics, namely the Cramer-Rao bound, dot product, and Euclidean distance, were calculated for each sampling scenario, and their relationships with the pH RMSE were investigated to examine their effectiveness for predicting the encoding capability of sampling schedules for pH quantification.

RESULTS

Both simulation and phantom studies revealed that the Cramer-Rao bound metric consistently exhibited superior performance for predicting the pH quantification error. Although dot product exhibited good encoding capability prediction in most sampling scenarios, it failed in the scenario with varied B₁ numbers. In contrast, Euclidean distance exhibited the worst performance among the 3 metrics in all scenarios.

CONCLUSION

Superior over dot product and Euclidean distance, the Cramer-Rao bound metric may reliably predicting the encoding capability of CEST MR fingerprinting sampling strategies and may be useful for guiding CEST MRI protocol optimization.

Magnetic Resonance Fingerprinting of the Pediatric Brain

Magnetic resonance fingerprinting (MRF) is increasingly being used to evaluate brain development and differentiate normal and pathologic tissues in children. MRF can provide reliable and accurate intrinsic tissue properties, such as T1 and T2 relaxation times. MRF is a powerful tool in evaluating brain disease in pediatric population. MRF is a new quantitative MR imaging technique for rapid and simultaneous quantification of multiple tissue properties.

Technical overview of magnetic resonance fingerprinting and its applications in radiation therapy

Magnetic resonance fingerprinting (MRF) is an emerging imaging technique for rapid and simultaneous quantification of multiple tissue properties. The technique has been developed for quantitative imaging of different organs. The obtained quantitative measures have the potential to improve multiple steps of a typical radiotherapy workflow and potentially further improve integration of magnetic resonance imaging guided clinical decision making. In this review paper, we first provide a technical overview of the MRF method from data acquisition to postprocessing, along with recent development in advanced reconstruction methods. We further discuss critical aspects that could influence its usage in radiation therapy, such as accuracy and precision, repeatability and reproducibility, geometric distortion, and motion robustness. Finally, future directions for MRF application in radiation therapy are discussed.

Channel Attention Networks for Robust MR Fingerprint Matching

Magnetic Resonance Fingerprinting (MRF) enables simultaneous mapping of multiple tissue parameters such as T1 and T2 relaxation times. The working principle of MRF relies on varying acquisition parameters pseudo-randomly, so that each tissue generates its unique signal evolution during scanning. Even though MRF provides faster scanning, it has disadvantages such as erroneous and slow generation of the corresponding parametric maps, which needs to be improved. Moreover, there is a need for explainable architectures for understanding the guiding signals to generate accurate parametric maps. In this paper, we addressed both of these shortcomings by proposing a novel neural network architecture (CONV-ICA) consisting of a channel-wise attention module and a fully convolutional network. The proposed approach, evaluated over 3 simulated MRF signals, reduces error in the reconstruction of tissue parameters by 8.88% for T1 and 75.44% for T2 with respect to state-of-the-art methods. Another contribution of this study is a new channel selection method: attention-based channel selection. Furthermore, the effect of patch size and temporal frames of MRF signal on channel reduction are analyzed by employing a channel-wise attention.

Current Applications and Future Development of Magnetic Resonance Fingerprinting in Diagnosis, Characterization, and Response Monitoring in Cancer

Magnetic resonance imaging (MRI) has enabled non-invasive cancer diagnosis, monitoring, and management in common clinical settings. However, inadequate quantitative analyses in MRI continue to limit its full potential and these often have an impact on clinicians' judgments. Magnetic resonance fingerprinting (MRF) has recently been introduced to acquire multiple quantitative parameters simultaneously in a reasonable timeframe. Initial retrospective studies have demonstrated the feasibility of using MRF for various cancer characterizations. Further trials with larger cohorts are still needed to explore the repeatability and reproducibility of the data acquired by MRF. At the moment, technical difficulties such as undesirable processing time or lack of motion robustness are limiting further implementations of MRF in clinical oncology. This review summarises the latest findings and technology developments for the use of MRF in cancer management and suggests possible future implications of MRF in characterizing tumour heterogeneity and response assessment.

