Image analysis – Applications – Biomedical applications
Reexamination Certificate
1999-12-17
2001-05-01
Bella, Matthew C. (Department: 2621)
Image analysis
Applications
Biomedical applications
C382S264000, C382S132000, C600S407000, C378S018000, C514S451000
Reexamination Certificate
active
06226393
ABSTRACT:
This application is the national phase under 35 U.S.C. §371 of PCT International Application No. PCT/DK98/00311 which has an International filing date of Jul. 6, 1998, which designated the United States of America.
The present invention relates to a method for estimating aspects of the bone quality or skeletal status of a vertebrate on the basis of two-dimensional image data comprising information relating to the cortical structure of at least a part of a bone, typically a long or tubular bone. Especially bones comprising a tube (diaphysis) and at least at one end an epiphysis comprising trabecular bone are suitable for use in the present method.
In the present context, skeletal status is taken to relate to e.g. biomechanical competence, cortical homogeneity, cortical resorption (intra-cortical and endosteal resorption), cortical thickness of the bone or a quantification of sensitivity, specificity or accuracy of diagnosis of osteoporosis.
Thus, in a first aspect, the invention relates to a method for estimating the bone quality or skeletal status of a vertebrate on the basis of two-dimensional image data comprising information relating to the cortical bone of at least a part of a bone of the vertebrate, the image data being data obtained by exposing at least the part of the bone to electromagnetic radiation, the method comprising:
determining at least one value representing a variation of a density of cortical bone of the at least part of the bone,
an estimation procedure in which the bone quality or skeletal status of the vertebrate is estimated on the basis of the at least one variation value, and optionally one or more features related to the bone of the vertebrate, and a predetermined relationship between the at least one variation value and optional feature(s) and reference skeletal status or bone quality parameters.
As indicated above, in the present context “skeletal status” is not simply equalled to “bone quantity”, such as Bone Mineral Density, as a loss of cortical tissue (cortical resorption) or cortical inhomogeneity may lead to a disproportionate loss of biomechanical competence.
In fact, it is contemplated that a better estimate of skeletal status or bone quality may be obtained by considering pertinent properties of cortical and cancellous bone individually instead of treating these two types of bones similarly, as it is done in standard densitometric methods.
Normally, at least one variation value will represent a variation in values reflecting a projected cortical density (PCD) in the image data, caused by the X-ray attenuating properties of cortical bone in the part of the bone.
Also, the image data will typically comprise discrete elements (pixels) each relating to an individual part of the bone and each having a value relating to a projected cortical density in the image data at the individual part of the bone.
In one embodiment, one or more of the at least one variation value is/are determined by
(a) determining values reflecting the projected cortical density (PCD) in the image data, caused by the X-ray attenuating properties of cortical bone in the part of the bone, for each of a number of locations or areas in the image data,
(b) deriving the one or more variation values from a variation of the determined PCD-values, preferably in a longitudinal direction of the bone.
Even though it is contemplated that the variation of PCD may be used in the present embodiment for substantially all directions in the bone, especially that of the longitudinal direction is found interesting, and especially in long bones, such as the tibia, ulna, and the distal radius, where strong relations to other measures of skeletal status or bone quality have been found.
In fact, it is contemplated that a large number of variation values derived from this variation of PCD in the longitudinal direction of the bone will provide a significant correlation with the failure load of e.g. the hip.
In order to facilitate the determination of the Regions Of Interest used in the bones and to enhance repeatability of the method, it is preferred that the PCD-values are determined at locations or areas situated longitudinally spaced apart within the bone. It is currently preferred that the locations or areas for which the PCD-values are determined are located at predetermined areas or positions in the image, such as along a reproducibly positionable curve in the image, preferably a substantially straight line substantially along a longitudinal axis of the bone.
The PCD-values may be determined in a number of ways, and they may be chosen to relate to different properties of the projected cortical structure. At present, the PCD-value determined for each of the locations or areas may be a value relating to first, second, third or higher order properties of the image data relating to the projected cortical structure at the location or in the area in question.
Preferably, the PCD-values are calculated on the basis of a determination of any property of the Fourier power spectrum, a parametric spectral estimate or of the grey level co-occurrence matrix of information in the image data at the location or in the area in question, the variance of the image data at the location or in the area in question, or a mean value of the size and density of edges in the image data at the location or in the area in question.
Preferably, a PCD-profile is defined by the determined PCD values and a (preferably reproducible) measure relating to the positions of the areas or positions in relation to each other or in relation to a fixed position within the bone, such as, the determined PCD-values as a function of eg a distance from the location or area for which the individual PCD-value was determined to a predetermined reference point. In this situation, at least one of the above one or more variation values is derived from the profile. Values from this profile have been found to correlate especially well with parameters of skeletal status or bone quality. Preferably, the predetermined reference point is also located on the line along which the PCD values are extracted.
Thus, the PCD-profile preferably describes the PCD as a function of the distance between the area or position at which the individual value was derived and the predetermined reference point.
This profile has the advantage that reproducible information may be derived from all bones independently of the size and shape of the bone in question. Moreover, this type of measurement gives a global description of the skeletal status or bone quality as opposed to measuring in a region of interest chosen more or less ad hoc.
It is preferred that the locations or areas for which the PCD-values are determined are situated at locations or areas within the bone at which the image data show substantially no projected trabecular structure.
The fluctuation or variation of the PCD-profile is of interest. Therefore, one or more variation values are preferably derived from this fluctuation or variation. Such a value may be derived from a polynomial fitted to the profile or from the standard deviation of the difference between the polynomial and the profile or a smoothed version thereof. In these situations, the polynomial is preferably of an order in the interval 1-10, such as 2-5, preferably 3-4.
An optional or additional variation value may be derived from the standard deviation of a difference between the profile or a smoothed version thereof and a morphological opening or closing of the profile or of a smoothed version of the morphological opening or closing of the profile.
In another embodiment, one or more of the variation values is/are determined as a variation within a predetermined Region Of Interest in the image data. Thus, instead of using a number of different locations or areas, the information from within a single area may be used in a calculation of one or more variation values all relating to the same area.
However, no matter whether the variation is determined from a number of positions or areas or from a single Region Of Interest, the calculations in the regi
Grunkin Michael
Rosholm Anders
Sadegh Payman
Bella Matthew C.
Choobin M B
Torsana Osteoporosis Diagnostics A/S
LandOfFree
Method for estimating the bone quality or skeletal status of... does not yet have a rating. At this time, there are no reviews or comments for this patent.
If you have personal experience with Method for estimating the bone quality or skeletal status of..., we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Method for estimating the bone quality or skeletal status of... will most certainly appreciate the feedback.
Profile ID: LFUS-PAI-O-2478273