Method for segmenting medical images and detecting surface anomalies in anatomical structures
First Claim
1. A computer assisted method for detecting surface anomalies in a three-dimensional surface model representing a three-dimensional structure, the method comprising:
- fitting a parametric patch to a local neighborhood around a vertex in the surface model;
computing a curvature characteristic of the fitted patch in the local neighborhood around the vertex;
classifying curvature of a cluster of connected vertices by comparing the computed curvature characteristic to a predetermined curvature classification; and
defining a surface anomaly as a cluster of connected vertices having a predetermined curvature classification.
1 Assignment
0 Petitions
Accused Products
Abstract
A region growing method segments three-dimensional image data of an anatomical structure using a tortuous path length limit to constrain voxel growth. The path length limit constrains the number of successive generations of voxel growth from a seed point to prevent leakage of voxels outside the boundary of the anatomical structure. Once segmented, a process for detecting surface anomalies performs a curvature analysis on a computer model of the surface of the structure. This process detects surface anomalies automatically by traversing the vertices in the surface model, computing partial derivatives of the surface at the vertices, and computing curvature characteristics from the partial derivatives. To identify possible anomalies, the process compares the curvature characteristics with predetermined curvature characteristics of anomalies and classifies the vertices. The process further refines potential anomalies by segmenting neighboring vertices that are classified as being part of an anomaly using curvature characteristics. Finally, the process colorizes the anomalies and computes a camera position and direction for each one to assist the user in viewing 2D renderings of the computer model.
201 Citations
13 Claims
-
1. A computer assisted method for detecting surface anomalies in a three-dimensional surface model representing a three-dimensional structure, the method comprising:
-
fitting a parametric patch to a local neighborhood around a vertex in the surface model;
computing a curvature characteristic of the fitted patch in the local neighborhood around the vertex;
classifying curvature of a cluster of connected vertices by comparing the computed curvature characteristic to a predetermined curvature classification; and
defining a surface anomaly as a cluster of connected vertices having a predetermined curvature classification. - View Dependent Claims (2, 3, 4, 5, 6, 7, 8, 9)
segmenting a region of the surface model by evaluating neighboring vertices of the vertex, and combining together a cluster of neighboring vertices that satisfy a predetermined curvature classification.
-
-
6. The method of claim 1 further including:
finding partial derivatives of the parametric patch at a vertex, deriving principal curvatures at the vertex, classifying curvature at the vertex as one of a predetermined curvature type based on the principal curvatures, and determining whether the curvature matches a known curvature characteristic of a surface anomaly to evaluate whether a surface anomaly exists at the vertex.
-
7. The method of claim 1 further including:
-
starting from a point in a three-dimensional data set derived from a scan of the structure, growing a region in the structure by modifying coordinates of a voxel in the region to find a next generation voxel, examining a characteristic of the next generation voxel, and adding the next generation voxel to the region when the characteristic of the voxel is within a threshold;
growing the region with successive generations of voxels along a path in the three-dimensional data set using a path length limit to limit the number of generations of voxel growth from the starting point;
creating the 3D surface model approximating points on a surface of the structure from the voxels identified as being part of the region; and
rendering the 3D surface model from a camera position directed at the surface anomaly to create a two-dimensional image representing the structure and the surface anomaly.
-
-
8. The method of claim 1 further including:
-
traversing the surface model to identify surface anomalies that have a predetermined curvature; and
flagging the surface anomalies to enable a user to view each of surface anomalies selectively as 2D image renderings of the surface model from camera positions directed at the surface anomalies.
-
-
9. A computer readable medium having instructions for performing the steps of claim 1.
-
10. A method for viewing 2D image renderings of a 3D surface model representing a surface of a three-dimensional structure, the method including:
-
identifying surface anomalies on the surface by identifying regions of the surface that have a predetermined curvature;
for each surface anomaly, computing representative coordinates approximating a location of the surface anomaly in the surface model, and computing a camera position for viewing the anomaly by aligning the camera position in a direction toward the representative coordinates and located at predetermined distance from the representative coordinates; and
in a 3D graphics rendering process used to view 2D image renderings of the surface model from selected camera positions, using the camera positions computed for each surface anomaly to enable a user to view selectively 2D image renderings of the 3D representation that include a rendering of the surface anomalies;
wherein the step of identifying surface anomalies includes;
fitting a parametric patch to a local neighborhood around a vertex in the surface model;
computing an extent of curvature in the local neighborhood around the vertex;
segmenting a region of the surface model by evaluating neighboring vertices of the vertex, and combining together a cluster of neighboring vertices that satisfy a predetermined curvature classification; and
defining a surface anomaly as a cluster of connected vertices having a predetermined curvature classification.