Training data distribution significantly impacts the estimation of tissue microstructure with machine learning

PURPOSE

Supervised machine learning (ML) provides a compelling alternative to traditional model fitting for parameter mapping in quantitative MRI. The aim of this work is to demonstrate and quantify the effect of different training data distributions on the accuracy and precision of parameter estimates when supervised ML is used for fitting.

METHODS

We fit a two- and three-compartment biophysical model to diffusion measurements from in-vivo human brain, as well as simulated diffusion data, using both traditional model fitting and supervised ML. For supervised ML, we train several artificial neural networks, as well as random forest regressors, on different distributions of ground truth parameters. We compare the accuracy and precision of parameter estimates obtained from the different estimation approaches using synthetic test data.

RESULTS

When the distribution of parameter combinations in the training set matches those observed in healthy human data sets, we observe high precision, but inaccurate estimates for atypical parameter combinations. In contrast, when training data is sampled uniformly from the entire plausible parameter space, estimates tend to be more accurate for atypical parameter combinations but may have lower precision for typical parameter combinations.

CONCLUSION

This work highlights that estimation of model parameters using supervised ML depends strongly on the training-set distribution. We show that high precision obtained using ML may mask strong bias, and visual assessment of the parameter maps is not sufficient for evaluating the quality of the estimates.

Deep Learning and Its Application to Function Approximation for MR in Medicine: An Overview

This article presents an overview of deep learning (DL) and its applications to function approximation for MR in medicine. The aim of this article is to help readers develop various applications of DL. DL has made a large impact on the literature of many medical sciences, including MR. However, its technical details are not easily understandable for non-experts of machine learning (ML).The first part of this article presents an overview of DL and its related technologies, such as artificial intelligence (AI) and ML. AI is explained as a function that can receive many inputs and produce many outputs. ML is a process of fitting the function to training data. DL is a kind of ML, which uses a composite of many functions to approximate the function of interest. This composite function is called a deep neural network (DNN), and the functions composited into a DNN are called layers. This first part also covers the underlying technologies required for DL, such as loss functions, optimization, initialization, linear layers, non-linearities, normalization, recurrent neural networks, regularization, data augmentation, residual connections, autoencoders, generative adversarial networks, model and data sizes, and complex-valued neural networks.The second part of this article presents an overview of the applications of DL in MR and explains how functions represented as DNNs are applied to various applications, such as RF pulse, pulse sequence, reconstruction, motion correction, spectroscopy, parameter mapping, image synthesis, and segmentation.

A data-driven T2 relaxation analysis approach for myelin water imaging: Spectrum analysis for multiple exponentials via experimental condition oriented simulation (SAME-ECOS)

PURPOSE

The decomposition of multi-exponential decay data into a T₂ spectrum poses substantial challenges for conventional fitting algorithms, including non-negative least squares (NNLS). Based on a combination of the resolution limit constraint and machine learning neural network algorithm, a data-driven and highly tailorable analysis method named spectrum analysis for multiple exponentials via experimental condition oriented simulation (SAME-ECOS) was proposed.

THEORY AND METHODS

The theory of SAME-ECOS was derived. Then, a paradigm was presented to demonstrate the SAME-ECOS workflow, consisting of a series of calculation, simulation, and model training operations. The performance of the trained SAME-ECOS model was evaluated using simulations and six in vivo brain datasets. The code is available at https://github.com/hanwencat/SAME-ECOS.

RESULTS

Using NNLS as the baseline, SAME-ECOS achieved over 15% higher overall cosine similarity scores in producing the T₂ spectrum, and more than 10% lower mean absolute error in calculating the myelin water fraction (MWF), as well as demonstrated better robustness to noise in the simulation tests. Applying to in vivo data, MWF from SAME-ECOS and NNLS was highly correlated among all study participants. However, a distinct separation of the myelin water peak and the intra/extra-cellular water peak was only observed in the mean T₂ spectra determined using SAME-ECOS. In terms of data processing speed, SAME-ECOS is approximately 30 times faster than NNLS, achieving a whole-brain analysis in 3 min.

CONCLUSION

Compared with NNLS, the SAME-ECOS method yields much more reliable T₂ spectra in a dramatically shorter time, increasing the feasibility of multi-component T₂ decay analysis in clinical settings.

