Automated segmentation of retinal blood vessels in spectral domain optical coherence tomography scans.  
Jump to Full Text  
MedLine Citation:

PMID: 22808422 Owner: NLM Status: PubMednotMEDLINE 
Abstract/OtherAbstract:

The correct segmentation of blood vessels in optical coherence tomography (OCT) images may be an important requirement for the analysis of intraretinal layer thickness in human retinal diseases. We developed a shape model based procedure for the automatic segmentation of retinal blood vessels in spectral domain (SD)OCT scans acquired with the Spectralis OCT system. The segmentation procedure is based on a statistical shape model that has been created through manual segmentation of vessels in a training phase. The actual segmentation procedure is performed after the approximate vessel position has been defined by a shadowgraph that assigns the lateral vessel positions. The active shape model method is subsequently used to segment blood vessel contours in axial direction. The automated segmentation results were validated against the manual segmentation of the same vessels by three expert readers. Manual and automated segmentations of 168 blood vessels from 34 Bscans were analyzed with respect to the deviations in the mean Euclidean distance and surface area. The mean Euclidean distance between the automatically and manually segmented contours (on average 4.0 pixels respectively 20 µm against all three experts) was within the range of the manually marked contours among the three readers (approximately 3.8 pixels respectively 18 µm for all experts). The area deviations between the automated and manual segmentation also lie within the range of the area deviations among the 3 clinical experts. Intra reader variability for the experts was between 0.9 and 0.94. We conclude that the automated segmentation approach is able to segment blood vessels with comparable accuracy as expert readers and will provide a useful tool in vessel analysis of whole Cscans, and in particular in multicenter trials. 
Authors:

Matthäus Pilch; Yaroslava Wenner; Elisabeth Strohmayr; Markus Preising; Christoph Friedburg; Erdmuthe Meyer Zu Bexten; Birgit Lorenz; Knut Stieger 
Publication Detail:

Type: Journal Article Date: 20120604 
Journal Detail:

Title: Biomedical optics express Volume: 3 ISSN: 21567085 ISO Abbreviation: Biomed Opt Express Publication Date: 2012 Jul 
Date Detail:

Created Date: 20120718 Completed Date: 20120823 Revised Date: 20141030 
Medline Journal Info:

Nlm Unique ID: 101540630 Medline TA: Biomed Opt Express Country: United States 
Other Details:

Languages: eng Pagination: 147891 Citation Subset:  
Export Citation:

APA/MLA Format Download EndNote Download BibTex 
MeSH Terms  
Descriptor/Qualifier:


Comments/Corrections 
Full Text  
Journal Information Journal ID (nlmta): Biomed Opt Express Journal ID (isoabbrev): Biomed Opt Express Journal ID (publisherid): BOE ISSN: 21567085 Publisher: Optical Society of America 
Article Information Download PDF ©2012 Optical Society of America openaccess: Received Day: 24 Month: 2 Year: 2012 Revision Received Day: 25 Month: 5 Year: 2012 Accepted Day: 28 Month: 5 Year: 2012 Electronic publication date: Day: 04 Month: 6 Year: 2012 collection publication date: Day: 01 Month: 7 Year: 2012 pmcrelease publication date: Day: 04 Month: 6 Year: 2012 Volume: 3 Issue: 7 First Page: 1478 Last Page: 1491 ID: 3395475 PubMed Id: 22808422 Publisher Id: 163622 DOI: 10.1364/BOE.3.001478 
Automated segmentation of retinal blood vessels in spectral domain optical coherence tomography scans  
Matthäus Pilch^{1}^{2}  
Yaroslava Wenner^{1}  
Elisabeth Strohmayr^{1}  
Markus Preising^{1}  
Christoph Friedburg^{1}  
Erdmuthe Meyer zu Bexten^{2}  
Birgit Lorenz^{1}  
Knut Stieger^{1}*  
1Department of Ophthalmology, JustusLiebigUniversity Giessen, Germany 

