METHOD FOR LUNG LESION LOCATION IDENTIFICATION

• US 20090185731A1
• Filed: 01/23/2008
• Published: 07/23/2009
• Est. Priority Date: 01/23/2008
• Status: Active Grant

• Current Assignee: CareStream Health Incorporated
• Original Assignee: CareStream Health Incorporated
• Inventors: Ray, Lawrence A., Simon, Richard A., Nicponski, Henry, Gindele, Edward B.
• Granted Patent: US 8,150,113 B2
• Time in Patent Office: Days
• US Class Current: 382/131
• CPC Class Codes: G06T 2207/10081 Computed x-ray tomography [CT]
• CPC Class Codes: G06T 2207/10088 Magnetic resonance imaging [MRI]
• CPC Class Codes: G06T 2207/20101 Interactive definition of point of interest, landmark or seed
• CPC Class Codes: G06T 2207/30061 Lung
• CPC Class Codes: G06T 2207/30101 Blood vessel; Artery; Vein; Vascular
• CPC Class Codes: G06T 7/0012 Biomedical image inspection
• CPC Class Codes: G06T 7/12 Edge-based segmentation
• CPC Class Codes: G06T 7/149 involving deformable models, e.g. active contour models
• CPC Class Codes: G06T 7/162 involving graph-based methods
• CPC Class Codes: Y10S 128/92 Computer assisted medical diagnostics
• CPC Class Codes: Y10S 128/922 including image analysis
• CPC Class Codes: Y10S 128/923 by comparison of patient data to other data

Abstract

A method and a system are disclosed for labeling an anatomical point associated with a lesion in an organ such as a lung. The method includes: a segmentation of a vessel tree anatomical structure starting from an autonomously determined initial image point; labeling the vessel segments of the vessel tree segmentation with segment labels based on a priori anatomical knowledge, thereby creating an individualized anatomical model; receiving a user-specified image point having a location from a user and locating a nearby vessel structure; tracking along the vessel structure in a direction towards a root of a parent vessel tree until a prior labeled vessel segment is encountered in the anatomical model, and assigning the label of the encountered prior labeled vessel segment from the anatomical model as an anatomical location label of the user-specified image point.

Claims

1. A method for labeling a user-specified image point with an anatomical label, comprising:

segmentation of a vessel tree anatomical structure starting from an autonomously determined initial image point;

labeling the vessel segments of the vessel tree segmentation with segment labels based on a priori anatomical knowledge, thereby creating an individualized anatomical model;

receiving a user-specified image point having a location from a user and locating a nearby vessel structure;

tracking along the vessel structure in a direction towards a root of a parent vessel tree until a prior labeled vessel segment is encountered in the anatomical model, and

assigning the label of the encountered prior labeled vessel segment from the anatomical model as an anatomical location label of the user-specified image point.

2. The method as recited in claim 1, wherein the user-specified image point identifies the location of an anatomical lesion.

3. The method as recited in claim 1, wherein segmentation and labeling of the vessel tree anatomical structure are confined to beginning parts of the vessel tree where segmentation and labeling can be performed with a highest accuracy.

4. The method as recited in claim 1, wherein the segmentation and labeling are performed for two or more parallel vessel tree anatomical structures associated to an organ.

5. The method as recited in claim 4, wherein an autonomously identified most reliable vessel tree with anatomical model and labeling thereof are used to assign the anatomical location label to the user-specified image point.

6. The method as recited in claim 1, wherein the process of locating a vessel nearby to a user-specified seed point comprises randomly searching around the image point location until a vessel is found.

7. The method as recited in claim 1, wherein the process of tracking along the vessel structure in the direction towards the root of the parent vessel tree until the prior labeled vessel segment is encountered includes the process of:

measuring a cost function along a path of tracking and summing a total path cost; and

rejecting the tracking as failed if the total path cost exceeds a fixed threshold thr_cost.

