Second-Derivative Splitting Filters for Bone Segmentation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional bone segmentation algorithms for small animal micro-CT imaging suffer from under-segmentation, over-segmentation, and incorrect placement of split lines/planes due to limitations in morphological approaches, particularly in low-resolution data where partial volume effects cause objects to become morphologically connected, leading to significant loss in segmentation accuracy.
Innovation Solution
The use of single and hybrid second-derivative splitting filters applied to gray-scale images and binary bone masks prior to watershed segmentation, which accurately identify split lines/planes and disconnect individual bones, enhancing segmentation robustness and accuracy by leveraging second-derivative functions.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If conventional morphological segmentation approaches (watershed transformation) are applied to binary bone masks, then the segmentation process is simple and fast, but segmentation accuracy deteriorates due to under-segmentation, over-segmentation, and incorrect placement of split lines
Solution Approach 1:
The patent applies preliminary action by performing second-derivative filtering on the gray-scale image data before converting to binary mask and applying watershed transformation. This preprocessing step enhances the visibility of bone boundaries and separation regions, ensuring that subsequent segmentation operations occur on pre-enhanced data with clearly defined split lines, thereby preventing under-segmentation and over-segmentation while maintaining computational efficiency
Solution Approach 2:
The patent changes parameters by transforming the image data through second-derivative operations that emphasize boundary regions. By applying differential operators to the gray-scale image, the method alters the parameter distribution to highlight separation zones between adjacent bones, enabling more accurate split line placement without requiring complex post-processing corrections
2Productivity
If low-resolution micro-CT imaging is used to speed up image acquisition, then imaging throughput increases, but segmentation accuracy deteriorates due to partial volume effects causing morphological connections between separate bones
Solution Approach 1:
The patent addresses partial volume effects by applying second-derivative filtering that is sensitive to intensity transitions. This parameter transformation enhances the detection of boundaries between adjacent bones even when they appear connected in low-resolution images, as the derivative operations amplify the subtle intensity variations at interfaces that are obscured by partial volume averaging
Solution Approach 2:
By performing second-derivative enhancement before binary thresholding and segmentation, the method prepares the low-resolution data in advance to compensate for resolution limitations. This preliminary processing creates enhanced boundary information that guides subsequent segmentation steps, enabling accurate bone separation despite the underlying partial volume effects in the original low-resolution scan
Data Source
Figure 1
Figure 2
Figure 3
AI summary
Presented herein, in certain embodiments, are approaches for robust bone splitting and segmentation in the context of small animal imaging, for example, microCT imaging. In certain embodiments, a method for calculating and applying single and hybrid second-derivative splitting filters to gray-scale images and binary bone masks is described. These filters can accurately identify the split lines/planes of the bones even for low-resolution data, and hence accurately morphologically disconnect the individual bones. The split bones can then be used as seeds in region growing techniques such as marker-controlled watershed segmentation. With this approach, the bones can be segmented with much higher robustness and accuracy compared to prior art methods.