GOP® methodology is unique in detecting structures by examining the significance of each pixel in an image in relation to the wider context in which they appear. In this way, once the structure is identified and analyzed, noise can be suppressed and the true structure, however weak, can be emphasized and seen more clearly.
Please observe that methods and images presented here are only examples and demonstrations of the GOP technology and do not represent areas where ContextVision is involved today.
GOP image enhancement:
- reduces noise and other errors in the digital image
- enhances edges and lines to display more sharply defined structures
- enhances image contrast, making vague structures strong
- provides latitude compression so that structures in both dense and translucent areas are visible simultaneously
For the clinician this means:
- better visibility and contrast of small and/or weak structures
- better depiction of overlapping structures in CR and DR
- better delineation of bone structures in CR and DR
- simultaneous viewing of bone and soft tissue in CR and DR
- better contrast between white and grey matter in MR brain scans
- reduced speckle and haze in ultrasound
Conventional image processing generally assumes that each individual pixel has a particular significance, e.g. that the intensity or color information about a pixel determines its relevance to the image as a whole. In many professions, this approach is sufficient to solve an image-processing task. However, a more accurate method is required in more complex image structures, where we only realize the significance of individual pixels when they are seen in their contextual environment.
An image contains different levels of data (see diagram above). The data levels create a hierarchical structure, with the lowest level representing the original image and higher levels representing more detailed information about the image, i.e. edges and lines, their orientation and curvature. This provides the basis for ContextVision’s GOP methodology, proving that all image information can be captured and examined without any loss of detail.
The information obtained from different levels in the processing is used for a variety of purposes. In normal image processing, a one-level analysis defines the final result. However, GOP image enhancement goes one stage further; images from higher levels control operations performed on lower levels. When no structures are found, noise reduction is performed, and where structures are detected, they are thus maintained. The method can be taken even further by not only detecting and retaining information, but also sharpening and clarifying structures.