2Department of Informatics, University of Applied Sciences Giessen, Germany 

*knut.stieger@uniklinikumgiessen.de 
Optical coherence tomography (OCT) has revolutionized the in vivo imaging capacities in ophthalmology. The technique uses low coherence infrared light to visualize the reflectance of retinal structures in depth, similar to ultrasound. Since the early days about 20 years ago, technical advances in image quality and acquisition have made this technique crucial in the diagnostic procedure of many retinal and optic nerve disorders. The transition from TimeDomain (TDOCT) to SpectralDomain (SDOCT) based OCT systems increased sensitivity, resolution (from 15 µm axial resolution to 6 µm) and recording speed (from 400 axial scans/s to 40000 axial scans/s) [^{1}]. In contrast to the image quality of modern devices, image analysis tools are still quite restricted. Most commercial OCT devices currently perform only the segmentation of the entire retina or segmentation of the retinal nerve fiber layer (RNFL). It is of note though that the high resolution of SDOCT scans theoretically allows for the analysis of more structures, including all discernible intraretinal layers, blood vessels, druses, and fluid filled regions. However, a meaningful analysis requires the identification and segmentation of all of these structures in the same OCT Bscan.
The analysis of retinal blood vessels can be important not only in the diagnosis of various retinal diseases like diabetic retinopathy or glaucoma, it may also be an important tool to align different images taken from patients at different clinical visits. In addition, the RNFL thickness measurement can be made more accurate if the vessels are separated from the actual nerve fibers by a reliable segmentation method. However, manual segmentation of small retinal structures like blood vessels is extremely time consuming, and reliability is limited due to inter and intragrader variability. Due to the huge amount of data to be analyzed in a Cscan, it is practically impossible to perform segmentation tasks by an expert in clinical routine.
To make the amount of data available for routine diagnostics, automated segmentation algorithms may be the solution, as they can reduce analysis time and increase reliability for repeatable and quantitative results.
Most published vessel segmentation approaches are based on fundus images. For example, a thresholding based method was proposed by Hoover et al. [^{2}], where the fundus image is examined in pieces and the pieces are tested for a number of criteria, and the algorithm is driven by a matched filter response. Budai et al. [^{3}] proposed a multiscale algorithm for automatic segmentation, which uses a Gaussian resolution hierarchy and an efficient neighborhood analysis approach on each level to segment vessels of different diameters. A neural network classification based method was described by Sinthanayothin et al. [^{4}], where the inputs for a multilayer perceptron neuronal network were derived from a principal component analysis (PCA) of the fundus image and edge detection of the first component of PCA. However, all these examples of vessel segmentation approaches were only performed on fundus images.
The first segmentation method based on OCT images was presented by Niemeijer et al. [^{5}]. The method is based on a supervised pixel classification of a 2D projection created by information from certain automatically segmented retinal layers in a 3D volume. The performance of the method depends on the robustness of the retinal layer segmentation and axial segmentation is not performed. A second classification based method to segment blood vessels in lateral direction was proposed by Xu et al. [^{6}], in which a training set of vessel and nonvessel Ascans is created to segment vessel Ascans in unseen images (i.e. all images that were not used for training) by classification. Again, an axial segmentation of the blood vessels is not performed.
The first method to segment vessels in lateral and axial direction was proposed by Lee et al. [^{7}]. The algorithm is based on a 3D graph search. A graph is composed of a set of nodes and a set of arcs and is created from the image, where the nodes correspond to the image pixels and the arcs connect the nodes. The retinal blood vessel segmentation in the 3D graph search method was based on a vessel model. To create the triangular mesh of an initial 3D vessel model in x and ydirection, the OCT projection image was calculated by averaging two retinal layer surfaces. Additional information for the model creation concerning the depth position in zdirection was gathered from optical Doppler tomography images.
In this paper we present for the first time the use of a shape model approach for automated segmentation of retinal blood vessels with cross sectional Bscans, where apriori knowledge about the shape characteristics of retinal blood vessels is used. The shape model is used to segment objects in unseen images, where plausible variations of the model are created and compared to the image data until the object fits best. The proposed method performs segmentation in lateral and axial direction compared to most other approaches where only the lateral position is determined.
