K-space Sampling for Organ Structure Identification in MRI

The patent applies local quality by differentiating the sampling density across different regions of k-space. Specifically, the outer region of k-space is sampled with higher density than the central region, creating non-uniform sampling that emphasizes outer contour information while maintaining overall image quality. This localized sampling strategy directly improves outer contour identification precision without requiring complex uniform sampling throughout the entire k-space.

The patent implements preliminary action by pre-emphasizing outer contour information through optimized k-space sampling before the segmentation process begins. The non-uniform sampling scheme预先 (in advance) enhances the outer contour signals in the acquired data, which then facilitates faster and more accurate automated or semi-automated segmentation. This preliminary enhancement reduces the time and effort required for subsequent manual or algorithmic segmentation while maintaining high accuracy.

The patent applies mechanics substitution by replacing the CT-based density mapping approach with an MR-based segmentation approach. Instead of using ionizing radiation (CT) to obtain anatomical information for treatment planning, the method uses optimized MR imaging with non-uniform k-space sampling to achieve precise organ contour identification. This substitution eliminates the need for additional patient radiation exposure from CT scans while maintaining the reliability needed for accurate radiation treatment planning through improved outer contour emphasis.