Patent US5881124A successfully refreshed from Google Patents.
US5881124A
Expired / Lapsed
US
⏱️ 0.0 years remaining
Automated method and system for the detection of lesions in medical computed tomographic scans
Assignee: Arch Development Corp
•
Filing: 1994-03-31
•
Granted: 1999-03-09
•
Expires: 2014-03-31
Abstract
Abstract A method and system for the automated detection of lesions in computed tomographic images, including generating image data from at least one selected portion of an object, for example, from CT images of the thorax. The image data are then analyzed in order to produce the boundary of the thorax. The image data within the thoracic boundary is then further analyzed to produce boundaries of the lung regions using predetermined criteria. Features within the lung regions are then extracted using multi-gray-level thresholding and correlation between resulting multi-level threshold images and between at least adjacent sections. Classification of the features as abnormal lesions or normal anatomic features is then performed using geometric features yielding a likelihood of being an abnormal lesion along with its location in either the 2-D image section or in the 3-D space of the object.
Inventors
- 👤 Maryellen L. Giger
- 👤 Kyongtae Ty Bae
- 👤 Kunio Doi
Bibliographic Details
Application Number
—
Jurisdiction
US
Data Source
google_patents
Last Synchronized
2026-11-09 22:38
CPC Classifications
G; G06; G06T; G06T7/00; G06T7/0002; G06T7/0012; G06T2207/00; G06T2207/10; G06T2207/10072; G06T2207/10081; G06T2207/30; G06T2207/30004; G06T2207/30061; G06T2207/30064; Y; Y10; Y10S; Y10S378/00; Y10S378/901
Claims Analysis
Total Claims: 49
Claim 1
Independent Claim
56 words
A method for the automated detection of nodules in a computed tomographic (CT) scan of a subject, comprising: detecting an anatomic region of said subject in said CT scan; and detecting a nodule in said anatomic region using a plurality of images of said anatomic region in said CT scan generated at different section-specific gray-level thresholds.
Claim 2
Independent Claim
56 words
A method for the automated detection of nodules in a computed tomographic (CT) scan of a subject, comprising: detecting an anatomic region of said subject in said CT scan; and detecting a nodule in said anatomic region using a plurality of images of said anatomic region in said CT scan generated at different section-specific gray-level thresholds.
Claim 3
Dependent Claim
48 words
A method as recited in claim 1, wherein said step of detecting said anatomic region comprises: forming a histogram of gray-values of pixels in said CT scan; determining a gray-level threshold using said histogram; and generating a binary image of said anatomic region using said gray level threshold.
Claim 4
Dependent Claim
97 words
A method as recited in claim 2, comprising: forming a histogram of gray-values of pixels in said CT scan from a corner of said CT scan to a center of said scan; wherein said step of selecting said gray level threshold comprises: determining a first cumulative mean of said histogram from said corner to said center of said CT scan; determining a second cumulative mean of said histogram from said center to said corner of said CT scan; determining a difference between said first and second cumulative means; and using said difference to determine said gray-level threshold.
Claim 5
Dependent Claim
35 words
A method as recited in claim 2, wherein said step of detecting said anatomic region further comprises: detecting a boundary of said binary image; and evaluating geometric parameters of a region enclosed by said boundary.
Claim 6
Dependent Claim
19 words
A method as recited in claim 4, wherein said step of detecting a boundary comprises detecting a thoracic boundary.
Claim 7
Dependent Claim
39 words
A method as recited in claim 4, wherein: said step of detecting said boundary of said binary image comprises using 8-point connectivity; and said step of evaluating geometric parameters comprises evaluating a size and a compactness of said region.
Claim 8
Dependent Claim
62 words
A method as recited in claim 6, wherein said step of evaluating geometric parameters further comprises: requiring said size of said region to be at least a predetermined portion of a size of said CT scan and requiring said compactness of said region to be at least a predetermined value in order for said region to be detected as said anatomic region.