OCT images were acquired from 8 different healthy eyes for statistical shape model creation and from 13 different healthy eyes for algorithm evaluation using the Spectralis OCT system (Heidelberg Engineering, Heidelberg, Germany) with an axial resolution of 4 µm, a transversal resolution of 14 µm and an acquisition speed of 40000 AScans/sec. The volume and circular scan protocols were used for acquisition of images with sizes of 768 × 496, 1024 × 496 pixels and 1536 × 496 pixels. The image data were exported as raw data, and no processing steps were done by the OCT device itself. The algorithms were written in C/C++.
The development of the algorithm consisted of two major parts. In the first part, a statistical shape model (SSM) was created in a training step via the point distribution model, in which training data were segmented manually by so called landmarks. In the second part, the actual segmentation procedure was performed on unseen images by creating a shadowgraph of the scan to assign the lateral vessel position followed by the active shape model (ASM) method to segment the vessels in vertical direction. Preprocessing procedures, such as noise reduction and contrast adjustment were done prior to the segmentation. The basic segmentation algorithm flow is summarized in Fig. 1.
Different speckle noise reduction approaches have been developed in the past years and were used by OCT segmentation algorithms for suppression. Fernández [^{8}] showed that denoising by a nonlinear complex diffusion filter [^{9}] is an essential processing step for detection of fluidfilled region boundaries by a deformable contour model (also known as snakes). Kajić et al. [^{10}] applied a dualtree wavelet algorithm [^{11}] for image data denoising for intraretinal layer segmentation.
Our algorithm uses a general Bayesian estimation based method for speckle noise reduction, in which the image is projected into the logarithmic space to estimate the noisefree data using a conditional posterior approach [^{12}]. Figure 2 shows that the method provides effective speckle noise reduction on Spectralis OCT scans in our set up. The algorithm was processed on a section of 300 AScans centered on the human healthy macula and provides effective noise reduction while preserving thin retinal layers and small retinal blood vessels.
The starting point of creating a statistical shape model is the generation of manually segmented samples. In our project, the retinal blood vessel shape model is based on the work of Cootes et al., who proposed a point distribution model (PDM), in which the shape of a manually segmented object is defined by points, so called landmarks [^{13}]. An object is defined by a vector of image coordinates of n landmarks _{x=(x1,y1,x2,y2,...,xn,yn)}, in which the number of landmarks must be the same for all objects of the sample. Corresponding landmarks of the manual segmentations are aligned by performing translation, rotation and scaling using a Procrustes analysis.
The mean shape vector _{x¯} represents the mean object shape of the sample with s elements, where the landmarks of all objects are averaged by Eq. (1):
x¯=1s∑t=1sxt 
The entire set of manually segmented shapes consists of 52 blood vessels from 8 Bscans from 8 different healthy eyes. Figure 3 shows 20 blood vessel shapes of the entire set of 52 blood vessels, in which each shape is defined by 10 landmarks in equal distance along the boundary. The resulting mean shape of the set of shapes is shown on the right side of Fig. 3.
To reduce the number of dimensions of the training set, a principal component analysis (PCA) is applied to find a small set of modes that describes the shape variation. The deviation of each shape dx_{t} from the mean shape is calculated by Eq. (2) and the covariance matrix is calculated by Eq. (3):
dxt=xt−x¯ 
S=1s∑t=1sdxtdxtT 
An eigendecomposition gives the principal modes of variation p_{l} (eigenvectors) and the corresponding variances λ_{l} (eigenvalues) of S. The matrix P is built from each eigenvector ordered in descending order of the eigenvalues λ_{l} in Eq. (4):
P=[p1…p2n] 
The goal of calculating a PCA is to reduce the number of parameters in the SSM. The t modes of variation are chosen until the accumulated variance λ_{l} reaches a certain ratio. Here the common ratio of 0.98 was chosen, and the retinal blood vessel model includes 98% of the training set shape variations.
A shape instance x of the training set is then approximated by Eq. (5):
x=x¯+Pb 
New conformable shapes of the model are created by varying the shape coefficients in b within suitable limits in order to position the population of landmarks between three times the standard deviation in Eq. (6):
−3λl≤bl≤3λl 
By defining pose parameters, an instance of the model is created in the image frame by Eq. (7):
X=M(s,θ)[x]+XC, 
where _{M(s,θ)[]} is a scaling by s and a rotation by θ, and _{XC} defines the center position.
Regions surrounding corresponding landmarks are similar in different images. Therefore, we used a formulation for the analysis of the greylevel appearance [^{14}]. On each landmark i of each image I_{j} the greylevel profile g_{ij} of n_{p} pixel length is sampled on the normal to the boundary (see Fig. 4). We used a profile length of 25 pixels. The kth greylevel element of the profile is sampled by Eq. (8):
gijk=Ij(xik,zik) 
The greylevel profile will be normalized in order to be less sensitive to variations in a profile by Eq. (9):
gij'=gij∑k=1npgijk 
