Evaluating and improving label fusion in atlas-based segmentation using the surface distance

Langerak, Thomas; Heide, Uulke A. van der; Kotte, Antj Alexis; Berendsen, Floris G; Pluim, Jpw Josien

doi:10.1117/12.877636

Cited by 8 publications

(8 citation statements)

References 14 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…We also showed that our model statistically outperformed the previously reported best methods of Birkbeck et al and Yuan et al and was slightly better than second observer results in DSC and Hausdorff distance. We also found that the surface of our segmentation output has a shorter Euclidean distance (Langerak et al 2011) from that of the ground truth shape. We suggest that the explanation for the better accuracy is through the RVM training and probabilistic classification.…”

Section: Discussionmentioning

confidence: 56%

A combinatorial Bayesian and Dirichlet model for prostate MR image segmentation using probabilistic image features

Wang

et al. 2016

Phys. Med. Biol.

View full text Add to dashboard Cite

Blurred boundaries and heterogeneous intensities make accurate prostate MR image segmentation problematic. To improve prostate MR image segmentation we suggest an approach that includes: (a) an image patch division method to partition the prostate into homogeneous segments for feature extraction; (b) an image feature formulation and classification method, using the relevance vector machine, to provide probabilistic prior knowledge for graph energy construction; (c) a graph energy formulation scheme with Bayesian priors and Dirichlet graph energy and (d) a non-iterative graph energy minimization scheme, based on matrix differentiation, to perform the probabilistic pixel membership optimization. The segmentation output was obtained by assigning pixels with foreground and background labels based on derived membership probabilities. We evaluated our approach on the PROMISE-12 dataset with 50 prostate MR image volumes. Our approach achieved a mean dice similarity coefficient (DSC) of 0.90 ± 0.02, which surpassed the five best prior-based methods in the PROMISE-12 segmentation challenge.

show abstract

Section: Discussionmentioning

confidence: 56%

A combinatorial Bayesian and Dirichlet model for prostate MR image segmentation using probabilistic image features

Wang

et al. 2016

Phys. Med. Biol.

View full text Add to dashboard Cite

show abstract

“…DSC has been popular in the literature of atlas-based segmentation. 24 However, as pointed out by the authors in, 24 it can over/underestimate the quality of segmentation due to its global characteristic. Therefore, we prefer to draw final conclusions based on ASSD.…”

Section: Resultsmentioning

confidence: 97%

“…In order to avoid segmentation bias toward the reference mask, 19 we computed the distances between each pair of training mask and selected the one whose average distance to the others is minimum. We used the Average Symmetric Surface Distance (ASSD) 24 for this task, because it can better detect the local differences between boundaries than global measures, such as Dice Similarity Coefficient (DSC), for instance. In order to allow locally fine tuning to the object's boundary, 9 segmentation is based on the posterior probability values as estimated for each new image.…”

Section: Segmentation Using Object Atlas Modelmentioning

confidence: 99%

Medical image segmentation using object atlas versus object cloud models

2015

View full text Add to dashboard Cite

Medical image segmentation is crucial for quantitative organ analysis and surgical planning. Since interactive segmentation is not practical in a production-mode clinical setting, automatic methods based on 3D object appearance models have been proposed. Among them, approaches based on object atlas are the most actively investigated. A key drawback of these approaches is that they require a time-costly image registration process to build and deploy the atlas. Object cloud models (OCM) have been introduced to avoid registration, considerably speeding up the whole process, but they have not been compared to object atlas models (OAM). The present paper fills this gap by presenting a comparative analysis of the two approaches in the task of individually segmenting nine anatomical structures of the human body. Our results indicate that OCM achieve a statistically significant better accuracy for seven anatomical structures, in terms of Dice Similarity Coefficient and Average Symmetric Surface Distance.

show abstract

“…The segmentation accuracy was based on the ASSD in millimeters, since it captures local errors along the segmented boundary. 38 The computational time of each method is divided into time for model construction and time for segmentation T I. Tuned values of α for combined image gradient computation per object, using SOSM-S and FOSM.…”

Section: B Evaluation Methodologymentioning

confidence: 99%

“…We adopt a standard procedure that usually differs only in the similarity metric. We use the average symmetric surface distance (ASSD) 38 to F. 3.…”

Section: Statistical Object Shape Modelsmentioning

confidence: 99%

Medical image segmentation via atlases and fuzzy object models: Improving efficacy through optimum object search and fewer models

2015

View full text Add to dashboard Cite

Although multiple models per object can usually improve segmentation efficacy, the optimum object search has shown to reduce the number of required models. The efficiency gain of FOSM over SOSM-S motivates its use for interactive applications and studies with large image data sets. FOSM and SOSM impose different degrees of shape constraints from the model, making one approach more suitable than the other, depending on contrast. This suggests the use of hybrid models that can take advantage from the strengths of fuzzy and statistical models.

show abstract

Evaluating and improving label fusion in atlas-based segmentation using the surface distance

Cited by 8 publications

References 14 publications

A combinatorial Bayesian and Dirichlet model for prostate MR image segmentation using probabilistic image features

A combinatorial Bayesian and Dirichlet model for prostate MR image segmentation using probabilistic image features

Medical image segmentation using object atlas versus object cloud models

Medical image segmentation via atlases and fuzzy object models: Improving efficacy through optimum object search and fewer models

Contact Info

Product

Resources

About