Skip to main content
eScholarship
Open Access Publications from the University of California

Scholarly Works (12 results)

  • Thesis
  • Peer Reviewed
UCLA Electronic Theses and Dissertations (2017)

Chemical exchange saturation transfer (CEST) is an emerging MR imaging technique that is sensitive to the metabolite accumulation/loss. This technique selectively saturates the exchangeable protons in certain molecules. After exchange with water protons, the saturation can be detected by the change of water signal. The primary focus of the work in this dissertation is to make the current CEST techniques clinically translatable and address the challenges for cardiac and lumbar spine applications.

CEST imaging has been applied in the heart for creatine mapping because cardiac dysfunction was linked to loss of metabolites in the creatine kinase system. Validation studies were performed in chronic myocardial infarction animal model using late gadolinium enhancement as reference. It is shown that the infarct region has lower CEST signal compared to remote myocardium. Spatially, the hypointense regions in the CEST contrast maps closely match the bright areas in the LGE images.

Intervertebral disc (IVD) degeneration is one of the leading causes of chronic low back pain. Previous studies associated low pH as the leading cause of discogenic low back pain in degenerate IVDs. We proposed a quantitative CEST (qCEST) imaging protocol for pH assessment in the IVD. It is shown in animal studies that the exchange rate measured from qCEST technique is correlated with the direct pH measurement using tissue pH-meter. In addition, gene analysis of harvested degenerated IVDs revealed strong positive correlation between up-regulation of pain markers and increase in qCEST signal. Collectively, these findings demonstrate that this approach can be used to measure pH in vivo within the IVD and has the potential to be used as a novel non-invasive method for the diagnosis of discogenic pain.

CEST fingerprinting framework was proposed to further improve CEST quantification. This is inspired by the magnetic resonance fingerprinting concept where the signal generated using randomized sequence parameters will be matched directly to a pre-defined dictionary instead of using a repeated, serial acquisition of data to fit to a particular equation. CEST fingerprinting utilizes CEST saturation with varying saturation power B1 amplitude and saturation time to create uniqueness of signal evolution for different exchange rates. Phantom studies demonstrated that CEST fingerprinting was more efficient (5x faster) compared to pulsed qCEST because there is no need for long saturation time and long TR. It is also shown that the proposed CEST fingerprinting technique can quantify exchange rate more accurately in the presence of MT effects.

In summary, with the technical improvements proposed in this dissertation, CEST imaging can be performed more efficiently and more accurately. These techniques can be potentially translated into clinical studies.

    Cover page: Chemical Exchange Saturation Transfer (CEST): A Metabolic Imaging Technique Using Magnetic Resonance Imaging (MRI)
    • Article
    • Peer Reviewed
    UCLA Previously Published Works (2018)

    Purpose

    There is an increased interest to determine the exchange rate using CEST to provide pH information. However, current CEST quantification methods require lengthy scan times and do not address magnetization transfer effects. The purpose of this work was to apply the magnetic resonance fingerprinting (MRF) concept to CEST to achieve more efficient and accurate exchange rate quantification.

    Methods

    The proposed CEST fingerprinting method used varying saturation powers and saturation times to create unique signal evolutions for different exchange rates. The acquired signal was matched to a predefined dictionary to determine the exchange rate. The magnetization transfer effects were also addressed in the framework of CEST fingerprinting: The simulated dictionary could predict the signal curves without magnetization transfer effects, and comparing the dictionary to the acquired signals allowed the correction of the magnetization transfer effects. The CEST fingerprinting method was compared with the conventional pulsed quantitative CEST method using omega plots in the creatine phantom study.

    Results

    The CEST fingerprinting method has a significantly reduced scan time (10 minutes versus 50 minutes) while providing more accurate exchange rate quantification using literature values as the reference.