For each landmark i the mean profile and the covariance matrix are calculated in Eq. (10) and Eq. (11):
g¯i=1s∑j=1sgij' 
Si=1s∑j=1s(gij−g¯i)(gij−g¯i)T 
In unseen images, blood vessel contours are calculated by the active shape model (ASM) [^{14}], in which the statistical shape model and the greylevel variations of each landmark are combined. The segmentation process starts by placing the mean shape nearby the object to be segmented. The shape coefficients are changed iteratively until the object fits optimally. We have included greylevel profile models for each landmark ensuring that the ASM searches for regions that fit best to the profile models in new images. On each model landmark a profile is sampled with a larger length then the model profile (>n_{p}). The model profile is compared to the sampled profile to find the best position. We used the Mahalanobis distance to calculate the optimal fit position between the greylevel profile of the object in the new image and the model profile for each landmark i by Eq. (12):
f(d)=(h(d)−g¯i)TSi−1(h(d)−g¯i), 
where h(d) is a subinterval of the model profile length n_{p} of the sampled profile (length > n_{p}) centered at d [^{14}]. Therefore, for each shape model landmark a vector with position movements results in Eq. (13):
dX=(dX0,dY0,dX1,dY1, …,dXn−1,dYn−1) 
The landmarks of the current model X should be moved to the best new locations (X+ dX), which need to be within conformable possible object shape variations. This is achieved by finding the deviations in translation (dX_{c}, dY_{c}), rotation dθ and scale ds that fit best the current instance to the new positions (X+ dX). This is done by the Procrustes analysis. The residual adjustments dx in the local model coordinate frame is given by Eq. (14):
M(s(1+ds), (θ+dθ))[x+dx]+(Xc+dXC)=(X+dX) 
This results in
dx=M((s(1+ds))−1, −(θ+dθ)[y]−x, 
where_{y=M(s,θ)[x]+dX−dXC}. To adjust the model points as closely to dx as possible, the changes of the shape coefficients db is defined by Eq. (16):
db=PTdx 
The segmentation process stops when changes to the current shape model instance are calculated. The ASM procedure is shown in Fig. 5 for one iteration.
An important prerequisite of the ASM segmentation procedure is that the initial model instance must be placed nearby the object in the unseen images. Otherwise, the algorithm will not be able to calculate the greylevel profiles of the vessels (see Fig. 5) to provide reliable results.
The easiest way to initialize the shape model would be by user interaction [^{15},^{16}]. The mean shape is placed nearby the object manually. Such a manual initialization was proposed by Hug et al. [^{17}] where the user defines a principal control polygon for the shape with a small number of landmarks. However, this procedure is time consuming and results in different segmentations, and is therefore not used in our algorithm.
To perform the initialization of the ASM automatically, we use a search strategy [^{18}–^{20}], in which parts are based on the initialization by a global genetic search algorithm [^{21},^{22}]. The respective calculation steps are summarized below.
In relation to the blood vessel sizes the search space of an OCT image is large. The maximum resolution of a Bscan at the Spectralis OCT device is 1536 × 496 = 761856 pixels, but samples of the greylevel profiles on the normals of each landmark are just 25 pixels long.
The search strategy for retinal blood vessels is based on the initialization algorithm of [^{20}]. The ASM optimization algorithm with the greylevel appearance model is used to find the initial pose and model parameters. The search algorithm starts with an initial population of shape model configurations. The best matching candidates survive and are reused in further estimation rounds. Candidates with the best configuration are used for initialization. The algorithm flow including pseudocode is described in detail in [^{20}].
As an important parameter, any available apriori knowledge is used to narrow the range of possible configurations. In our case, the retinal structure intensity above and below vessels in an OCT Ascan differs compared to regions without vessels.
Wehbe et al. [^{23}] determined the lateral blood vessel positions by a so called shadowgraph, where two basic processing steps are calculated. First, a retinal layer segmentation approach was performed to remove the upper retinal layers. Second, the sum of each Ascan was calculated, in which nonvessel Ascans got a large sum, whereas Ascans with vessel got a small sum. Hence, the lateral positions of the blood vessels were determined by the first derivative of a shadowgraph. However, the graph in our algorithm is not created by the sum but by the greylevel centers of each Ascan position xEq. (17):
zcentre, x=∑z=0h−1z*I(x,z)∑z=0h−1I(x,z), 
where h is the height of the Bscan image. Figure 6a shows the switch from a nonvessel Ascan to an Ascan with blood vessel and the effect on the greylevel center, and Fig. 6b shows the center positions for all Ascans of the image marked as green line. In Ascans with blood vessels, the peaks of the greylevel center are strong, which means that the lateral position and the width of the blood vessel shadows were determined by thresholding.
