Abstract:

BACKGROUND: Super-resolution reconstruction has been extensively studied and used in many fields, such as medical diagnostics, military surveillance, frame freeze in video, and remote sensing.
OBJECTIVE: In order to obtain high-resolution magnetic resonance images, gradient magnetic field is required and the signal-to-noise will be reduced due to the decrease in voxel size with traditional scan. The present study used a self-adaptive regularized super-resolution reconstruction algorithm to acquire high-resolution magnetic resonance images from four half-pixel-shifted low resolution images.
METHODS: The least squares algorithm was used as a cost function. The derivative of the cost function was calculated to obtain an iterative formula of super-resolution reconstruction. In the process of iterative process, the parameter and step size of image resolution were regularized.
RESULTS AND CONCLUSION: The new regularization parameter makes cost function of the new algorithm convex within the definition region. The piori information is involved in the regularization parameter that can improve the high-frequency components of the restored image. As shown from the results obtained in the phantom imaging, the proposed super-resolution technique can improve the resolution of magnetic resonance image.