Claim 9
Dependent Claim
43 words
A method as recited in claim 2, comprising: detecting a boundary in said binary image; forming a second gray-level histogram of pixels within said boundary; determining a gray-level threshold based on said histogram; generating a second binary image using said second gray-level threshold.
Claim 10
Dependent Claim
29 words
A method as recited in claim 8, comprising: detecting a second boundary in said second binary image; and evaluating geometric parameters of a region enclosed by said second boundary.
Claim 11
Dependent Claim
30 words
A method as recited in claim 9, wherein said step of evaluating geometric parameters comprises: determining an area of said region; and correcting said area using a corner correction technique.
Claim 12
Dependent Claim
51 words
A method as recited in claim 9, wherein said step of evaluating geometric parameters comprises: requiring said region to be located in said first boundary and requiring a compactness of said region to be at least a predetermined value in order for said region to be detected as said anatomic region.
Claim 13
Dependent Claim
23 words
A method as recited in claim 11, wherein: detecting said first boundary comprises detecting a thoracic boundary; and said region is a lung.
Claim 14
Dependent Claim
40 words
A method as recited in claim 1, further comprising: detecting features in said anatomic region as suspected nodules using said plurality of images; determining geometric descriptors for selected ones of said features; and detecting said nodule using said geometric descriptors.
Claim 15
Dependent Claim
35 words
A method as recited in claim 13, wherein said step of determining geometric descriptors comprises: determining plural of perimeter, area, compactness, elongation, circularity, distance measure and total score for said selected ones of said features.
Claim 16
Dependent Claim
30 words
A method as recited in claim 13, wherein said step of detecting said nodule using said geometric descriptors comprises assigning each of said features a likelihood of being a nodule.
Claim 17
Dependent Claim
41 words
A method as recited in claim 15, wherein: said anatomic region is a lung pair; and said step of detecting said nodule using said geometric descriptors comprises assigning each of said features a likelihood of being a nodule or a vessel.
Claim 18
Dependent Claim
35 words
A method as recited in claim 1, further comprising: detecting features in said anatomic region as suspected nodules using said plurality of images; and performing a corner correction routine on selected ones of said features.
Claim 19
Dependent Claim
61 words
A method as recited in claim 17, wherein performing said corner correction routine comprises: detecting a boundary of each of said selected ones of said features; determining chain codes for pixels on said boundary of each of said selected ones of said features; and selectively reducing or expanding an area of said pixels on said border based upon said chain codes.
Claim 20
Dependent Claim
99 words
A method as recited in claim 18, comprising: determining chain codes B1-B7 for a possible next boundary pixel for each pixel A on said boundary of said feature; selectively reducing or expanding an area of each pixel A on said boundary according to said chain codes B1-B7 as: B1--reduce area of pixel by 3/8 B2--reduce area of pixel by 2/8 B3--reduce area of pixel by 1/8 B4--no reduction or expansion B5--expand area of pixel by 1/8 B6--expand area of pixel by 2/8 B7--expand area of pixel by 3/8; and generating a corner correction term based on said chain codes.
Claim 21
Dependent Claim
47 words
A method as recited in claim 19, comprising: said chain codes B1-B7 having direction values of B1=5, B2=6, B3=7, B4=0, B5=1, B6=2 and B7=3; said pixel A having a direction value of 8; and determining Σ(Bi-A) over all pixels on said boundary as said corner correction term.
Claim 22
Dependent Claim
27 words
A method as recited in claim 1, comprising: determining features in anatomic region using said images; and forming a tree structure of features in said binary images.
Claim 23
Dependent Claim
85 words
A method as recited in claim 21, comprising: generating said images using a corresponding plurality of successively stricter gray-level thresholds; wherein said step of forming tree structure comprises: selecting one of said images; detecting a first node within said one image; selecting a second one of said images generated at a next stricter gray-level threshold; detecting a second node within said second image; and determining if a position of said second node lies within a corresponding position of said first node in said second image.