The information of the shadowgraph was used to preestimate a number of initial search configurations and specific search ranges. The number of vessels, their locations in lateral direction, the widths and the locations in axial direction were approximated.
The above described algorithm for automated blood vessel segmentation in Bscans taken with a Spectralis OCT device was evaluated for accuracy and reliability against three expert readers. In total, 168 retinal blood vessels in 34 OCT Bscans from 13 healthy eyes were manually segmented with markers. The first method to validate the performance of the algorithm was to compare the mean Euclidean distance between the contours defined by the algorithm and the contours defined by the experts. The method has been widely used to analyze algorithms for contour segmentation [^{9}]. In a second validation step, we compared the areas of the blood vessels.
The mean Euclidean distance between a set of landmark points A (in this case A is the set of automated landmarks) and a second set M (in this case M is the set of expert’s landmarks) is calculated by Eq. (18):
d¯(A,M)=1n∑m=1nmina=1,...,j‖A(a)−M(m)‖, 
where n is the number of landmarks within the set M and b within the set A. A(a) and M(m) represent the coordinate vector of the ath and mth landmark.
However, the mean Euclidean distance of Eq. (18) is not symmetric, _{d¯(A,M)≠d¯(M,A)}. To get a symmetric definition, Eq. (19) is calculated:
d¯AM=d¯(A,M)+d¯(M,A)2 
Figure 7 shows three Bscans together with the detail of the respective infrared fundus image. The segmentation results of the algorithm and one expert (expert 1) are shown as an example in order to visualize the obtained vessel contours. The average automated processing time for the three Bscans with 13 retinal blood vessels is about 9 seconds (Windows 7 Professional 64 bit, used one processor core of Intel Core i7 @ 2.8 GHz and 4 GB RAM). In contrast, expert readers need at least 120 seconds for all three Bscans.
The average distances of the measured contours within the group of experts and between each expert and the algorithm for all 168 blood vessels are detailed in Table 1. The distances between algorithm and the three experts are on average approximately 4 pixels respectively (resp.) 20 µm. The minimal distance is 4.01 ± 1.97 pixel resp. 19.58 ± 9.76 µm between the automated method and expert 2. The maximal distance is 4.13 ± 1.967 pixel resp. 20.44 ± 10.06 µm between the automated method and expert 1. Compared to the distances among the expert readers, there is a small disagreement with a minimal interreader distance of 3.66 ± 1.73 pixel resp. 18.06 ± 8.91 µm and a maximal interreader distance of 3.85 ± 1.92 pixel resp. 19.67 ± 9.63 µm.
In order to evaluate the accuracy of the segmentation results and to correlate the measurements with the anatomical structures, we measured the diameter of the vessel contour of all vessels in Fig. 7 on the fundus image and correlated them with the calculated contour diameter on the B scan. The results are shown in Table 2. It is of note to mention that the diameter of the contour and not the real vessel diameter was compared, as the vessel diameter can only be evaluated if the angle of intersection of the vessel on the fundus image or OCT scan is taken into account. On average, the correlation of the contour diameter on the OCT scan and the vessel contour diameter on the fundus image was high (0.99), indicating a meaningful segmentation of the vessels by expert readers and algorithm.
The second method to evaluate the accuracy of our algorithm was to compare the obtained surface areas of the contours. Table 3 summarizes the mean area and the standard deviation for the segmentations of the experts and the automated method for all 168 blood vessels. The minimal surface area is 11040 ± 7254 µm^{2} for the segmentations of expert 3 and the maximal surface area is 13969 ± 9461 µm^{2} for the segmentations of expert 2. The automated segmentations were well within the range of the minimal and maximal surface areas (12435 ± 8738 µm^{2}).
The area deviations of all vessels from the 168 Bscans are shown in a BlandAltman plot [^{24}] in Fig. 8 in order to compare the interreader deviations and the deviations of each expert with the algorithm.
It can be seen on all plots that as the measurement increases, so does the disagreement, exceeding the standard deviation more likely for larger vessels. The smallest discrepancy between experts and the algorithm is a mean area deviation of −249 ± 9973 µm^{2} (2×SD) for expert 1, while the biggest discrepancy is a mean area deviation of 1397 ± 9214 µm^{2} (2×SD) for expert 3. Similarly, the smallest discrepancy among the experts is between expert 1 and expert 2 with a mean area deviation of −1153 ± 10856 µm^{2} (2×SD), while the biggest discrepancy is between the segmentations of expert 2 and expert 3 with a mean area deviation of 3008 ± 9402 µm^{2} (2×SD).
The intra reader variability was also tested using the intraclass correlation coefficient (ICC) [^{25}]. The repeated segmentations of 23 blood vessels of 4 different eyes at several time points showed good agreement. An ICC of 0.93 was obtained for the first expert, 0.91 for the second, and 0.94 for the third expert. An ICC between 0.4 and 0.75 is considered fair to good and a higher value excellent agreement.