    Conclusion

    In this study, we demonstrate that CEST fingerprinting is more efficient (5 times faster) compared with pulsed quantitative CEST. It is also shown that the results of the proposed CEST fingerprinting technique are much closer to the literature values than pulsed quantitative CEST at 3 T.
      • Article
      • Peer Reviewed
      UCLA Previously Published Works (2018)
      Magnetic resonance image (MRI) in high spatial resolution provides detailed anatomical information and is often necessary for accurate quantitative analysis. However, high spatial resolution typically comes at the expense of longer scan time, less spatial coverage, and lower signal to noise ratio (SNR). Single Image Super-Resolution (SISR), a technique aimed to restore high-resolution (HR) details from one single low-resolution (LR) input image, has been improved dramatically by recent breakthroughs in deep learning. In this paper, we introduce a new neural network architecture, 3D Densely Connected Super-Resolution Networks (DCSRN) to restore HR features of structural brain MR images. Through experiments on a dataset with 1,113 subjects, we demonstrate that our network outperforms bicubic interpolation as well as other deep learning methods in restoring 4× resolution-reduced images.
        • Article
        • Peer Reviewed
        UC Irvine Previously Published Works (2018)
        Objective: To quantitatively evaluate lumbar disc degeneration with recently-developed quantitative magnetic resonance imaging (MRI) techniques. A series of MRI parameters, including T2*, T1rho relaxation time, apparent diffusion coefficient and gagCEST, were compared and correlated with the Pfirrmann semi-quantitative classification of lumbar intervertebral disc degeneration; the most accurate and relevant MRI parameters of lumbar disc degeneration were identified. Materials and Methods: Thirty-seven subjects (age range, 18-74 years) with non-specific low back pain (LBP) for more than 6 months were enrolled. The L1/2-L5/S1 discs of each subject were measured and then analyzed. Results: The gagCEST value of the discs showed the best negative correlation with degeneration level (nucleus pulposus: r = -0.951, P < 0.001; annulus fibrosus: r = -0.938, P < 0.001). The discriminant analysis results showed that this parameter also had the highest correct rate using a single index (gagCEST discriminant accuracy = 82%). Conclusion: Early stage lumbar disc degeneration can be quantitatively evaluated with MRI using the chemical exchange saturation transfer technique.
          Cover page: Multi-parameter evaluation of lumbar intervertebral disc degeneration using quantitative magnetic resonance imaging techniques.Creative Commons 'BY' version 4.0 license
          • Article
          UCLA Previously Published Works (2020)
          High-resolution (HR) magnetic resonance imaging (MRI) provides detailed anatomical information that is critical for diagnosis in the clinical application. However, HR MRI typically comes at the cost of long scan time, small spatial coverage, and low signal-to-noise ratio (SNR). Recent studies showed that with a deep convolutional neural network (CNN), HR generic images could be recovered from low-resolution (LR) inputs via single image super-resolution (SISR) approaches. Additionally, previous works have shown that a deep 3D CNN can generate high-quality SR MRIs by using learned image priors. However, 3D CNN with deep structures, have a large number of parameters and are computationally expensive. In this paper, we propose a novel 3D CNN architecture, namely a multi-level densely connected super-resolution network (mDCSRN), which is light-weight, fast and accurate. We also show that with the generative adversarial network (GAN)-guided training, the mDCSRN-GAN provides appealing sharp SR images with rich texture details that are highly comparable with the referenced HR images. Our results from experiments on a large public dataset with 1,113 subjects showed that this new architecture outperformed other popular deep learning methods in recovering 4x resolution-downgraded images in both quality and speed.
            Cover page: MRI Super-Resolution with GAN and 3D Multi-Level DenseNet: Smaller, Faster, and Better
            • Article
            • Peer Reviewed
            UCLA Previously Published Works (2019)

            Purpose

            To evaluate the accuracy and repeatability of a free-breathing, non-electrocardiogram (ECG), continuous myocardial T1 and extracellular volume (ECV) mapping technique adapted from the Multitasking framework.