Claim 24
Dependent Claim
37 words
A method as recited in claim 22, comprising: determining a boundary of said first node; and determining if a center of said second node lies within a corresponding boundary of said first node in said second image.
Claim 25
Dependent Claim
31 words
A method as recited in claim 22, comprising: said CT scan comprising a plurality of CT sections; and generating said plurality of images for each of said plurality of CT sections.
Claim 26
Dependent Claim
101 words
A method as recited in claim 1, wherein: said step of detecting a nodule within said anatomic region comprises generating a plurality of binary images of said anatomic region at a corresponding plurality of different gray-level thresholds; said method comprises: forming a histogram of gray values of pixels in said anatomic region; generating said plurality of binary images of said anatomic region at gray-level threshold values corresponding to pixel values at predetermined portions of an area under said histogram; detecting features in said anatomic region as suspected nodules; and detecting a boundary of each feature in each of said binary images.
Claim 27
Dependent Claim
105 words
A method as recited in claim 1, wherein: said CT scan comprises a plurality of CT sections; said step of detecting a nodule comprises generating a plurality of images for each of said CT sections; said step of detecting a nodule comprises: determining features in anatomic region using said images; determining geometric descriptors for selected ones of said features; and using a rule-based system having rules based on said geometric descriptors to assign said selected ones of said features a likelihood of being a nodule; and said method comprises using said rule-based system having rules determined from cumulative frequencies of said features being a nodule.
Claim 28
Dependent Claim
120 words
A method as recited in claim 26, comprising: forming a tree structure defining as nodes said features detected in said plurality of images of said anatomic region; determining plural of perimeter, area compactness, elongation, circularity, distance measure and total score as said geometric descriptors; determining a boundary for said selected ones of said features; wherein said step of using a rule-based system comprises: determining whether one of said nodes has a corresponding node in an adjacent one of said plurality of images and, if said corresponding node is present in said adjacent one image deleting said one node and evaluating geometric descriptors of said corresponding node; evaluating geometric descriptors of said one node if said corresponding node is not present.
Claim 29
Dependent Claim
27 words
A method as recited in claim 27, wherein said step evaluating geometric descriptors of said one node comprises determining whether said one node is said anatomic region.
Claim 30
Dependent Claim
21 words
A method as recited in claim 26, further comprising generating a analyzed CT scan with identified features in said anatomic region.
Claim 31
Dependent Claim
37 words
A method as recited in claim 27, wherein said CT scan comprises a plurality of CT sections; and said method further comprises generating a plurality of analyzed CT sections each having identified features in said anatomic region.
Claim 32
Dependent Claim
110 words
A method as recited in claim 30, wherein said step of comparing comprises: assigning each of said nodes a likelihood of being a nodule; determining which of said analyzed CT sections contain nodes that have not been definitely identified; comparing each of said analyzed CT sections having nodes that have not been definitely identified with adjacent ones of said analyzed CT sections; and updating a likelihood of being a nodule for each node having a corresponding node in said adjacent ones of said analyzed CT sections to a likelihood of being a nodule of said corresponding node when said corresponding node has a more definite likelihood of being a nodule.
Claim 33
Dependent Claim
92 words
A method as recited in claim 31, wherein said step of comparing each of said analyzed CT sections comprises: determining locations of all nodes in one analyzed CT section assigned a likelihood of being a normal anatomic feature; determining nodes having corresponding locations in an analyzed CT section adjacent to said one analyzed CT section which overlap said locations each of said nodes in said one analyzed CT section; updating a likelihood of being a nodule for each of said nodes in said one analyzed CT section based upon corresponding overlapping nodes.
Claim 34
Dependent Claim
63 words
A method as recited in claim 32, wherein said step of updating a likelihood comprises: determining a first count of pixels of a first node in said one analyzed section; determining a second count of pixels of nodes in adjacent analyzed CT sections determined to overlap said first node; updating a likelihood of said first node based upon said first and second counts.