In this work we present an algorithm for automatic segmentation of retinal blood vessels in SDOCT Bscans in lateral and vertical direction. The algorithm uses a model based approach based on a SSM for initial training of blood vessel contours and an ASM application for blood vessel segmentation in unseen images. Compared to the approach for 3D vessel segmentation based on the graph search [^{7}], retinal surface segmentation is not required. A limitation of our method is that it performs 2D Bscans rather than 3D Cscans, as performed in the graph search approach [^{7}]. However, it is possible to segment an entire Cscan by processing the segmentation for each Bscan separately.
We used Bscans from healthy eyes for the verification of our approach, and therefore, the robustness of the proposed algorithm in the context of diseased eyes remains to be demonstrated. Further studies characterizing blood vessels in diseased retinae are underway.
Bscans used in this study were obtained from Spectralis OCT devices and, therefore, are averaged over a defined number of scans in order to reduce speckle noise. Other OCT devices use single Bscans, which may possess a higher speckle noise and can potentially alter detection efficiency. Even though increased noise may therefore be a limitation of the approach, we think that the algorithm performs well under such conditions. The robustness of a SSM for intraretinal layer segmentation under conditions of increased noise was successfully demonstrated elsewhere [^{10}].
To avoid manual interaction for approximating vessel position prior to segmentation, we implemented a shadowgraph built by the greylevel centers of each Ascan [^{23}]. This feature makes the algorithm fully automatic, which is important for the segmentation of large Cscans and when time is an issue.
The algorithm was validated against manual segmentations on a large set of 168 blood vessels of different sizes from 34 Bscans from 13 healthy eyes done by three different experts. Contour diameter analysis was done between fundus image measurements and Bscan measurements in order to demonstrate proximity of the obtained data to the real anatomical structures. The correlation was always above 0.98. In order to verify the accuracy of vessel boundary segmentation, the mean Euclidean distances of the boundaries were evaluated between the automatic and manual segmentations. Furthermore, the area discrepancy between automatically and manually segmented vessels was tested. The results of the validation indicate that the SSM approach for initial training of the algorithm, and the ASM for segmentation in unseen images, are feasible approaches to obtain accurate segmentations. Overall, the mean Euclidean distances between algorithm and experts are very small in all three situations, approximately 4 pixels resp. 20 µm. The distances among the three experts are approximately 3.8 pixels resp. 18 µm. In the paper proposing the 3D graph search approach [^{7}], only the distances between manually marked blood vessel center points and the automated contour center points were evaluated, rather than comparing entire contour positions, which makes it impossible to compare the accuracy of such an approach with our algorithm.
The evaluation demonstrates that the proposed algorithm is similarly accurate compared to expert readers, and with a very short segmentation time it would be advantageous to use the algorithm in large scale analyses, such as multicenter clinical trials.
Nevertheless, as seen in Fig. 5, even our algorithm is not able to segment the boundaries of every given blood vessel with 100% accuracy. This small discrepancy is due to the limited adaption potential of the ASM, as this is restricted to the available range of trained variations, and local high frequency deformations cannot be fitted exactly by the model. This limitation of automatic segmentation algorithms for locally limited structures is known and cannot be avoided. In any case, the obtained results displayed very high accuracy.
Evaluation of the surface area measurements also revealed a good correlation between automatic segmentation and expert results. The deviations between the algorithm and each of the three experts were not larger than the deviations among the experts themselves. The disagreement among experts and between experts and algorithm increased when the size of the vessel was larger (Fig. 8), which is due to the multiplication of the errors during surface calculation. However, as the aim of the verification method was to demonstrate that the measurements of the algorithm would not differ from expert measurements, this observation is not in disagreement with the general conclusion. This is particularly important, as errors in the surface area measurement would have marked effects on the reliability of the data.
Measurement of blood vessel surfaces is a valuable clinical application, as knowledge about the vessel surface on a Bscan enables further analytical possibilities for the ophthalmologist. For example, the retinal nerve fiber layer (RNFL) can be characterized with or without embedded blood vessels. As large blood vessels make an important part of the RNFL thickness, changes may be better detectable without the presence of the vessels. This algorithm would be a useful tool in the diagnosis of diseases, in which the RNFL thickness may be reduced, such as glaucoma or other optic neuropathies.
