Search

Fitting rate constants to Hyperpolarized [1- 13 C]Pyruvate (HP C13) MRI data is a promising approach for quantifying metabolism in vivo. Current methods typically fit each voxel of the dataset using a least-squares objective. With these methods, each voxel is considered independently, and the spatial relationships are not considered during fitting. In this work, we use a convolutional neural network, a U-Net, with convolutions across the 2D spatial dimensions to estimate pyruvate-to-lactate conversion rate, k PL , maps from dynamic HP C13 datasets. We designed a framework for creating simulated anatomically accurate brain data that matches typical HP C13 characteristics to provide large amounts of data for training with ground truth results. The U-Net is initially trained with the digital phantom data and then further trained with in vivo datasets for regularization. In simulation where ground-truth k PL maps are available, the U-Net outperforms voxel-wise fitting with and without spatiotemporal denoising, particularly for low SNR data. In vivo data was evaluated qualitatively, as no ground truth is available, and before regularization the U-Net predicted k PL maps appear oversmoothed. After further training with in vivo data, the resulting k PL maps appear more realistic. This study demonstrates how to use a U-Net to estimate rate constant maps for HP C13 data, including a comprehensive framework for generating a large amount of anatomically realistic simulated data and an approach for regularization. This simulation and architecture provide a foundation that can be built upon in the future for improved performance., Competing Interests: Declaration of competing interest The authors declare the following financial interests/personal relationships which may be considered as potential competing interests: Peder Larson reports a relationship with GE Healthcare that includes: funding grants. If there are other authors, they declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper., (Copyright © 2025 Elsevier Inc. All rights reserved.)

Magnetic resonance imaging is subject to slow acquisition times due to the inherent limitations in data sampling. Recently, supervised deep learning has emerged as a promising technique for reconstructing sub-sampled MRI. However, supervised deep learning requires a large dataset of fully-sampled data. Although unsupervised or self-supervised deep learning methods have emerged to address the limitations of supervised deep learning approaches, they still require a database of images. In contrast, scan-specific deep learning methods learn and reconstruct using only the sub-sampled data from a single scan. Here, we introduce Scan-Specific Self-Supervised Bayesian Deep Non-Linear Inversion (DNLINV) that does not require an auto calibration scan region. DNLINV utilizes a Deep Image Prior-type generative modeling approach and relies on approximate Bayesian inference to regularize the deep convolutional neural network. We demonstrate our approach on several anatomies, contrasts, and sampling patterns and show improved performance over existing approaches in scan-specific calibrationless parallel imaging and compressed sensing.

A major remaining challenge for magnetic resonance-based attenuation correction methods (MRAC) is their susceptibility to sources of magnetic resonance imaging (MRI) artifacts (e.g., implants and motion) and uncertainties due to the limitations of MRI contrast (e.g., accurate bone delineation and density, and separation of air/bone). We propose using a Bayesian deep convolutional neural network that in addition to generating an initial pseudo-CT from MR data, it also produces uncertainty estimates of the pseudo-CT to quantify the limitations of the MR data. These outputs are combined with the maximum-likelihood estimation of activity and attenuation (MLAA) reconstruction that uses the PET emission data to improve the attenuation maps. With the proposed approach uncertainty estimation and pseudo-CT prior for robust MLAA (UpCT-MLAA), we demonstrate accurate estimation of PET uptake in pelvic lesions and show recovery of metal implants. In patients without implants, UpCT-MLAA had acceptable but slightly higher root-mean-squared-error (RMSE) than Zero-echotime and Dixon Deep pseudo-CT when compared to CTAC. In patients with metal implants, MLAA recovered the metal implant; however, anatomy outside the implant region was obscured by noise and crosstalk artifacts. Attenuation coefficients from the pseudo-CT from Dixon MRI were accurate in normal anatomy; however, the metal implant region was estimated to have attenuation coefficients of air. UpCT-MLAA estimated attenuation coefficients of metal implants alongside accurate anatomic depiction outside of implant regions.

