Kidney glomeruli measurement systems and methods
View Patent ↗Methods and systems for identifying blobs, for example kidney glomeruli, are disclosed. A raw image may be smoothed via a difference of Gaussians filter, and a Hessian analysis may be conducted on the smoothed image to mark glomeruli candidates. Exemplary candidate features are identified, such as average intensity A T , likelihood of blobness R T , and flatness S T . A clustering algorithm may be utilized to post prune the glomeruli candidates.
1. A method of measuring glomeruli in a kidney, the method comprising:
obtaining a 3 dimensional image of a kidney;
smoothing the image via a Difference of Gaussians filter;
conducting a Hessian analysis on the smoothed image to mark glomeruli candidates;
identifying, for each glomeruli candidate, average intensity A T , likelihood of blobness R T , and flatness S T ; and
executing a clustering algorithm to post prune the glomeruli candidates, wherein the clustering algorithm evaluates, for each glomeruli candidate, the average intensity A T , the likelihood of blobness R T , and the flatness S T to determine if a glomeruli candidate should be counted as a true glomeruli,
wherein
R
T
=
3
×
λ
1
′
λ
2
′
λ
3
′
2
3
λ
1
′
λ
2
′
+
λ
2
′
λ
3
′
+
λ
1
′
λ
3
′
,
and
wherein λ 1 ′, λ 2 ′, λ 3 ′ are eigenvalues of a regional Hessian H T .
2. The method of claim 1 , wherein the clustering algorithm is a Variational Bayesian Gaussian Mixture Model.
3. The method of claim 2 , wherein the Variational Bayesian Gaussian Mixture Model automatically identifies a number of clusters for optimum performance of the clustering algorithm without the need for initialization and subjective parameter settings.
4. The method of claim 1 , wherein the 3 dimensional image is a magnetic resonance imaging (MRI) image.
5. The method of claim 1 , further comprising utilizing the output of the clustering algorithm to evaluate the susceptibility of an individual to chronic kidney and/or cardiovascular disease.
6. The method of claim 1 , wherein the Hessian analysis delineates the boundary of each glomeruli candidate.
7. The method of claim 1 , wherein
S
T
=
(
λ
1
′
+
λ
2
′
+
λ
3
′
)
2
-
2
(
λ
1
′
λ
2
′
+
λ
2
′
λ
3
′
+
λ
1
′
λ
3
′
)
.
8. The method of claim 1 , further comprising extracting features of the individual glomeruli candidates.
9. The method of claim 8 , wherein the features comprise at least one of average intensity, divergence, distance to kidney boundary, region volume, shape index, or Laplacian of Gaussian (LoG).
10. The method of claim 1 , wherein the identifying, for each glomeruli candidate, the likelihood of blobness R T and the flatness S T utilizes a calculation of trace and determinant of the regional Hessian H T .