In conclusion, the algorithm presented herein is an accurate and reliable tool for the automated segmentation of retinal blood vessels. This enables the ophthalmologist to analyze blood vessels in B and Cscans and may open the way for new and innovative diagnostic possibilities in the prevention of ocular diseases.
The work was supported by a grant from the von Behring Röntgen Foundation (580038).
References and links
1.  Geitzenauer W.,Hitzenberger C. K.,SchmidtErfurth U. M.. , “Retinal optical coherence tomography: past, present and future perspectives,” Br. J. Ophthalmol.95(2), 171–177 (Year: 2011).10.1136/bjo.2010.18217020675732 
2.  Hoover A.,Kouznetsova V.,Goldbaum M.. , “Locating blood vessels in retinal images by piecewise threshold probing of a matched filter response,” IEEE Trans. Med. Imaging19(3), 203–210 (Year: 2000).10.1109/42.84517810875704 
3.  A. Budai, G. Michelson, and J. Hornegger, “Multiscale Blood Vessel Segmentation in Retinal Fundus Images,” in Proceedings of Bildverarbeitung für die Medizin (Springer Verlag, 2010), pp. 261–265. 
4.  Sinthanayothin C.,Boyce J. F.,Cook H. L.,Williamson T. H.. , “Automated localisation of the optic disc, fovea, and retinal blood vessels from digital colour fundus images,” Br. J. Ophthalmol.83(8), 902–910 (Year: 1999).10.1136/bjo.83.8.90210413690 
5.  Niemeijer M.,Garvin M. K.,van Ginneken B.,Sonka M.,Abràmoff M. D.. , “Vessel segmentation in 3D spectral OCT scans of the retina,” Proc. SPIE6914, 69141R, 69141R8 (Year: 2008).10.1117/12.772680 
6.  J. Xu, D. Tolliver, H. Ishikawa, C. Wollstein, and J. Schuman, “Blood vessel segmentation with threedimensional spectral domain optical coherence tomography,” International Patent no. WO/2010/138645 (Feb. 12, 2010). 
7.  K. Lee, “Segmentations of the intraretinal surfaces, optic disc and retinal blood vessels in 3DOCT scans,” Ph.D. dissertation (University of Iowa, 2009), pp. 57–69. 
8.  Fernández D. C.. , “Delineating fluidfilled region boundaries in optical coherence tomography images of the retina,” IEEE Trans. Med. Imaging24(8), 929–945 (Year: 2005).10.1109/TMI.2005.84865516092326 
9.  Perona P.,Malik J.. , “Scalespace and edge detection using anisotropic diffusion,” IEEE Trans. Pattern Anal. Mach. Intell.12(7), 629–639 (Year: 1990).10.1109/34.56205 
10.  Kajić V.,Považay B.,Hermann B.,Hofer B.,Marshall D.,Rosin P. L.,Drexler W.. , “Robust segmentation of intraretinal layers in the normal human fovea using a novel statistical model based on texture and shape analysis,” Opt. Express18(14), 14730–14744 (Year: 2010).10.1364/OE.18.01473020639959 
11.  Selesnick I. W.,Baraniuk R. G.,Kingsbury N. G.. , “The dualtree complex wavelet transform,” IEEE Signal Process. Mag.22(6), 123–151 (Year: 2005).10.1109/MSP.2005.1550194 
12.  Wong A.,Mishra A.,Bizheva K.,Clausi D. A.. , “General Bayesian estimation for speckle noise reduction in optical coherence tomography retinal imagery,” Opt. Express18(8), 8338–8352 (Year: 2010).10.1364/OE.18.00833820588679 
13.  Cootes T.,Taylor C.,Cooper D.,Graham J.. , “Active shape models—their training and application,” Comput. Vis. Image Underst.61(1), 38–59 (Year: 1995).10.1006/cviu.1995.1004 
14.  T. Cootes, C. Taylor, A. Lanitis, D. Cooper, and J. Graham, “Building and using flexible models incorporating greylevel information,” in Fourth International Conference on Computer Vision, 1993. Proceedings (1993), pp. 242–246. 
15.  Kelemen A.,Székely G.,Gerig G.. , “Elastic modelbased segmentation of 3D neuroradiological data sets,” IEEE Trans. Med. Imaging18(10), 828–839 (Year: 1999).10.1109/42.81126010628943 
16.  Lamecker H.,Seebaß M.,Hege H.C.,Deuflhard P.. , “A 3D statistical shape model of the pelvic bone for segmentation,” Proc. SPIE5370, 1341–1351 (Year: 2004).10.1117/12.534145 
17.  J. Hug, C. Brechbühler, and G. Székely, “Modelbased Initialisation for Segmentation,” in Computer Vision—ECCV 2000 (Springer, 2000), pp. 290–306. 
18.  G. Edwards, T. Cootes, and C. Taylor, “Advances in active appearance models,” The Proceedings of the Seventh IEEE International Conference on Computer Vision (IEEE, 1999), Vol. 1, pp. 137–142. 
19.  R. Fisker, N. Schultz, N. Duta, and J. Carstensen, “A general scheme for training and optimization of the Grenander deformable template model,” in IEEE Conference on Computer Vision and Pattern Recognition,2000. Proceedings (IEEE, 2000), pp. 698–705. 
20.  M. Stegmann, R. Fisker, and B. Ersbøll, “Extending and applying active appearance models for automated, high precision segmentation in different image modalities,” in Scandinavian Conference on Image Analysis, (2001), pp. 90–97. 
21.  D. Goldberg, Genetic Algorithms in Search, Optimization, and Machine Learning (AddisonWesley, Bonn, 1989). 
22.  Hill A.,Taylor C.. , “Modelbased image interpretation using genetic algorithms,” Image Vis. Comput.10(5), 295–300 (Year: 1992).10.1016/02628856(92)900455 
23.  Wehbe H.,Ruggeri M.,Jiao S.,Gregori G.,Puliafito C. A.,Zhao W.. , “Automatic retinal blood flow calculation using spectral domain optical coherence tomography,” Opt. Express15(23), 15193–15206 (Year: 2007).10.1364/OE.15.01519319550803 
24.  Bland J. M.,Altman D. G.. , “Statistical methods for assessing agreement between two methods of clinical measurement,” Lancet327(8476), 307–310 (Year: 1986).10.1016/S01406736(86)9083782868172 
25.  H. Fleiss, Statistical Methods for Rates and Proportions, 2nd ed. (Wiley New York, 1981). 
Figures
Tables
Mean Euclidean distance ± SD