            Methods

            The Multitasking framework is adapted to quantify both myocardial native T1 and ECV with a free-breathing, non-ECG, continuous acquisition T1 mapping method. We acquire interleaved high-spatial resolution image data and high-temporal resolution auxiliary data following inversion-recovery pulses at set intervals and perform low-rank tensor imaging to reconstruct images at 344 inversion times, 20 cardiac phases, and 6 respiratory phases. The accuracy and repeatability of Multitasking T1 mapping in generating native T1 and ECV maps are compared with conventional techniques in a phantom, a simulation, 12 healthy subjects, and 10 acute myocardial infarction patients.

            Results

            In phantoms, Multitasking T1 mapping correlated strongly with the gold-standard spin-echo inversion recovery (R2 = 0.99). A simulation study demonstrated that Multitasking T1 mapping has similar myocardial sharpness to the fully sampled ground truth. In vivo native T1 and ECV values from Multitasking T1 mapping agree well with conventional MOLLI values and show good repeatability for native T1 and ECV mapping for 60 seconds, 30 seconds, or 15 seconds of data. Multitasking native T1 and ECV in myocardial infarction patients correlate positively with values from MOLLI.

            Conclusion

            Multitasking T1 mapping can quantify native T1 and ECV in the myocardium with free-breathing, non-ECG, continuous scans with good image quality and good repeatability in vivo in healthy subjects, and correlation with MOLLI T1 and ECV in acute myocardial infarction patients.
              • Article
              • Peer Reviewed
              UCLA Previously Published Works (2018)
              High-resolution (HR) magnetic resonance images (MRI) provide detailed anatomical information important for clinical application and quantitative image analysis. However, HR MRI conventionally comes at the cost of longer scan time, smaller spatial coverage, and lower signal-to-noise ratio (SNR). Recent studies have shown that single image super-resolution (SISR), a technique to recover HR details from one single low-resolution (LR) input image, could provide high quality image details with the help of advanced deep convolutional neural networks (CNN). However, deep neural networks consume memory heavily and run slowly, especially in 3D settings. In this paper, we propose a novel 3D neural network design, namely a multi-level densely connected super-resolution network (mDCSRN) with generative adversarial network (GAN)–guided training. The mDCSRN trains and inferences quickly, and the GAN promotes realistic output hardly distinguishable from original HR images. Our results from experiments on a dataset with 1,113 subjects shows that our new architecture outperforms other popular deep learning methods in recovering 4x resolution-downgraded images and runs 6x faster.
                • Article
                • Peer Reviewed
                UCLA Previously Published Works (2017)
                BACKGROUND:Previous studies have linked cardiac dysfunction to loss of metabolites in the creatine kinase system. Chemical exchange saturation transfer (CEST) is a promising metabolic cardiovascular magnetic resonance (CMR) imaging technique and has been applied in the heart for creatine mapping. However, current limitations include: (a) long scan time, (b) residual cardiac and respiratory motion, and (c) B0 field variations induced by respiratory motion. An improved CEST CMR technique was developed to address these problems. METHODS:Animals with chronic myocardial infarction (N = 15) were scanned using the proposed CEST CMR technique and a late gadolinium enhancement (LGE)  sequence as reference. The major improvements of the CEST CMR technique are: (a) Images were acquired by single-shot FLASH, significantly increasing the scan efficiency. (b) All images were registered to reduce the residual motion. (c) The acquired Z-spectrum was analyzed using 3-pool-model Lorentzian-line fitting to generate CEST signal, reducing the impact of B0 field shifting due to respiratory motion. Feasibility of the technique was tested in a porcine model with chronic myocardial infarction. CEST signal was measured in the scar, border zone and remote myocardium. Initial studies were performed in one patient. RESULTS:In all animals, healthy remote myocardial CEST signal was elevated (0.16 ± 0.02) compared to infarct CEST signal (0.09 ± 0.02, P < 0.001) and the border zone (0.12 ± 0.02, P < 0.001). For both animal and patient studies, the hypointense regions in the CEST contrast maps closely match the bright areas in the LGE images. CONCLUSIONS:The proposed CEST CMR technique was developed to address long scan times, respiratory and cardiac motion, and B0 field variations. Lower CEST signal in bright region of the LGE image is consistent with the fact that myocardial infarction has reduced metabolic activity.
                  Cover page: Optimized CEST cardiovascular magnetic resonance for assessment of metabolic activity in the heart
                  • Article
                  • Peer Reviewed
                  UCLA Previously Published Works (2017)