Claim 35
Independent Claim
159 words
A system for the automated detection of lesions in computed tomographic (CT) scan, comprising; an image generation device to generate said CT scan of an anatomic region; a device for generating gray-level images of said CT scan at a plurality of gray-level thresholds; and a device for analyzing features in said gray-level images; wherein said device for generating gray-level threshold images comprises: a histogram generator which generates a histogram of gray-values of pixels in said CT scan; a threshold generator for determining a plurality of gray-level thresholds using said histogram; and a binary image generator for generating a respective plurality of binary images of said anatomic region using said gray-level thresholds; said CT scan comprising a plurality of CT sections; said binary image generator generating a respective plurality of binary images for each of said CT sections; and a tree structure generator producing a tree structure of features in said respective plurality of binary images for each CT section.
Claim 36
Independent Claim
159 words
A system for the automated detection of lesions in computed tomographic (CT) scan, comprising; an image generation device to generate said CT scan of an anatomic region; a device for generating gray-level images of said CT scan at a plurality of gray-level thresholds; and a device for analyzing features in said gray-level images; wherein said device for generating gray-level threshold images comprises: a histogram generator which generates a histogram of gray-values of pixels in said CT scan; a threshold generator for determining a plurality of gray-level thresholds using said histogram; and a binary image generator for generating a respective plurality of binary images of said anatomic region using said gray-level thresholds; said CT scan comprising a plurality of CT sections; said binary image generator generating a respective plurality of binary images for each of said CT sections; and a tree structure generator producing a tree structure of features in said respective plurality of binary images for each CT section.
Claim 37
Dependent Claim
46 words
A system as recited in claim 34, comprising: a feature detector for detecting features using said binary image of said anatomic region; a geometric parameter generator to generate geometric parameters of said features; and a rule-based feature analyzer which analyzes said features using said geometric parameters.
Claim 38
Dependent Claim
71 words
A system as recited in claim 35, comprising: said CT scan comprising a plurality of CT sections; said rule-based feature analyzer generating a plurality of analyzed CT sections having features assigned a status indicating a likelihood of being a nodule; a section comparer for comparing said plurality of CT sections and updating a status of a feature in one section based upon a status of a feature in an adjacent section.
Claim 39
Independent Claim
113 words
A method for the automated detection of nodules in a computed tomographic (CT) scan of a subject, comprising: detecting an anatomic region of said subject in said CT scan; detecting a nodule in said anatomic region using a plurality of images of said anatomic region in said CT scan generated at different gray-level thresholds; detecting features in said anatomic region as suspected nodules using said plurality of images; determining geometric descriptors for selected ones of said features; and detecting said nodule using said geometric descriptors; wherein said step of determining geometric descriptors comprises determining plural of perimeter, area, compactness, elongation, circularity, distance measure and total score for said selected ones of said features.
Claim 40
Independent Claim
113 words
A method for the automated detection of nodules in a computed tomographic (CT) scan of a subject, comprising: detecting an anatomic region of said subject in said CT scan; detecting a nodule in said anatomic region using a plurality of images of said anatomic region in said CT scan generated at different gray-level thresholds; detecting features in said anatomic region as suspected nodules using said plurality of images; determining geometric descriptors for selected ones of said features; and detecting said nodule using said geometric descriptors; wherein said step of determining geometric descriptors comprises determining plural of perimeter, area, compactness, elongation, circularity, distance measure and total score for said selected ones of said features.
Claim 41
Independent Claim
108 words
A method for the automated detection of nodules in a computed tomographic (CT) scan of a subject, comprising: detecting an anatomic region of said subject in said CT scan; detecting a nodule in said anatomic region using a plurality of images of said anatomic region in said CT scan generated at different gray-level thresholds; detecting features in said anatomic region as suspected nodules using said plurality of images; determining geometric descriptors for selected ones of said features; and detecting said nodule using said geometric descriptors; wherein said step of detecting said nodule using said geometric descriptors comprises assigning each of said features a likelihood of being a nodule.