As in conventional 1 H MRI, T 1 and T 2 relaxation times of hyperpolarized (HP) 13 C nuclei can provide important biomedical information. Two new approaches were developed for simultaneous T 1 and T 2 mapping of HP 13 C probes based on balanced steady state free precession (bSSFP) acquisitions: a method based on sequential T 1 and T 2 mapping modules, and a model-based joint T 1 /T 2 approach analogous to MR fingerprinting. These new methods were tested in simulations, HP 13 C phantoms, and in vivo in normal Sprague-Dawley rats. Non-localized T 1 values, low flip angle EPI T 1 maps, bSSFP T 2 maps, and Bloch-Siegert B 1 maps were also acquired for comparison. T 1 and T 2 maps acquired using both approaches were in good agreement with both literature values and data from comparative acquisitions. Multiple HP 13 C compounds were successfully mapped, with their relaxation time parameters measured within heart, liver, kidneys, and vasculature in one acquisition for the first time., Competing Interests: Declaration of Competing Interest The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper., (Copyright © 2020 Elsevier Inc. All rights reserved.)

Purpose: This study demonstrated a magnetic resonance (MR) signal multitask learning method for three-dimensional (3D) simultaneous segmentation and relaxometry of human brain tissues., Materials and Methods: A 3D inversion-prepared balanced steady-state free precession sequence was used for acquiring in vivo multicontrast brain images. The deep neural network contained three residual blocks, and each block had 8 fully connected layers with sigmoid activation, layer norm, and 256 neurons in each layer. Online-synthesized MR signal evolutions and labels were used to train the neural network batch-by-batch. Empirically defined ranges of T1 and T2 values for the normal gray matter, white matter, and cerebrospinal fluid (CSF) were used as the prior knowledge. MRI brain experiments were performed on three healthy volunteers. The mean and standard deviation for the T1 and T2 values in vivo were reported and compared to literature values. Additional animal (N = 6) and prostate patient (N = 1) experiments were performed to compare the estimated T1 and T2 values with those from gold standard methods and to demonstrate clinical applications of the proposed method., Results: In animal validation experiment, the differences/errors (mean difference ± standard deviation of difference) between the T1 and T2 values estimated from the proposed method and the ground truth were 113 ± 486 and 154 ± 512 ms for T1, and 5 ± 33 and 7 ± 41 ms for T2, respectively. In healthy volunteer experiments (N = 3), whole brain segmentation and relaxometry were finished within ~ 5 s. The estimated apparent T1 and T2 maps were in accordance with known brain anatomy, and not affected by coil sensitivity variation. Gray matter, white matter, and CSF were successfully segmented. The deep neural network can also generate synthetic T1- and T2-weighted images., Conclusion: The proposed multitask learning method can directly generate brain apparent T1 and T2 maps, as well as synthetic T1- and T2-weighted images, in conjunction with segmentation of gray matter, white matter, and CSF., (© 2019 American Association of Physicists in Medicine.)

Accurate quantification of uptake on PET images depends on accurate attenuation correction in reconstruction. Current MR-based attenuation correction methods for body PET use a fat and water map derived from a 2-echo Dixon MRI sequence in which bone is neglected. Ultrashort-echo-time or zero-echo-time (ZTE) pulse sequences can capture bone information. We propose the use of patient-specific multiparametric MRI consisting of Dixon MRI and proton-density-weighted ZTE MRI to directly synthesize pseudo-CT images with a deep learning model: we call this method ZTE and Dixon deep pseudo-CT (ZeDD CT). Methods: Twenty-six patients were scanned using an integrated 3-T time-of-flight PET/MRI system. Helical CT images of the patients were acquired separately. A deep convolutional neural network was trained to transform ZTE and Dixon MR images into pseudo-CT images. Ten patients were used for model training, and 16 patients were used for evaluation. Bone and soft-tissue lesions were identified, and the SUV max was measured. The root-mean-squared error (RMSE) was used to compare the MR-based attenuation correction with the ground-truth CT attenuation correction. Results: In total, 30 bone lesions and 60 soft-tissue lesions were evaluated. The RMSE in PET quantification was reduced by a factor of 4 for bone lesions (10.24% for Dixon PET and 2.68% for ZeDD PET) and by a factor of 1.5 for soft-tissue lesions (6.24% for Dixon PET and 4.07% for ZeDD PET). Conclusion: ZeDD CT produces natural-looking and quantitatively accurate pseudo-CT images and reduces error in pelvic PET/MRI attenuation correction compared with standard methods., (© 2018 by the Society of Nuclear Medicine and Molecular Imaging.)