8. The method as recited in claim 4, wherein an autonomously-identified hierarchical ranking of the reliability of vessel tree anatomical models with labeling thereof is used to assign an anatomical label to the user-specified image point, according to:

assigning the anatomical location label of a vessel segment from the a priori most reliable anatomical model as the label of the user-specified image point, if it is possible to track along a vessel structure from the user-specified image point in the direction towards the root of the vessel tree until a prior labeled vessel segment is encountered, the success of the tracking being assessed by;

a) randomly searching around the user-specified image point until a vessel is found;

b) finding a minimum-cost path back to the root of the anatomical model by measuring a cost function along a path of tracking and summing a total path cost;

c) rejecting the tracking operation as failed if the total path cost exceeds a fixed threshold thr_cost, otherwise accepting the tracking operation by assigning the label of the vessel segment and terminating the labeling process; and

d) repeating (a)-(c), if necessary, a fixed number of times to initiate the tracking from different starting vessel points until success occurs;

repeating, if the labeling operation using the first anatomical model and vessel tree did not complete successfully, the vessel structure tracking operation using the a priori next most reliable anatomical model and vessel tree until a prior labeled vessel segment is encountered, and assigning the label of the vessel segment if the tracking operation in successful; and

repeating this labeling and tracking operation using all segmented vessel trees in a priori decreasing priority order, as necessary.

9. The method as recited in claim 4, wherein a user-identified hierarchical ranking of the reliability of vessel tree anatomical models with labeling thereof is used to assign the anatomical location label to the user-specified image point, according to:

assigning the label of a vessel segment from a user-identified most reliable anatomical model as the anatomical location label of the user-specified image point, if it is possible to track along a vessel structure from the user-specified image point in the direction towards the root of the vessel tree until a prior labeled vessel segment is encountered, the success of the tracking being assessed by;

a) randomly searching around the user-specified image point until a vessel is found;

b) finding a minimum-cost path back to the root of the anatomical model by measuring a cost function along a path of tracking and summing a total path cost;

c) rejecting the tracking operation as failed if the total path cost exceeds a fixed threshold thr_cost, otherwise accepting the tracking operation by assigning the label of the vessel segment and terminating the labeling process; and

d) repeating (a)-(c), if necessary, a fixed number of times to initiate the tracking from different starting vessel points until success occurs;

repeating, if the labeling operation using the first anatomical model and vessel tree did not complete successfully, the vessel structure tracking operation using the user-identified next most reliable anatomical model and vessel tree until a prior labeled vessel segment is encountered, and assigning the label of the vessel segment if the tracking operation in successful; and

repeating this labeling and tracking operation using all segmented vessel trees in user-identified decreasing priority order, as necessary.

10. The method as recited in claim 4, wherein a plurality of vessel tree anatomical models with labeling thereof are used to assign the anatomical location label to the user-specified image point with increased reliability, according to:

assigning the label of a vessel segment from each anatomical model to the user-specified image point, if it is possible to track along a vessel structure from the user-specified image point in the direction towards the root of the vessel tree until a prior labeled vessel segment is encountered, the success of the tracking being assessed by;

a) randomly searching around the user-specified image point until a vessel is found;

b) finding a minimum-cost path back to the root of the anatomical model by measuring a cost function along the path of tracking and summing the total path cost;

c) rejecting the tracking operation as failed if the total path cost exceeds a fixed threshold thr_cost, otherwise accepting the tracking operation by assigning the label of the vessel segment and terminating the labeling process; and

d) repeating (a)-(c) if necessary a fixed number of times to initiate the tracking from different starting vessel points until success occurs; and

accumulating a set of anatomical labels, one from each anatomical model.

11. The method as recited in claim 10, wherein the user-specified image point is assigned the anatomical label derived from all anatomical models, when the anatomical models unanimously provide the same label.