Auto.—Expert 1  Auto.—Expert 2  Auto.—Expert 3 
4.13 ± 1.97 pixel 20.44 ± 10.06 µm  4.01 ± 1.97 pixel 19.58 ± 9.76 µm  4.05 ± 1.74 pixel 20.05 ± 8.95 µm 
Expert 1—Expert 2  Expert 1—Expert 3  Expert 2—Expert 3 

3.85 ± 1.92 pixel 19.67 ± 9.63 µm  3.66 ± 1.73 pixel 18.06 ± 8.91 µm  3.72 ± 1.7 pixel 18.26 ± 8.48 µm 
BScan  Blood vessel 
Contour diameter [µm]



Fundus  Automated Seg.  Expert 1  Expert 2  Expert 3  
Fig. 7(a)  1  183.07  173.73  173.73  179.16  195.45 
2  114.66  124.87  119.44  103.15  130.3  
3  237.79  249.74  244.31  228.6  228.02  
4  113.19  114.01  124.87  119.44  105.58  
Fig. 7(b)  1  73.97  91.67  80.21  85.94  98.51 
2  170.24  161.17  161.17  177.62  166.16  
3  199.35  217.73  203.53  206.27  217.73  
4  145.71  143.24  132.01  137.56  154.51  
Fig.7(c)  1  167.45  159.78  170.8  181.82  154.27 
2  85.68  104.78  88.15  110.19  77.13  
3  74.50  82.64  82.64  82.64  71.62  
4  81.11  82.64  82.64  71.62  77.13  
5  79.18  88.15  77.13  82.64  77.13 
Mean Area [µm^{2}] ± SD [µm^{2}]

Correlation of the automated method and the 3 experts  

Expert 1  Expert 2  Expert 3  Automated method  
12810 ± 8344  13969 ± 9461  11040 ± 7254  12435 ± 8738  0.83 
Article Categories:
Keywords: OCIS codes: (170.4500) Optical coherence tomography, (110.6880) Threedimensional image acquisition, (100.0100) Image processing, (100.3008) Image recognition, algorithms and filters. 
Previous Document: Integrative analyses reveal novel strategies in HPV11,16 and 45 early infection.
Next Document: Image formation by linear and nonlinear digital scanned lightsheet fluorescence microscopy with Gau...