Simultaneous positron emission tomography and MRI (PET/MRI) is a technology that combines the anatomic and quantitative strengths of MR imaging with physiologic information obtained from PET. PET and computed tomography (PET/CT) performed in a single scanning session is an established technology already in widespread and accepted use worldwide. Given the higher cost and complexity of operating and interpreting the studies obtained on a PET/MRI system, there has been question as to which patients would benefit most from imaging with PET/MRI versus PET/CT. In this article, we compare PET/MRI with PET/CT, detail the applications for which PET/MRI has shown promise and discuss impediments to future adoption. It is our hope that future work will prove the benefit of PET/MRI to specific groups of patients, initially those in which PET/CT and MRI are already performed, leveraging simultaneity and allowing for greater degrees of multiparametric evaluation., Level of Evidence: 5 Technical Efficacy: Stage 5 J. Magn. Reson. Imaging 2017;46:1247-1262., (© 2017 International Society for Magnetic Resonance in Medicine.)

Purpose: This study introduces a new hybrid ZTE/Dixon MR-based attenuation correction (MRAC) method including bone density estimation for PET/MRI and quantifies the effects of bone attenuation on metastatic lesion uptake in the pelvis., Methods: Six patients with pelvic lesions were scanned using fluorodeoxyglucose (18F-FDG) in an integrated time-of-flight (TOF) PET/MRI system. For PET attenuation correction, MR imaging consisted of two-point Dixon and zero echo-time (ZTE) pulse sequences. A continuous-value fat and water pseudoCT was generated from a two-point Dixon MRI. Bone was segmented from the ZTE images and converted to Hounsfield units (HU) using a continuous two-segment piecewise linear model based on ZTE MRI intensity. The HU values were converted to linear attenuation coefficients (LAC) using a bilinear model. The bone voxels of the Dixon-based pseudoCT were replaced by the ZTE-derived bone to produce the hybrid ZTE/Dixon pseudoCT. The three different AC maps (Dixon, hybrid ZTE/Dixon, CTAC) were used to reconstruct PET images using a TOF-ordered subset expectation maximization algorithm with a point-spread function model. Metastatic lesions were separated into two classes, bone lesions and soft tissue lesions, and analyzed. The MRAC methods were compared using a root-mean-squared error (RMSE), where the registered CTAC was taken as ground truth., Results: The RMSE of the maximum standardized uptake values (SUV max ) is 11.02% and 7.79% for bone (N = 6) and soft tissue lesions (N = 8), respectively, using Dixon MRAC. The RMSE of SUVmax for these lesions is significantly reduced to 3.28% and 3.94% when using the new hybrid ZTE/Dixon MRAC. Additionally, the RMSE for PET SUVs across the entire pelvis and all patients are 8.76% and 4.18%, for the Dixon and hybrid ZTE/Dixon MRAC methods, respectively., Conclusion: A hybrid ZTE/Dixon MRAC method was developed and applied to pelvic regions in an integrated TOF PET/MRI, demonstrating improved MRAC. This new method included bone density estimation, through which PET quantification is improved., (© 2017 American Association of Physicists in Medicine.)