Method and apparatus for automated detection of target structures from medical images using a 3d morphological matching algorithm
First Claim
1. A method for automated detection of target structures shown in digital medical images, the method of comprising:
- generating a three dimensional (3D) volumetric data set of a patient region within which the target structure resides from a plurality of segmented medical image slices;
grouping contiguous structures that are depicted in the 3D volumetric data set to create corresponding grouped structure data sets;
assigning each grouped structure data set to one of a plurality of detection algorithms, each detection algorithm being configured to detect a different type of target structure; and
processing each grouped structure data set according to its assigned detection algorithm to thereby detect whether any target structures are present in the medical images.
0 Assignments
0 Petitions
Accused Products
Abstract
A method for the detection of target structures shown in digital medical images, the method of comprising: (1) generating a three dimensional (3D) volumetric data set of a patient region within which the target structure resides from a plurality of segmented medical image slices; (2) grouping contiguous structures that are depicted in the 3D volumetric data set to create corresponding grouped structure data sets; (3) assigning each grouped structure data set to one of a plurality of detection algorithms, each detection algorithm being configured to detect a different type of target structure; and (4) processing each grouped structure data set according to its assigned detection algorithm to thereby detect whether any target structures are present in the medical images. Preferably, the target structures are pulmonary nodules, and a specialized detection algorithm is applied to image data classified as a candidate for depicting perivascular nodules. To segment perivascular nodule candidates from surrounding vessels, the image data is preferably correlated with a plurality of 3D morphological filters.
54 Citations
68 Claims
-
1. A method for automated detection of target structures shown in digital medical images, the method of comprising:
-
generating a three dimensional (3D) volumetric data set of a patient region within which the target structure resides from a plurality of segmented medical image slices;
grouping contiguous structures that are depicted in the 3D volumetric data set to create corresponding grouped structure data sets;
assigning each grouped structure data set to one of a plurality of detection algorithms, each detection algorithm being configured to detect a different type of target structure; and
processing each grouped structure data set according to its assigned detection algorithm to thereby detect whether any target structures are present in the medical images. - View Dependent Claims (2, 3, 4, 5, 6, 7, 8, 9, 10)
-
-
11. A device for detecting whether pulmonary nodules are present in a patient'"'"'s lung region from a three-dimensional (3D) data set representative of a volumetric image of the patient'"'"'s lung region, the device comprising:
a processor configured to (1) identify contiguous structures in the 3D data set, (2) classify the identified contiguous structures according to a plurality of classifications, the classifications comprising a vessel contiguous structure classification and a non-vessel contiguous structure classification, (3) apply a first nodule detection operation to each vessel contiguous structure to determine a nodule status therefor, and (4) apply a second nodule detection operation to each non-vessel contiguous structure to determine a nodule status therefor, wherein the first nodule detection operation is different than the second nodule detection algorithm. - View Dependent Claims (12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22, 23, 24, 25, 26, 27, 28, 29, 30, 31, 32, 33, 34, 35, 36, 67, 68)
-
37. A device for analyzing a 3D data set representative of a patient'"'"'s lung region, the device comprising:
a processor configured to (1) group the data set into data subsets, each subset being representative of a contiguous structure, (2) identify each data subset that corresponds to a vessel, and (3) segment any perivascular nodule candidates from each identified subset by correlating that identified subset with at least one 3D morphological filter that is tuned to an expected shape of a perivascular nodule. - View Dependent Claims (38, 39, 40, 41, 42, 43, 44, 45, 46, 47, 48, 49, 50)
-
51. A computer-readable medium for detecting whether pulmonary nodules are present in a patient'"'"'s lung region from a 3D data set representative of a volumetric image of the patient'"'"'s lung region, the computer-readable medium comprising:
a plurality of instructions executable by a processor for;
(1) identifying contiguous structures in the data set;
(2) classifying the identified contiguous structures according to a plurality of classifications, the classifications comprising a vessel contiguous structure classification and a non-vessel contiguous structure classification;
(3) applying a first nodule detection operation to each vessel contiguous structure to determine a nodule status therefor; and
(4) applying a second nodule detection operation to each non-vessel contiguous structure to determine a nodule status therefor, wherein the first nodule detection operation is different than the second nodule detection operation.- View Dependent Claims (52, 53, 54, 55)
-
56. A device configured to automatically detect the presence of pulmonary nodules depicted within a three-dimensional data set representative of a patient'"'"'s lung region, the data set comprising a plurality of contiguous structures associated with a first classification and a plurality of contiguous structures associated with a second classification, the device comprising:
a processor configured to (1) apply a first nodule detection algorithm to contiguous structures associated with the first classification, and (2) apply a second nodule detection algorithm to contiguous structures associated with the second classification. - View Dependent Claims (57, 58)
-
59. A system for automatic detection of pulmonary nodules shown in CT images, the system comprising:
-
a CT scanner for scanning a patient'"'"'s lung region to generate a CT image data set;
a computer configured to (1) segment CT image data corresponding to the patient'"'"'s lung region from the CT image data set, (2) generate a three dimensional volumetric data set of a patient'"'"'s lung region from the segmented CT image data, (3) group contiguous structures that are depicted in the volumetric data set to create corresponding grouped structure data sets, (4) classify each grouped structure data set as either a vessel group data set or a non-vessel group data set, (5) for each non-vessel group data set, determine a nodule status for the structure depicted therein based on geometric criteria, and (6) for each vessel group data set, (a) convolve the vessel group data set with a predetermined morphological filter to thereby compute a correlation value between the vessel group data set and the morphological filter, (b) select a vessel group data set if its correlation value is within a predetermined range, and (c) determine a nodule status for the structure depicted in the selected vessel group data set based on geometric criteria.
-
-
60. A computer readable medium for use with computer-aided diagnosis of pulmonary nodules present in computed tomography (CT) images, the computer readable medium comprising:
-
a code segment configured to generate a three dimensional (3D) volumetric data set of a patient'"'"'s lung region from a plurality of segmented CT images;
a code segment configured to classify each grouped structure data set as either a vessel group data set or a non-vessel group data set;
a code segment configured to, for each non-vessel group data set, determine a nodule status for the structure depicted therein based on geometric criteria; and
a code segment configured to, for each vessel group data set, (1) convolve the vessel group data set with a predetermined morphological filter to thereby compute a correlation value between the vessel group data set and the morphological filter, (2) select a vessel group data set if its correlation value is within a predetermined range, and (3) determine a nodule status for the structure depicted in the selected vessel group data set based on geometric criteria.
-
-
61. A method of analyzing a data set representative of a region of the patient'"'"'s body, the method comprising:
-
grouping contiguous structures depicted in the data set;
identifying contiguous structures that correspond to a predefined classification; and
segmenting target structure candidates from the identified structures by correlating each identified structure with at least one filter that is tuned to an expected shape of a target structure. - View Dependent Claims (62, 63, 64, 65, 66)
-
Specification