Claim 42
Independent Claim
108 words
A method for the automated detection of nodules in a computed tomographic (CT) scan of a subject, comprising: detecting an anatomic region of said subject in said CT scan; detecting a nodule in said anatomic region using a plurality of images of said anatomic region in said CT scan generated at different gray-level thresholds; detecting features in said anatomic region as suspected nodules using said plurality of images; determining geometric descriptors for selected ones of said features; and detecting said nodule using said geometric descriptors; wherein said step of detecting said nodule using said geometric descriptors comprises assigning each of said features a likelihood of being a nodule.
Claim 43
Independent Claim
80 words
A method for the automated detection of nodules in a computed tomographic (CT) scan of a subject, comprising: detecting an anatomic region of said subject in said CT scan; detecting a nodule in said anatomic region using a plurality of images of said anatomic region in said CT scan generated at different gray-level thresholds; detecting features in said anatomic region as suspected nodules using said plurality of images; and performing a corner correction routine on selected ones of said features.
Claim 44
Independent Claim
80 words
A method for the automated detection of nodules in a computed tomographic (CT) scan of a subject, comprising: detecting an anatomic region of said subject in said CT scan; detecting a nodule in said anatomic region using a plurality of images of said anatomic region in said CT scan generated at different gray-level thresholds; detecting features in said anatomic region as suspected nodules using said plurality of images; and performing a corner correction routine on selected ones of said features.
Claim 45
Independent Claim
73 words
A method for the automated detection of nodules in a computed tomographic (CT) scan of a subject, comprising: detecting an anatomic region of said subject in said CT scan; detecting a nodule in said anatomic region using a plurality of images of said anatomic region in said CT scan generated at different gray-level thresholds; determining features in anatomic region using said images; and forming a tree structure of features in said binary images.
Claim 46
Independent Claim
73 words
A method for the automated detection of nodules in a computed tomographic (CT) scan of a subject, comprising: detecting an anatomic region of said subject in said CT scan; detecting a nodule in said anatomic region using a plurality of images of said anatomic region in said CT scan generated at different gray-level thresholds; determining features in anatomic region using said images; and forming a tree structure of features in said binary images.
Claim 47
Dependent Claim
85 words
A method as recited in claim 40, comprising: generating said images using a corresponding plurality of successively stricter gray-level thresholds; wherein said step of forming tree structure comprises: selecting one of said images; detecting a first node within said one image; selecting a second one of said images generated at a next stricter gray-level threshold; detecting a second node within said second image; and determining if a position of said second node lies within a corresponding position of said first node in said second image.
Claim 48
Dependent Claim
37 words
A method as recited in claim 41, comprising: determining a boundary of said first node; and determining if a center of said second node lies within a corresponding boundary of said first node in said second image.
Claim 49
Dependent Claim
31 words
A method as recited in claim 41, comprising: said CT scan comprising a plurality of CT sections; and generating said plurality of images for each of said plurality of CT sections.
References Cited (Backward Citations)
US4777598A
Cited
US5146923A
Cited
US4907156A
Cited
US5185809A
Cited
US5133020A
Cited
US5319549A
Cited
WO1995015536A1
Cited
USPTO Annuity & Maintenance Windows
| Annuity Stage | Window Opens | Payment Deadline | Surcharge Period Until | Estimated Fee | Status |
|---|---|---|---|---|---|
| 3.5 Years (Year 4) | 2002-03-08 | 2002-09-07 | 2003-03-09 | $2000.00 | Window Closed / Lapsed |
| 7.5 Years (Year 8) | 2006-03-08 | 2006-09-07 | 2007-03-09 | $3760.00 | Window Closed / Lapsed |
| 11.5 Years (Year 12) | 2010-03-08 | 2010-09-07 | 2011-03-09 | $7700.00 | Window Closed / Lapsed |