Michael Schacht Hansen

SNRAware, a model-agnostic approach for training MRI denoising models that leverages information from the image reconstruction process, improved performance and enhanced generalization to unseen imaging applications.

Purpose: Quantitative MRI techniques such as MR fingerprinting (MRF) promise more objective and comparable measurements of tissue properties at the point‐of‐care than weighted imaging. However, few direct cross‐modal comparisons of MRF's repeatability and reproducibility versus weighted acquisitions have been performed. This work proposes a novel fully automated pipeline for quantitatively comparing cross‐modal imaging performance in vivo via atlas‐based sampling.Methods: We acquire whole‐brain 3D‐MRF, turbo spin echo, and MPRAGE sequences three times each on two scanners across 10 subjects, for a total of 60 multimodal datasets. The proposed automated registration and analysis pipeline uses linear and nonlinear registration to align all qualitative and quantitative DICOM stacks to Montreal Neurological Institute (MNI) 152 space, then samples each dataset's native space through transformation inversion to compare performance within atlas regions across subjects, scanners, and repetitions.Results: Voxel values within MRF‐derived maps were found to be more repeatable (σ_T1 = 1.90, σ_T2 = 3.20) across sessions than vendor‐reconstructed MPRAGE (σ_T1w = 6.04) or turbo spin echo (σ_T2w = 5.66) images. Additionally, MRF was found to be more reproducible across scanners (σ_T1 = 2.21, σ_T2 = 3.89) than either qualitative modality (σ_T1w = 7.84, σ_T2w = 7.76). Notably, differences between repeatability and reproducibility of in vivo MRF were insignificant, unlike the weighted images.Conclusion: MRF data from many sessions and scanners can potentially be treated as a single dataset for harmonized analysis or longitudinal comparisons without the additional regularization steps needed for qualitative modalities.

Purpose: A new real‐time MR‐thermometry pipeline was developed to measure multiple temperature images per heartbeat with 1.6×1.6×3 mm³spatial resolution. The method was evaluated on 10 healthy volunteers and during radiofrequency ablation (RFA) in sheep.Methods: Multislice, electrocardiogram‐triggered, echo‐planar imaging was combined with parallel imaging, under free breathing conditions. In‐plane respiratory motion was corrected on magnitude images by an optical flow algorithm. Motion‐related susceptibility artifacts were compensated on phase images by an algorithm based on Principal Component Analysis. Correction of phase drift and temporal filter were included in the pipeline implemented in the Gadgetron framework. Contact electrograms were recorded simultaneously with MR thermometry by an MR‐compatible ablation catheter.Results: The temporal standard deviation of temperature in the left ventricle remained below 2 °C on each volunteer. In sheep, focal heated regions near the catheter tip were observed on temperature images (maximal temperature increase of 38 °C) during RFA, with contact electrograms of acceptable quality. Thermal lesion dimensions at gross pathology were in agreement with those observed on thermal dose images.Conclusion: This fully automated MR thermometry pipeline (five images/heartbeat) provides direct assessment of lesion formation in the heart during catheter‐based RFA, which may improve treatment of cardiac arrhythmia by ablation. Magn Reson Med 77:673–683, 2017. © 2016 International Society for Magnetic Resonance in Medicine

Background The value of late-gadolinium-enhancement (LGE) imaging in the diagnosis and management of pediatric and congenital heart disease is clear; however current acquisition techniques are susceptible to error and artifacts when performed in children because of children’s higher heart rates, higher prevalence of sinus arrhythmia, and inability to breath-hold. Commonly used techniques in pediatric LGE imaging include breath-held segmented FLASH (segFLASH) and steady-state free precession-based (segSSFP) imaging. More recently, single-shot SSFP techniques with respiratory motion-corrected averaging have emerged. Objective This study tested and compared single-shot free-breathing LGE techniques with standard segmented breath-held techniques in children undergoing LGE imaging. Materials and methods Thirty-two consecutive children underwent clinically indicated late-enhancement imaging using intravenous gadobutrol 0.15 mmol/kg. Breath-held segSSFP, breath-held segFLASH, and free-breathing single-shot SSFP LGE sequences were performed in consecutive series in each child. Two blinded reviewers evaluated the quality of the images and rated them on a scale of 1–5 (1 = poor, 5 = superior) based on blood pool–myocardial definition, presence of cardiac motion, presence of respiratory motion artifacts, and image acquisition artifact. We used analysis of variance (ANOVA) to compare groups. Results Patients ranged in age from 9 months to 18 years, with a mean +/− standard deviation (SD) of 13.3 +/− 4.8 years. R-R interval at the time of acquisition ranged 366–1,265 milliseconds (ms) (47–164 beats per minute [bpm]), mean +/− SD of 843+/−231 ms (72+/−21 bpm). Mean +/− SD quality ratings for long-axis imaging for segFLASH, segSSFP and single-shot SSFP were 3.1+/−0.9, 3.4+/−0.9 and 4.0+/−0.9, respectively (P < 0.01 by ANOVA). Mean +/− SD quality ratings for short-axis imaging for segFLASH, segSSFP and single-shot SSFP were 3.4+/−1, 3.8+/−0.9 and 4.3+/−0.7, respectively (P < 0.01 by ANOVA). Conclusion Single-shot late-enhancement imaging with motion-corrected averaging is feasible in children, robust at high heart rates and with variable R-R intervals, and can be performed without breath-holding with higher image quality ratings than standard breath-held techniques. Use of free-breathing single-shot motion-corrected technique does not compromise LGE image quality in children who can hold their breath, and it can significantly improve image quality in children who cannot hold their breath or who have significant arrhythmia.

Purpose: This work proposes the ISMRM Raw Data format as a common MR raw data format, which promotes algorithm and data sharing.Methods: A file format consisting of a flexible header and tagged frames of k‐space data was designed. Application Programming Interfaces were implemented in C/C++, MATLAB, and Python. Converters for Bruker, General Electric, Philips, and Siemens proprietary file formats were implemented in C++. Raw data were collected using magnetic resonance imaging scanners from four vendors, converted to ISMRM Raw Data format, and reconstructed using software implemented in three programming languages (C++, MATLAB, Python).Results: Images were obtained by reconstructing the raw data from all vendors. The source code, raw data, and images comprising this work are shared online, serving as an example of an image reconstruction project following a paradigm of reproducible research.Conclusion: The proposed raw data format solves a practical problem for the magnetic resonance imaging community. It may serve as a foundation for reproducible research and collaborations. The ISMRM Raw Data format is a completely open and community‐driven format, and the scientific community is invited (including commercial vendors) to participate either as users or developers. Magn Reson Med 77:411–421, 2017.