Real-Time Mapping of Tissue Properties for Magnetic Resonance Fingerprinting

Magnetic resonance fingerprinting (MRF) is a relatively new multi-parametric quantitative imaging method that involves a two-step process: (i) reconstructing a series of time frames from highly-undersampled non-Cartesian spiral k-space data and (ii) pattern matching using the time frames to infer tissue properties (e.g., T 1 and T 2 relaxation times). In this paper, we introduce a novel end-to-end deep learning framework to seamlessly map the tissue properties directly from spiral k-space MRF data, thereby avoiding time-consuming processing such as the non-uniform fast Fourier transform (NUFFT) and the dictionary-based fingerprint matching. Our method directly consumes the non-Cartesian k -space data, performs adaptive density compensation, and predicts multiple tissue property maps in one forward pass. Experiments on both 2D and 3D MRF data demonstrate that quantification accuracy comparable to state-of-the-art methods can be accomplished within 0.5 s, which is 1,100 to 7,700 times faster than the original MRF framework. The proposed method is thus promising for facilitating the adoption of MRF in clinical settings.

Quantitative MR relaxation using MR fingerprinting with fractional-order signal evolution

The fractional-order Bloch equations have been shown to describe a wider range of experimental situations involving heterogeneous, porous, or composite materials. This paper introduces a novel dictionary of quantitative MR fingerprinting generated by signal evolution model with fractional-order Bloch equations to describe magnetic resonance (MR) relaxation. Here, the fractional-order relaxation models are implemented into Bloch equations through phase transitions using EPG simulation. In the phantom experiments, the fractional-order analysis showed smaller root mean squared error (T1: RMSE = 5.21%, T2: RMSE=3.75%) using the proposed method compared to using conventional method. Among the in vivo experiments of human brains, the estimated T1 and T2 values (mean ± SD) were 843 ± 46.3 ms and 70 ± 4.7 ms in white matter, 1323 ± 28.5 ms and 95 ± 3.8 ms in gray matter. So the proposed method can provide well extensions of current MR fingerprinting and has shown potential to apply into the phantom experiments and the in vivo applications to approach the standard methods for quantitative imaging.

Machine Learning in Drug Discovery: A Review

This review provides the feasible literature on drug discovery through ML tools and techniques that are enforced in every phase of drug development to accelerate the research process and deduce the risk and expenditure in clinical trials. Machine learning techniques improve the decision-making in pharmaceutical data across various applications like QSAR analysis, hit discoveries, de novo drug architectures to retrieve accurate outcomes. Target validation, prognostic biomarkers, digital pathology are considered under problem statements in this review. ML challenges must be applicable for the main cause of inadequacy in interpretability outcomes that may restrict the applications in drug discovery. In clinical trials, absolute and methodological data must be generated to tackle many puzzles in validating ML techniques, improving decision-making, promoting awareness in ML approaches, and deducing risk failures in drug discovery.

A deep learning approach for magnetic resonance fingerprinting: Scaling capabilities and good training practices investigated by simulations

MR fingerprinting (MRF) is an innovative approach to quantitative MRI. A typical disadvantage of dictionary-based MRF is the explosive growth of the dictionary as a function of the number of reconstructed parameters, an instance of the curse of dimensionality, which determines an explosion of resource requirements. In this work, we describe a deep learning approach for MRF parameter map reconstruction using a fully connected architecture. Employing simulations, we have investigated how the performance of the Neural Networks (NN) approach scales with the number of parameters to be retrieved, compared to the standard dictionary approach. We have also studied optimal training procedures by comparing different strategies for noise addition and parameter space sampling, to achieve better accuracy and robustness to noise. Four MRF sequences were considered: IR-FISP, bSSFP, IR-FISP-B₁, and IR-bSSFP-B₁. A comparison between NN and the dictionary approaches in reconstructing parameter maps as a function of the number of parameters to be retrieved was performed using a numerical brain phantom. Results demonstrated that training with random sampling and different levels of noise variance yielded the best performance. NN performance was at least as good as the dictionary-based approach in reconstructing parameter maps using Gaussian noise as a source of artifacts: the difference in performance increased with the number of estimated parameters because the dictionary method suffers from the coarse resolution of the parameter space sampling. The NN proved to be more efficient in memory usage and computational burden, and has great potential for solving large-scale MRF problems.

