Magnetic resonance image synthesis using a flexible model

Optimization of T2-weighted imaging for shoulder magnetic resonance arthrography by synthetic magnetic resonance imaging

Background Synthetic magnetic resonance imaging (MRI) allows reformatting of various synthetic images by adjustment of scanning parameters such as repetition time (TR) and echo time (TE). Optimized MR images can be reformatted from T1, T2, and proton density (PD) values to achieve maximum tissue contrast between joint fluid and adjacent soft tissue. Purpose To demonstrate the method for optimization of TR and TE by synthetic MRI and to validate the optimized images by comparison with conventional shoulder MR arthrography (MRA) images. Material and Methods Thirty-seven shoulder MRA images acquired by synthetic MRI were retrospectively evaluated for PD, T1, and T2 values at the joint fluid and glenoid labrum. Differences in signal intensity between the fluid and labrum were observed between TR of 500-6000 ms and TE of 80-300 ms in T2-weighted (T2W) images. Conventional T2W and synthetic images were analyzed for diagnostic agreement of supraspinatus tendon abnormalities (kappa statistics) and image quality scores (one-way analysis of variance with post-hoc analysis). Results Optimized mean values of TR and TE were 2724.7 ± 1634.7 and 80.1 ± 0.4, respectively. Diagnostic agreement for supraspinatus tendon abnormalities between conventional and synthetic MR images was excellent (κ = 0.882). The mean image quality score of the joint space in optimized synthetic images was significantly higher compared with those in conventional and synthetic images (2.861 ± 0.351 vs. 2.556 ± 0.607 vs. 2.750 ± 0.439; P < 0.05). Conclusion Synthetic MRI with optimized TR and TE for shoulder MRA enables optimization of soft-tissue contrast.

Quantitative T2 Mapping of Knee Cartilage: Comparison between the Synthetic MR Imaging and the CPMG Sequence

The purpose was to evaluate the feasibility of quantitative MRI T₂ mapping based on the quantitative MRI (QRAPMASTER) sequence for the quantitative assessment of knee cartilage. The T₂ values from the phantom study showed excellent correlation between the two techniques (r² = 0.998). The cartilage T₂ values exhibited strong correlations (r² = 0.867–0.982). Quantitative MRI (qMRI) T₂ mapping can be used as an alternative to multi-echo T₂ mapping, with relatively short scan time.

Functional Brain Imaging Synthesis Based on Image Decomposition and Kernel Modeling: Application to Neurodegenerative Diseases

The rise of neuroimaging in research and clinical practice, together with the development of new machine learning techniques has strongly encouraged the Computer Aided Diagnosis (CAD) of different diseases and disorders. However, these algorithms are often tested in proprietary datasets to which the access is limited and, therefore, a direct comparison between CAD procedures is not possible. Furthermore, the sample size is often small for developing accurate machine learning methods. Multi-center initiatives are currently a very useful, although limited, tool in the recruitment of large populations and standardization of CAD evaluation. Conversely, we propose a brain image synthesis procedure intended to generate a new image set that share characteristics with an original one. Our system focuses on nuclear imaging modalities such as PET or SPECT brain images. We analyze the dataset by applying PCA to the original dataset, and then model the distribution of samples in the projected eigenbrain space using a Probability Density Function (PDF) estimator. Once the model has been built, we can generate new coordinates on the eigenbrain space belonging to the same class, which can be then projected back to the image space. The system has been evaluated on different functional neuroimaging datasets assessing the: resemblance of the synthetic images with the original ones, the differences between them, their generalization ability and the independence of the synthetic dataset with respect to the original. The synthetic images maintain the differences between groups found at the original dataset, with no significant differences when comparing them to real-world samples. Furthermore, they featured a similar performance and generalization capability to that of the original dataset. These results prove that these images are suitable for standardizing the evaluation of CAD pipelines, and providing data augmentation in machine learning systems -e.g. in deep learning-, or even to train future professionals at medical school.

High dynamic range processing for magnetic resonance imaging

Purpose

To minimize feature loss in T₁- and T₂-weighted MRI by merging multiple MR images acquired at different T_R and T_E to generate an image with increased dynamic range.

Materials and Methods

High Dynamic Range (HDR) processing techniques from the field of photography were applied to a series of acquired MR images. Specifically, a method to parameterize the algorithm for MRI data was developed and tested. T₁- and T₂-weighted images of a number of contrast agent phantoms and a live mouse were acquired with varying T_R and T_E parameters. The images were computationally merged to produce HDR-MR images. All acquisitions were performed on a 7.05 T Bruker PharmaScan with a multi-echo spin echo pulse sequence.

Results

HDR-MRI delineated bright and dark features that were either saturated or indistinguishable from background in standard T₁- and T₂-weighted MRI. The increased dynamic range preserved intensity gradation over a larger range of T₁ and T₂ in phantoms and revealed more anatomical features in vivo.

Conclusions

We have developed and tested a method to apply HDR processing to MR images. The increased dynamic range of HDR-MR images as compared to standard T₁- and T₂-weighted images minimizes feature loss caused by magnetization recovery or low SNR.

Rapid T1 quantification based on 3D phase sensitive inversion recovery

Background

In Contrast Enhanced Magnetic Resonance Imaging fibrotic myocardium can be distinguished from healthy tissue using the difference in the longitudinal T₁ relaxation after administration of Gadolinium, the so-called Late Gd Enhancement. The purpose of this work was to measure the myocardial absolute T₁ post-Gd from a single breath-hold 3D Phase Sensitivity Inversion Recovery sequence (PSIR). Equations were derived to take the acquisition and saturation effects on the magnetization into account.

Methods

The accuracy of the method was investigated on phantoms and using simulations. The method was applied to a group of patients with suspected myocardial infarction where the absolute difference in relaxation of healthy and fibrotic myocardium was measured at about 15 minutes post-contrast. The evolution of the absolute R₁ relaxation rate (1/T₁) over time after contrast injection was followed for one patient and compared to T₁ mapping using Look-Locker. Based on the T₁ maps synthetic LGE images were reconstructed and compared to the conventional LGE images.

Results

The fitting algorithm is robust against variation in acquisition flip angle, the inversion delay time and cardiac arrhythmia. The observed relaxation rate of the myocardium is 1.2 s^-1, increasing to 6 - 7 s^-1 after contrast injection and decreasing to 2 - 2.5 s^-1 for healthy myocardium and to 3.5 - 4 s^-1 for fibrotic myocardium. Synthesized images based on the T₁ maps correspond very well to actual LGE images.

Conclusions

The method provides a robust quantification of post-Gd T₁ relaxation for a complete cardiac volume within a single breath-hold.

Contrast-modified gradient echo imaging using rotary echo preparatory pulses

The use of on-resonance 121 binomial composite pulses in two- or three-dimensional magnetization-prepared gradient-recalled echo magnetic resonance imaging experiments generates rotary echoes, leading to an increase in contrast range that is, in part, determined by the ratio of T2 to T1. In comparison with other fast gradient-recalled echo imaging techniques designed for enhanced T2 contrast, this method is more robust with respect to radiofrequency field inhomogeneity and less sensitive with respect to motion artifacts. Three-dimensional parametric images may be calculated using least-squares fitting based on a simple model for steady-state longitudinal magnetization during the imaging sequences.