3D cell nuclei segmentation based on gradient flow tracking
© Li et al; licensee BioMed Central Ltd. 2007
Received: 17 January 2007
Accepted: 04 September 2007
Published: 04 September 2007
Skip to main content
© Li et al; licensee BioMed Central Ltd. 2007
Received: 17 January 2007
Accepted: 04 September 2007
Published: 04 September 2007
Reliable segmentation of cell nuclei from three dimensional (3D) microscopic images is an important task in many biological studies. We present a novel, fully automated method for the segmentation of cell nuclei from 3D microscopic images. It was designed specifically to segment nuclei in images where the nuclei are closely juxtaposed or touching each other. The segmentation approach has three stages: 1) a gradient diffusion procedure, 2) gradient flow tracking and grouping, and 3) local adaptive thresholding.
Both qualitative and quantitative results on synthesized and original 3D images are provided to demonstrate the performance and generality of the proposed method. Both the over-segmentation and under-segmentation percentages of the proposed method are around 5%. The volume overlap, compared to expert manual segmentation, is consistently over 90%.
The proposed algorithm is able to segment closely juxtaposed or touching cell nuclei obtained from 3D microscopy imaging with reasonable accuracy.
Reliable segmentation of cell nuclei from three dimensional (3D) microscopic images is an important task in many biological studies as it is required for any subsequent comparison or classification of the nuclei. For example, zebrafish somitogenesis is governed by a clock that generates oscillations in gene expression within the presomitic mesoderm [1, 2]. The subcellular localization of oscillating mRNA in each nucleus, imaged through multi-channel microscopy, can be used to identify different phases within the oscillation. To automate the classification of the phase of an individual nucleus, each nucleus within the presomitic mesoderm first needs to be accurately segmented.
In recent years, there has been significant effort towards the development of automated methods for 3D cell or cell nuclei image segmentation [3–9, 15, 16]. Thresholding, watershed and active surface based methods are among the most commonly used techniques for 3D cell or cell nuclei segmentation. Unfortunately, thresholding-based methods often have difficulties in dealing with images that do not have a well-defined constant contrast between the objects and the background. Given this characteristic of the thresholding-based methods, they often have difficulties in segmenting images with clustered or juxtaposed nuclei. Watershed-based methods are also very popular for segmentation of clustered cell nuclei [3–5, 10]. However, these methods often result in the over-segmentation of clustered cell nuclei. In order to deal with this issue, heuristic rules have been developed for region merging [3–5] as a post-processing step. Segmentation problems have also been targeted through the use of active surface-based methods [8, 9, 15, 16] in the literature. However, such algorithms suffer from an inherent dependency on the initial guess. If the initial guess is wrong, these methods have difficulties in dealing with clustered cell nuclei.
Despite active research and progress in the literature, development of a fully automated and robust computational algorithm for 3D cell nuclei segmentation still remains a challenge when dealing with significant inherent nuclei shape and size variations in image data. Examples include cases where the contrast between nuclei and background is low, where there are differences in shapes and sizes of nuclei, and where we are dealing with 3D images of low quality [3, 4, 6–8]. Complications also arise when nuclei are juxtaposed or connected to one another, increasing the rate of over-segmentation or under-segmentation.
In this paper, we present a novel automated method that aims to tackle the aforementioned challenges of segmentation of clustered or connected 3D cell nuclei. We approach the segmentation problem by first generating the gradient vector field corresponding to the 3D volume image, and then diffusing the gradient vector field with an elastic deformable transform. After the elastic deformable transform is completed, the noisy gradient vector field is smoothed and the gradient vectors with large magnitude are propagated to the areas with weak gradient vectors. This gradient diffusion procedure results in a gradient flow field, in which the gradient vectors are smoothly flowing towards or outwards from the centers of the nuclei. Subsequently, a gradient flow tracking procedure is performed from each vector point to find the corresponding center to which the points flow. We group all points that flow to the same center into a region, and refer to this region as the attraction basin of the center. Once we have completed the process of tracking the gradient flow, the boundaries of juxtaposed nuclei are formed naturally and hence these juxtaposed nuclei are divided. The final step includes performing local thresholding in each attraction basin in order to extract the nuclei from their corresponding background. We have evaluated and validated this algorithm and have presented results attesting its validity.
