Rib Segmentation Using Probability Maps and Hit-or-Miss Labeling
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Manual segmentation of ribs in medical images is time-consuming, and automatic segmentation is challenging due to unclear rib boundaries, adhesion to other bones, missing ribs, or pathological conditions, making it difficult to accurately analyze anatomical structures and diagnose diseases.
Innovation Solution
A system and method using artificial intelligence algorithms to acquire and process medical images, determining rib regions, generating rib-probability-maps, and tracking rib segments to segment and label ribs accurately, even in complex cases like adhesion to vertebrae or missing parts.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If manual segmentation is used to label ribs in medical images, then segmentation accuracy can be maintained, but the processing time and labor cost increase significantly
Solution Approach 1:
The system performs preliminary actions by pre-processing the medical image to enhance rib structures, generate initial segmentation masks, and identify candidate rib regions before the main segmentation process. This preliminary preparation reduces the complexity of subsequent automatic segmentation while maintaining accuracy standards.
Solution Approach 2:
The patent introduces an intermediary semi-automatic mode where user input serves as a mediator between manual and fully automatic segmentation. Users can provide minimal guidance (e.g., clicking on starting points or correcting obvious errors) while the system handles the rest automatically, achieving both speed and accuracy.
2Productivity
If automatic segmentation algorithms are used to process rib images, then processing efficiency is improved, but segmentation accuracy deteriorates due to unclear boundaries and adhesion to other bones
Solution Approach 1:
The patent applies segmentation by dividing the complex rib segmentation task into multiple independent stages: pre-processing to enhance rib structures, generation of probability maps to identify rib locations, model-based tracking to follow rib contours, and post-processing to refine boundaries. Each stage handles specific aspects of the problem, improving overall accuracy while maintaining automation.
Solution Approach 2:
The system dynamically adjusts processing parameters based on local image characteristics. Different regions of the image may use different threshold values, tracking step sizes, or model parameters adapted to local bone density, boundary clarity, and anatomical variations, thereby maintaining accuracy across diverse cases.
3Adaptability or versatility
If the segmentation system processes complex cases with adhesion or missing ribs, then diagnostic capability is improved, but the algorithm complexity increases
Solution Approach 1:
The patent implements dynamic adaptation where the segmentation algorithm automatically adjusts its behavior based on detected anatomical variations. When adhesion to vertebrae or missing ribs are detected, the system dynamically modifies tracking parameters, boundary detection sensitivity, and model fitting approaches to handle these complex cases appropriately.
Data Source
AI summary
Methods and systems for image processing are provided. Image data may be obtained. The image data may include a plurality of voxels corresponding to a first plurality of ribs of an object. A first plurality of seed points may be identified for the first plurality of ribs. The first plurality of identified seed points may be labelled to obtain labelled seed points. A connected domain of a target rib of the first plurality of ribs may be determined based on at least one rib segmentation algorithm. A labelled target rib may be obtained by labelling, based on a hit-or-miss operation, the connected domain of the target rib, wherein the hit-or-miss operation may be performed using the labelled seed points to hit the connected domain of the target rib.