Limits...
Automated detection of soma location and morphology in neuronal network cultures.

Ozcan B, Negi P, Laezza F, Papadakis M, Labate D - PLoS ONE (2015)

Bottom Line: In addition to the usual difficulties associated with the processing of fluorescent images, these types of stacks contain a small number of images so that only a small number of pixels are available along the z-direction and it is challenging to apply conventional 3D filters.Extensive numerical experiments show that our algorithms are computationally efficient and highly accurate in segmenting the somas and separating contiguous ones.The algorithms presented in this paper will facilitate the development of a high-throughput quantitative platform for the study of neuronal networks for HCS applications.

View Article: PubMed Central - PubMed

Affiliation: Dept. of Mathematics, University of Houston, Houston, Texas, United States of America.

ABSTRACT
Automated identification of the primary components of a neuron and extraction of its sub-cellular features are essential steps in many quantitative studies of neuronal networks. The focus of this paper is the development of an algorithm for the automated detection of the location and morphology of somas in confocal images of neuronal network cultures. This problem is motivated by applications in high-content screenings (HCS), where the extraction of multiple morphological features of neurons on large data sets is required. Existing algorithms are not very efficient when applied to the analysis of confocal image stacks of neuronal cultures. In addition to the usual difficulties associated with the processing of fluorescent images, these types of stacks contain a small number of images so that only a small number of pixels are available along the z-direction and it is challenging to apply conventional 3D filters. The algorithm we present in this paper applies a number of innovative ideas from the theory of directional multiscale representations and involves the following steps: (i) image segmentation based on support vector machines with specially designed multiscale filters; (ii) soma extraction and separation of contiguous somas, using a combination of level set method and directional multiscale filters. We also present an approach to extract the soma's surface morphology using the 3D shearlet transform. Extensive numerical experiments show that our algorithms are computationally efficient and highly accurate in segmenting the somas and separating contiguous ones. The algorithms presented in this paper will facilitate the development of a high-throughput quantitative platform for the study of neuronal networks for HCS applications.

No MeSH data available.


Related in: MedlinePlus

Illustration of 2D soma detection algorithm.(A) Denoised image, obtained using a shearlet-based denoising routine on the MIP of the image stack. Image size = 512 × 512 pixels (1 pixel = 0.28 × 0.28μm). (B) Segmented binary image. (C) Directional Ratio plot; the values range between 1, in red color (corresponding to more isotropic regions), and 0, in blue color (corresponding to more anisotripc regions); note that the Directional Ratio is only computed inside the segmented region, i.e., inside the red region in Panel B. (D) Detection of initial soma region, obtained by applying a threshold to the values of the Directional Ratio in Panel C. (E) Soma detection, obtained by applying the level set method with the initialization curve determined by the boundary of the initial soma region in Panel D. (F) Separation of contiguous somas; two regions from Panel E are recognized as too large and hence divided using the level set method.
© Copyright Policy
Related In: Results  -  Collection

License
getmorefigures.php?uid=PMC4390318&req=5

pone.0121886.g010: Illustration of 2D soma detection algorithm.(A) Denoised image, obtained using a shearlet-based denoising routine on the MIP of the image stack. Image size = 512 × 512 pixels (1 pixel = 0.28 × 0.28μm). (B) Segmented binary image. (C) Directional Ratio plot; the values range between 1, in red color (corresponding to more isotropic regions), and 0, in blue color (corresponding to more anisotripc regions); note that the Directional Ratio is only computed inside the segmented region, i.e., inside the red region in Panel B. (D) Detection of initial soma region, obtained by applying a threshold to the values of the Directional Ratio in Panel C. (E) Soma detection, obtained by applying the level set method with the initialization curve determined by the boundary of the initial soma region in Panel D. (F) Separation of contiguous somas; two regions from Panel E are recognized as too large and hence divided using the level set method.

Mentions: We considered ten image stacks with different sizes to test our 2D algorithm for soma detection. The stacks we considered comprise between 10 and 25 images each and contain between 2 and 8 neurons. From each stack, we generated the MIP images and then processed the resulting 2D images using the algorithm described above. Fig. 10 illustrates the various steps of our algorithm on a representative MIP image from this set of ten image stacks. In particular, it shows the pre-processing, segmentation and soma detection on an images of size 512 × 512 pixels containing eight somas. The figure also illustrates the capability of the algorithm to separate contiguous somas, using the method of Directional ratio described above.