-
-
11. A method for viewing 2D image renderings of a 3D surface model representing a surface of a three-dimensional structure, the method including:
-
identifying surface anomalies on the surface by identifying regions of the surface that have a predetermined curvature;
for each surface anomaly, computing representative coordinates approximating a location of the surface anomaly in the surface model, and computing a camera position for viewing the anomaly by aligning the camera position in a direction toward the representative coordinates and located at predetermined distance from the representative coordinates;
in a 3D graphics rendering process used to view 2D image renderings of the surface model from selected camera positions, using the camera positions computed for each surface anomaly to enable a user to view selectively 2D image renderings of the 3D representation that include a rendering of the surface anomalies; and
finding partial derivatives of the parametric patch at a vertex, deriving principal curvatures at the vertex, classifying curvature at the vertex as one of a predetermined curvature type based on the principal curvatures, and determining whether the curvature matches a known curvature characteristic of a surface anomaly to evaluate whether a surface anomaly exists at the vertex;
wherein the step of identifying surface anomalies includes;
fitting a parametric patch to a local neighborhood around a vertex in the surface model;
computing an extent of curvature in the local neighborhood around the vertex;
segmenting a region of the surface model by evaluating neighboring vertices of the vertex, and combining together a cluster of neighboring vertices that satisfy a predetermined curvature classification; and
defining a surface anomaly as a cluster of connected vertices having a predetermined curvature classification.
-
-
12. A method for viewing 2D image renderings of a 3D surface model representing a surface of a three-dimensional structure, the method including:
-
identifying surface anomalies on the surface by identifying regions of the surface that have a predetermined curvature;
for each surface anomaly, computing representative coordinates approximating a location of the surface anomaly in the surface model, and computing a camera position for viewing the anomaly by aligning the camera position in a direction toward the representative coordinates and located at predetermined distance from the representative coordinates;
in a 3D graphics rendering process used to view 2D image renderings of the surface model from selected camera positions, using the camera positions computed for each surface anomaly to enable a user to view selectively 2D image renderings of the 3D representation that include a rendering of the surface anomalies;
starting from a point in a three-dimensional data set produced in a scan of a real anatomical structure, growing a region in the structure by modifying coordinates of a voxel in the region to find a next generation voxel, examining a characteristic of the next generation voxel, and adding the next generation voxel to the region when the characteristic of the voxel is within a threshold; and
growing the region with successive generations of voxels along a path in the three-dimensional data set using a path length limit to limit the number of generations of voxel growth from the starting point.
-
-
13. A method for viewing 2D image renderings of a 3D surface model representing a surface of a three-dimensional structure, the method including:
-
identifying surface anomalies on the surface by identifying regions of the surface that have a predetermined curvature;
for each surface anomaly, computing representative coordinates approximating a location of the surface anomaly in the surface model, and computing a camera position for viewing the anomaly by aligning the camera position in a direction toward the representative coordinates and located at predetermined distance from the representative coordinates; and
in a 3D graphics rendering process used to view 2D image renderings of the surface model from selected camera positions, using the camera positions computed for each surface anomaly to enable a user to view selectively 2D image renderings of the 3D representation that include a rendering of the surface anomalies;
wherein the step of identifying the surface anomalies includes;
applying a 3D filter to an array of voxels representing the structure to compute partial derivatives of the surface, where the surface model is used as a guide to determine where to apply the 3D filter in the array of voxels;
from the partial derivatives, computing curvature characteristics of the surface; and
comparing the curvature characteristics with the predetermined curvature characteristics to identify regions having the predetermined curvature characteristics.
-
Specification