12. The method as recited in claim 10, wherein the autonomously-identified hierarchical ranking of the reliability of vessel tree anatomical structures with labeling thereof is used to assign the anatomical location label to the user-specified image point, where furthermore the anatomical models fail to provide an unanimous label for the point, in which case the label from the a priori highest reliability vessel tree is assigned to the user-specified image point;

and the system user is notified of the lack of unanimous labeling.

13. The method as recited in claim 10, wherein an user-specified hierarchical ranking of the reliability of vessel tree anatomical structures with labeling thereof is used to assign an anatomical label to the user-specified image point, where furthermore the anatomical models fail to provide an unanimous label for the user-specified point, in which case the label from the user-specified highest reliability vessel tree is assigned to the user-specified image point;

and a system user is notified of the lack of unanimous labeling.

14. The method as recited in claim 1, wherein the segmentation of vessel tree anatomical structures comprises:

transforming image pixels using a threshold;

examining image slices one slice at a time until a location and size of a connected component match anatomical criteria to identify the root of the vessel tree anatomical structure; and

creating a binary recursive tree starting from the root having a pre-determined number of bifurcations by examining image slices one slice at a time and growing a vessel segment responsive to known anatomical vessel dimensions.

15. The method as recited in claim 14, wherein the labeling of vessel tree anatomical structures is further comprising:

creating a tree graph from thin skeletons of the vessel segments; and

labeling segment branches of the graph using a priori anatomical knowledge.

16. The method as recited in claim 15, wherein the labeling of lung bronchial vessel tree segments uses known constraints on the spatial arrangements of the lung organ sub-volumes supplied by primary vessel segments and further comprises:

starting with an uppermost image slice that contains part of a clear field of a left lung, selecting an arbitrary image voxel, tracing a path backwards until a segmentation for a left primary bronchial vessel branch is encountered, which is then labeled as an upper left main branch;

labeling an other branch of two main left lung bronchial branches as a lower left main branch;

starting with the uppermost image slice that contains part of the clear field of a right lung, selecting an arbitrary image voxel, tracing a path backwards until the segmentation for a right primary bronchial vessel branch is encountered, which is then labeled as a right eparterial branch;

starting with a lowermost image slice that contains part of the clear field of a right lung, selecting an arbitrary image voxel, tracing a path backwards until the segmentation for a right primary bronchial vessel branch is encountered, which is then labeled as a right lower hyparterial branch;

labeling a remaining middle right primary bronchial vessel branch as a right middle hyparterial branch.

17. A system, comprising:

a computer performing;

segmentation of a vessel tree anatomical structure starting from an autonomously determined initial image point;

labeling the vessel segments of the vessel tree segmentation with segment labels based on a priori anatomical knowledge, thereby creating an individualized anatomical model;

receiving a user-specified image point having a location from a user and locating a nearby vessel structure;

tracking along the vessel structure in a direction towards a root of a parent vessel tree until a prior labeled vessel segment is encountered in the anatomical model; and

assigning the label of the encounter prior labeled vessel segment from the anatomical model as an anatomical location label of the user-specified image point; and

a display displaying a report showing the anatomical location label for the user-specified image point.

18. A method, comprising:

vessel tree tracking associated with a seed point tree root; and

mapping vessel connectivity to an anatomical label for a user specified point using an anatomical model.

19. A method, comprising:

vessel tree tracking associated with a seed point tree root, and

mapping vessel connectivity to an anatomical label using an anatomical model by labeling a specified image point with the label of an anatomically labeled segment.

20. A method, comprising:

vessel tree tracking associated with a seed point tree root, and

mapping vessel connectivity to an anatomical label using an anatomical model;

labeling a specified image point with the label of an anatomically labeled segment, comprising;

randomly searching around the image point until a vessel is found;

tracing the vessel until a labeled segment is found; and

creating a report record correlating the specified image point to the labeled segment where the report includes a entry having the lesion location specified image point and a text segment label for the found segment.