Reconstruction Of Magnetic Esonance Imaging From Partial 2D K-space

Magnetic Resonance Imaging(MRI) is a new medical imaging method emerging in the recent years.It makes images based on nuclear magnetic resonance signals in the static magnetic field.It has many features such as high image resolution, multi-imaging parameters, non-invasive and arbitrary slice imaging, etc. Now MRI is one of the most advanced medical imaging methods, and is more and more widely used in the clinical applications.However, normal magnetic resonance imaging takes too long times, leading to high cost. Considering that the movement of the patient makes the image blur and contrast distortion, MRI cannot be used to research movement organs and nerve system before. Fast magnetic resonance imaging become the research hotspot in recent. Clinical applications in order to shorten imaging time, a common strategy is to reduce the amount of scan data collected, that is partial k-space method. This method can short imaging time without changing the hardware of MRI and the scan way. It is a fast imaging way with low cost. Nowadays, MRI from 1D partial k-space which has missing spectrum data in phase-encoding or frequency encoding is commonly used in clinical application. This paper presents the concept of MRI reconstruction from 2D partial k-space, which has missing spectrum data in both phase-encoding and frequency encoding. This strategy can further accelerate the imaging process.However, reduction of the number of phase-encoded signals results in the truncation artifacts when the truncated k-space data is Fourier transformed to reconstruct the image. The truncation artifact has long been a hindrance to the image reconstruction, study of how to eliminate truncation artifacts is a more urgent. To date, scholars have proposed many methods to reduce truncation artifacts, including the filtering method, signal analysis method, phase correction method, singular spectral decomposition method. However, these methods only used for MR imaging from 1D partial k-space. Based on researching these existing methods of removing truncation artifacts, including advantages and disadvantages. This paper proposed application of POCS method and TV means to remove truncation artifacts when MRI reconstruction from 2D partial k-space. POCS method is based on the conjugate phase correction and synthesis of iterative reconstruction algorithms, low-frequency data is used for phase correction, to recover the missing high-frequency data, iterative process repeated until the results stabilized. Total Variation (TV) method is the earliest and best PDE-based denoising model for edge-preserving. By changing the image model to minimize total variation, this method purposes to remove the unwanted image noise and small-scale details of the same time and protect those who are non-consecutive region mutations.This paper proposed two methods is verified with much experiment of digital model images and simulation images. We can conclude that reconstruction of these two methods are better than zero-padding method, their reconstruction errors are small. For the digital model image, TV method is better than POCS method, while POCS reconstruction errors are smaller in the simulation image.