Multi-Rate Acquisition for Dead Time Reduction in Magnetic Resonance Receivers: Application to Imaging With Zero Echo Time

A magnetic resonance imaging (MRI) exam typically consists of the acquisition of multiple MR pulse sequences, which are required for a reliable diagnosis. With the rise of generative deep learning models, approaches for the synthesis of MR images are developed to either synthesize additional MR contrasts, generate synthetic data, or augment existing data for AI training. While current generative approaches allow only the synthesis of specific sets of MR contrasts, we developed a method to generate synthetic MR images with adjustable image contrast. Therefore, we trained a generative adversarial network (GAN) with a separate auxiliary classifier (AC) network to generate synthetic MR knee images conditioned on various acquisition parameters (repetition time, echo time, and image orientation). The AC determined the repetition time with a mean absolute error (MAE) of 239.6 ms, the echo time with an MAE of 1.6 ms, and the image orientation with an accuracy of 100%. Therefore, it can properly condition the generator network during training. Moreover, in a visual Turing test, two experts mislabeled 40.5% of real and synthetic MR images, demonstrating that the image quality of the generated synthetic and real MR images is comparable. This work can support radiologists and technologists during the parameterization of MR sequences by previewing the yielded MR contrast, can serve as a valuable tool for radiology training, and can be used for customized data generation to support AI training.

Download Full-text

ID: 63: RAPID 3D PRECLINICAL QUANTITATIVE LUNG IMAGING WITH ULTRASHORT-ECHO TIME (UTE) MRI IN A MOUSE MODEL OF CYSTIC FIBROSIS LUNG DISEASE

Journal of Investigative Medicine ◽

10.1136/jim-2016-000120.133 ◽

2016 ◽

Vol 64 (4) ◽

pp. 975.1-975

Author(s):

C Anderson ◽

C Flask

Keyword(s):

Cystic Fibrosis ◽

Magnetic Resonance ◽

Mouse Model ◽

Lung Disease ◽

Lung Tissue ◽

Lung Imaging ◽

Early Disease ◽

Lung Anatomy ◽

Echo Time ◽

Regional Disease

Currently, the life expectancy for cystic fibrosis (CF) lung disease is less than 40 years due to decreasing lung function despite significant advances in the care and treatment of these patients. As patients live longer, the preservation of healthy lung tissue becomes of paramount importance to improve patient quality of life and increase life span. To do this, an understanding of the early disease processes is needed as is an ability to monitor the efficacy of therapeutic interventions early in life. CF lung disease, similar to other lung diseases, is a regional disease causing local dysfunction in the lung tissue and changes in lung anatomy. It is important for any monitoring or diagnostic tool to be sensitive to early regional disease which current methods (spirometry) are not. This lack of sensitivity to regional disease limits the ability of physicians and researchers to track the earliest stages of disease and assess treatment efficacy in these initial disease stages, ideally in infants and young children. Three dimensional imaging presents a unique solution to this problem by providing a non-invasive, volumetric investigation of the lung tissue. Computed tomography has long been the first choice in clinical lung imaging offering excellent resolution and fast imaging times but results in repeated exposure to ionizing radiation. Because the patient populations of interest are infants and children, avoidance of unnecessary, repeated radiation exposure during longitudinal monitoring is desirable. This combination of clinical and research need has led us to the exploration of rapid MRI techniques for lung imaging. We are interested in developing a novel, robust quantitative Magnetic Resonance Imaging technique that allows for 3D investigation of the lung tissue and is sensitive to early disease changes. Our hypothesis is that quantitative imaging will be able to detect changes in regional lung anatomy as an indication of early disease before disease is detected by standard methods. To accomplish this goal, we are proposing the implementation of multiple advanced quantitative MRI techniques including T1-mapping using Saturation-Recovery Look-Locker mapping and simultaneous multiple parameter mapping (combinations of T1, T2, T2*) using the recently developed Magnetic Resonance Fingerprinting method. An ultra-short echo time acquisition will be used to ensure imaging of the rapidly decaying MRI signal in the lung is possible. Using a radial acquisition, we plan to include an undersampled acquisition to reduce imaging time and generate an imaging method that is rapid and insensitive to patient motion. Our goal is to initially apply these quantitative measures in a mouse model of cystic fibrosis to establish the ability of the imaging methods to be sensitive to regional disease in CF mice. We expect to see changes in the quantitative parameters in areas that correspond to diseased areas of the lung upon histological investigation. These quantitative measurements should give unambiguous indications of disease and allow identification of changes in lung anatomy early in the disease process. This work will lay the foundation for translation of clinical CF monitoring in a pediatric population. Translational studies such as these will hopefully provide a measurement of disease progression and provide a new opportunity to evaluate early disease therapeutics offering insight into the earliest manifestations of CF lung disease.

Download Full-text