A multi-inversion multi-echo spin and gradient echo echo planar imaging sequence with low image distortion for rapid quantitative parameter mapping and synthetic image contrasts

PURPOSE

Brain imaging exams typically take 10-20 min and involve multiple sequential acquisitions. A low-distortion whole-brain echo planar imaging (EPI)-based approach was developed to efficiently encode multiple contrasts in one acquisition, allowing for calculation of quantitative parameter maps and synthetic contrast-weighted images.

METHODS

Inversion prepared spin- and gradient-echo EPI was developed with slice-order shuffling across measurements for efficient acquisition with T1 , T2 , and T 2 ∗ weighting. A dictionary-matching approach was used to fit the images to quantitative parameter maps, which in turn were used to create synthetic weighted images with typical clinical contrasts. Dynamic slice-optimized multi-coil shimming with a B0 shim array was used to reduce B0 inhomogeneity and, therefore, image distortion by >50%. Multi-shot EPI was also implemented to minimize distortion and blurring while enabling high in-plane resolution. A low-rank reconstruction approach was used to mitigate errors from shot-to-shot phase variation.

RESULTS

The slice-optimized shimming approach was combined with in-plane parallel-imaging acceleration of 4× to enable single-shot EPI with more than eight-fold distortion reduction. The proposed sequence efficiently obtained 40 contrasts across the whole-brain in just over 1 min at 1.2 × 1.2 × 3 mm resolution. The multi-shot variant of the sequence achieved higher in-plane resolution of 1 × 1 × 4 mm with good image quality in 4 min. Derived quantitative maps showed comparable values to conventional mapping methods.

CONCLUSION

The approach allows fast whole-brain imaging with quantitative parameter maps and synthetic weighted contrasts. The slice-optimized multi-coil shimming and multi-shot reconstruction approaches result in minimal EPI distortion, giving the sequence the potential to be used in rapid screening applications.

Efficiency analysis for quantitative MRI of T1 and T2 relaxometry methods

This study presents a comparison of quantitative MRI methods based on an efficiency metric that quantifies their intrinsic ability to extract information about tissue parameters. Under a regime of unbiased parameter estimates, an intrinsic efficiency metricηwas derived for fully-sampled experiments which can be used to both optimize and compare sequences. Here we optimize and compare several steady-state and transient gradient-echo based qMRI methods, such as magnetic resonance fingerprinting (MRF), for jointT1andT2mapping. The impact of undersampling was also evaluated, assuming incoherent aliasing that is treated as noise by parameter estimation.In vivovalidation of the efficiency metric was also performed. Transient methods such as MRF can be up to 3.5 times more efficient than steady-state methods, when spatial undersampling is ignored. If incoherent aliasing is treated as noise during least-squares parameter estimation, the efficiency is reduced in proportion to the SNR of the data, with reduction factors of 5 often seen for practical SNR levels.In vivovalidation showed a very good agreement between the theoretical and experimentally predicted efficiency. This work presents and validates an efficiency metric to optimize and compare the performance of qMRI methods. Transient methods were found to be intrinsically more efficient than steady-state methods, however the effect of spatial undersampling can significantly erode this advantage.

Lesion probability mapping in MS patients using a regression network on MR fingerprinting

BACKGROUND

To develop a regression neural network for the reconstruction of lesion probability maps on Magnetic Resonance Fingerprinting using echo-planar imaging (MRF-EPI) in addition to [Formula: see text], [Formula: see text], NAWM, and GM- probability maps.

METHODS

We performed MRF-EPI measurements in 42 patients with multiple sclerosis and 6 healthy volunteers along two sites. A U-net was trained to reconstruct the denoised and distortion corrected [Formula: see text] and [Formula: see text] maps, and to additionally generate NAWM-, GM-, and WM lesion probability maps.

RESULTS

WM lesions were predicted with a dice coefficient of [Formula: see text] and a lesion detection rate of [Formula: see text] for a threshold of 33%. The network jointly enabled accurate [Formula: see text] and [Formula: see text] times with relative deviations of 5.2% and 5.1% and average dice coefficients of [Formula: see text] and [Formula: see text] for NAWM and GM after binarizing with a threshold of 80%.

CONCLUSION

DL is a promising tool for the prediction of lesion probability maps in a fraction of time. These might be of clinical interest for the WM lesion analysis in MS patients.