Automated detection of soma location and morphology in neuronal network cultures.

Ozcan B, Negi P, Laezza F, Papadakis M, Labate D - PLoS ONE (2015)

Illustration of 2D soma detection algorithm.(A) Denoised image, obtained using a shearlet-based denoising routine on the MIP of the image stack. Image size = 512 × 512 pixels (1 pixel = 0.28 × 0.28μm). (B) Segmented binary image. (C) Directional Ratio plot; the values range between 1, in red color (corresponding to more isotropic regions), and 0, in blue color (corresponding to more anisotripc regions); note that the Directional Ratio is only computed inside the segmented region, i.e., inside the red region in Panel B. (D) Detection of initial soma region, obtained by applying a threshold to the values of the Directional Ratio in Panel C. (E) Soma detection, obtained by applying the level set method with the initialization curve determined by the boundary of the initial soma region in Panel D. (F) Separation of contiguous somas; two regions from Panel E are recognized as too large and hence divided using the level set method.
© Copyright Policy
Related In: Results  -  Collection

License
Show All Figures
getmorefigures.php?uid=PMC4390318&req=5

pone.0121886.g010: Illustration of 2D soma detection algorithm.(A) Denoised image, obtained using a shearlet-based denoising routine on the MIP of the image stack. Image size = 512 × 512 pixels (1 pixel = 0.28 × 0.28μm). (B) Segmented binary image. (C) Directional Ratio plot; the values range between 1, in red color (corresponding to more isotropic regions), and 0, in blue color (corresponding to more anisotripc regions); note that the Directional Ratio is only computed inside the segmented region, i.e., inside the red region in Panel B. (D) Detection of initial soma region, obtained by applying a threshold to the values of the Directional Ratio in Panel C. (E) Soma detection, obtained by applying the level set method with the initialization curve determined by the boundary of the initial soma region in Panel D. (F) Separation of contiguous somas; two regions from Panel E are recognized as too large and hence divided using the level set method.
Mentions: We considered ten image stacks with different sizes to test our 2D algorithm for soma detection. The stacks we considered comprise between 10 and 25 images each and contain between 2 and 8 neurons. From each stack, we generated the MIP images and then processed the resulting 2D images using the algorithm described above. Fig. 10 illustrates the various steps of our algorithm on a representative MIP image from this set of ten image stacks. In particular, it shows the pre-processing, segmentation and soma detection on an images of size 512 × 512 pixels containing eight somas. The figure also illustrates the capability of the algorithm to separate contiguous somas, using the method of Directional ratio described above.

Bottom Line: In addition to the usual difficulties associated with the processing of fluorescent images, these types of stacks contain a small number of images so that only a small number of pixels are available along the z-direction and it is challenging to apply conventional 3D filters.Extensive numerical experiments show that our algorithms are computationally efficient and highly accurate in segmenting the somas and separating contiguous ones.The algorithms presented in this paper will facilitate the development of a high-throughput quantitative platform for the study of neuronal networks for HCS applications.

View Article: PubMed Central - PubMed

Affiliation: Dept. of Mathematics, University of Houston, Houston, Texas, United States of America.

ABSTRACT
Automated identification of the primary components of a neuron and extraction of its sub-cellular features are essential steps in many quantitative studies of neuronal networks. The focus of this paper is the development of an algorithm for the automated detection of the location and morphology of somas in confocal images of neuronal network cultures. This problem is motivated by applications in high-content screenings (HCS), where the extraction of multiple morphological features of neurons on large data sets is required. Existing algorithms are not very efficient when applied to the analysis of confocal image stacks of neuronal cultures. In addition to the usual difficulties associated with the processing of fluorescent images, these types of stacks contain a small number of images so that only a small number of pixels are available along the z-direction and it is challenging to apply conventional 3D filters. The algorithm we present in this paper applies a number of innovative ideas from the theory of directional multiscale representations and involves the following steps: (i) image segmentation based on support vector machines with specially designed multiscale filters; (ii) soma extraction and separation of contiguous somas, using a combination of level set method and directional multiscale filters. We also present an approach to extract the soma's surface morphology using the 3D shearlet transform. Extensive numerical experiments show that our algorithms are computationally efficient and highly accurate in segmenting the somas and separating contiguous ones. The algorithms presented in this paper will facilitate the development of a high-throughput quantitative platform for the study of neuronal networks for HCS applications.

No MeSH data available.


Related in: MedlinePlus