                  Purpose

                  The presence of subendocardial dark-rim artifact (DRA) remains an ongoing challenge in first-pass perfusion (FPP) cardiac magnetic resonance imaging (MRI). We propose a free-breathing FPP imaging scheme with Cartesian sampling that is optimized to minimize the DRA and readily enables near-instantaneous image reconstruction.

                  Materials and methods

                  The proposed FPP method suppresses Gibbs ringing effects-a major underlying factor for the DRA-by "shaping" the underlying point spread function through a two-step process: 1) an undersampled Cartesian sampling scheme that widens the k-space coverage compared to the conventional scheme; and 2) a modified parallel-imaging scheme that incorporates optimized apodization (k-space data filtering) to suppress Gibbs-ringing effects. Healthy volunteer studies (n = 10) were performed to compare the proposed method against the conventional Cartesian technique-both using a saturation-recovery gradient-echo sequence at 3T. Furthermore, FPP imaging studies using the proposed method were performed in infarcted canines (n = 3), and in two symptomatic patients with suspected coronary microvascular dysfunction for assessment of myocardial hypoperfusion.

                  Results

                  Width of the DRA and the number of DRA-affected myocardial segments were significantly reduced in the proposed method compared to the conventional approach (width: 1.3 vs. 2.9 mm, P < 0.001; number of segments: 2.6 vs. 8.7; P < 0.0001). The number of slices with severe DRA was markedly lower for the proposed method (by 10-fold). The reader-assigned image quality scores were similar (P = 0.2), although the quantified myocardial signal-to-noise ratio was lower for the proposed method (P < 0.05). Animal studies showed that the proposed method can detect subendocardial perfusion defects and patient results were consistent with the gold-standard invasive test.

                  Conclusion

                  The proposed free-breathing Cartesian FPP imaging method significantly reduces the prevalence of severe DRAs compared to the conventional approach while maintaining similar resolution and image quality.

                  Level of evidence

                  2 J. Magn. Reson. Imaging 2017;45:542-555.
                    Cover page: First‐pass myocardial perfusion MRI with reduced subendocardial dark‐rim artifact using optimized Cartesian sampling
                    • Article
                    • Peer Reviewed
                    UCLA Previously Published Works (2022)

                    Purpose

                    To perform fast 3D steady-state CEST (ss-CEST) imaging using MR Multitasking.

                    Methods

                    A continuous acquisition sequence with repetitive ss-CEST modules was developed. Each ss-CEST module contains a single-lobe Gaussian saturation pulse, followed by a spoiler gradient and eight FLASH readouts (one "training line" + seven "imaging lines"). Three-dimensional Cartesian encoding was used for k-space acquisition. Reconstructed CEST images were quantified with four-pool Lorentzian fitting.

                    Results

                    Steady-state CEST with whole-brain coverage was performed in 5.6 s per saturation frequency offset at the spatial resolution of 1.7 × 1.7 × 3.0 mm3 . The total scan time was 5.5 min for 55 different frequency offsets. Quantitative CEST maps from multipool fitting showed consistent image quality across the volume.

                    Conclusion

                    Three-dimensional ss-CEST with whole-brain coverage can be done at 3 T within 5.5 min using MR Multitasking.
                      Cover page: Whole‐brain steady‐state CEST at 3 T using MR Multitasking
                      Top