In this section, a series of experiments are designed to evaluate and validate the gradient flow tracking method for segmentation of 3D images with juxtaposed nuclei. Both qualitative and quantitative results on synthesized and original 3D images are provided to demonstrate the performance and general applicability of the proposed method.
where R a is the automated extracted region and R g is the ground truth region. The ⋂ operator takes the intersection of two regions. S(·) is the volume of the region.
We have synthesized seven 3D nuclei images. In order to present a quantitative measure for the validation of our results, the average value of volume overlap measurement for the seven cases is around 0.971, and the standard deviation is 0.014. As it is clear from these results, our proposed segmentation method achieves significant volume overlap with the ground truth, indicating the accurate performance of the gradient flow tracking method.
The segmentation results of 3D C. elegans embryo images
The segmentation results of 3D cell nuclei image of zebrafish
Our method has several advantages over previous approaches. The major advantage of the method is the ability to robustly segment densely packed, touching, or connected nuclei. Additionally, no sophisticated rules are used. The only assumption is that the centers of nuclei are brighter or darker than a nearby region. The fundamental difference between our method and existing methods lies in the diffused gradient vector information. In existing methods such as the threshold or watershed methods, intensity is the only adopted information, hence those methods are sensitive to the noise in the image, which usually results in over-segmentation. In contrast, in our method the gradient vector diffusion procedure propagates gradient vectors with large magnitudes to the areas with weak gradient vectors and smoothes the noisy gradient vector field. Meanwhile, it preserves the potential structural information of the gradient vector field. For example when two nuclei are touching each other, the diffused gradient vectors point toward the corresponding centers of the nuclei. This step greatly contributes to the success of touching nuclei segmentation. The disadvantage of this method is that it may have difficulty in processing the images of textured blob objects, since in that situation the gradient vector at the centers of nuclei are cluttered and the condition is violated. Currently the method is implemented using the C/C++ language, without using any other common library. Without any optimization, it takes less than 50 seconds on an Intel Pentium4 2.4 GHz machine with 1 GB memory to segment a volume image with a size of 230*162*80. The running time can be reduced further with multi-resolution implementation and code optimisation. After evaluating this method on larger and more diverse image datasets, we intend to release the algorithm to the cell biology community.
We presented a novel, automated algorithm for 3D cell nuclei segmentation based on gradient flow tracking. To validate the efficacy and performance of the proposed segmentation algorithm, we evaluated it by using synthesized and real biological images. The results show that the algorithm is able to segment juxtaposed nuclei correctly, a persistent problem in the field of cellular image analysis.
Gradient information is an important factor in three-dimensional (3D) nuclei segmentation due to the fact that in any given nuclei image, the gradient vectors either point towards the central area of a bright nucleus, or outwards from the central area of a dark nucleus. However, in practice, the gradient magnitude is very small, and the direction of the gradient vector is usually not trustworthy due to the noise present in the image when approaching the central area of a nucleus. Additionally, when we are dealing with nuclei that are of irregular shapes, the gradient vectors tend to be cluttered. Motivated by these facts, here, we introduce a physical model that incorporates the diffused gradient vectors from the boundaries of the image to generate a smooth gradient field. Our gradient vector diffusion procedure propagates gradient vectors with large magnitudes to the areas with weak gradient vectors and smoothes the noisy gradient vector field . For a detailed introduction to gradient vector diffusion, we refer to . We adopt an elastic deformation transformation, under which the image is modeled as elastic sheets warped by an external force field to achieve gradient vector diffusion. This model has been previously employed for image registration [12, 13], where the deformation of boundary points are fixed and then the deformation field is propagated to inner regions of the image by solving the elastic model equation. Here, we extend this model to analyze 3D microscopic nuclei images.
The diffused gradient vector field v(x, y, z) = (u(x, y, z), v(x, y, z), w(x, y, z)) (u(x, y, z), v(x, y, z) and w(x, y, z) are three components of the diffused gradient vector projecting to x, y and z axis respectively) in a 3D image is defined to be a solution to the partial differential equation (PDE), also known as a Navier-Stokes equation, describing the deformation of an elastic sheet :
μ∇2v + (λ + μ)∇div(v) + q × (∇f - v) = 0, (1)
where ∇2 is the Laplacian operator, div is the divergence operator, ∇ is the gradient operator, ∇f is the original gradient vector field, and Lame's coefficients μ and λ refer to the elastic properties of the material. In this paper, we aim to diffuse the gradient vectors toward the central areas of nuclei objects to obtain a gradient flow field. Therefore, f is set to be
f (x, y, z) = Gσ (x, y, z)*I(x, y, z),
where v t (x, y, z, t) denotes the partial derivative of v(x, y, z, t) with respect to time t. The equation is decoupled as:
u t (x, y, z, t) = μ∇2u(x, y, z, t) + (λ + μ) (∇div(v(x, y, z, t))) x + q(x, y, z)((∇f(x, y, z)) x - u(x, y, z, t))
v t (x, y, z, t) = μ∇2v(x, y, z, t) + (λ + μ) (∇div(v(x, y, z, t))) y + q(x, y, z)((∇f(x, y, z)) y - v(x, y, z, t))
w t (x, y, z, t) = μ∇2w(x, y, z, t) + (λ + μ) (∇div(v(x, y, z, t))) z + q(x, y, z)((∇f(x, y, z)) z - w(x, y, z, t))
The algorithm of the gradient flow tracking is summarized as follows.
1. Randomly select a point x as the initial point x0.
2. Obtain xn + 1(n = 0,1,2...) using Equation 3 based on x n .
3. Compute the angle θ n of diffused gradient vector between xn + 1and x n with Equation 4. If θ n is larger than , stop.
4. Replace x n with xn + 1. Return to step 2.
Gradient flow tracking is applied to each point in the image. All points in the same attraction basin are grouped into the same cluster. Since it is time consuming to run the tracking algorithm for every point, in order to improve the performance of our method, gradient flow tracking is not applied to the points that have already been on the gradient flow trajectory of a previously processed pixel. Instead, these visited points are directly associated with the sink to which the path flows. This improvement not only speeds up the segmentation, but also yields reproducible segmentation results.
After the gradient flow tracking step, the image is segmented into smaller regions each of which is expected to contain only a single nucleus. From here the nuclei segmentation problem is turned into binary classification problem where we are interested in distinguishing the nuclei from their background in a small region. Therefore an intensity thresholding method is appropriate for extracting the nuclei from the background. In order to approach this problem, we can take advantage of the method employed by Otsu in , which has the ability to extract the nucleus from each attraction basin. Another approach for dealing with this problem is through designing a more involved method that employs techniques such as graph cut, level set, etc. Here, we employ the locally adaptive method of Otsu  because of its ability to deal with situations where the intensity of nuclei and background are not constant across an image. In each segmented region, pixels whose intensities are larger than the automatically determined local Otsu threshold are grouped as nuclei, otherwise they are grouped as background. Finally, an optional procedure is performed after extracting the nuclei to eliminate small regions, which contain a lower number of pixels than a threshold.
The algorithm of the 3D cell nuclei segmentation method based on gradient flow tracking is summarized as follows.
1. Obtain the diffuse gradient vector field using the elastic deformation transformation
2. From each point, run the gradient flow tracking procedure, and label each passed pixel with a converged sink position.
3. Combine the attraction basins of the sinks whose distance is less than three pixels.
4. Assign the same label to the points in the same attraction basin.
5. Perform local adaptive thresholding in each attraction basin to extract the nucleus.
6. Optional procedure: eliminate regions with smaller number of pixels than a threshold T.
This work is funded by a Bioinformatics Research Center Program Grant from Harvard Center for Neurodegeneration and Repair, Harvard Medical School (STCW). We would like to thank Dr. Sean Megason of California Institute of Technology for providing the C. elegans embryo images.
This article is published under license to BioMed Central Ltd. This is an Open Access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/2.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.