A New Marching Cubes Algorithm for Interactive Level Set with Application to MR Image Segmentation

Feltell, David; Li, Bai

doi:10.1007/978-3-642-17289-2_36

Cited by 3 publications

(4 citation statements)

References 12 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Here, we have developed a MophoNet-based method for interactive 3D visualization of segmentation quality measures, by projecting VJI values on 3D meshes created from 3D confocal images ( Fig 13 ). The 3D meshes were created by the marching cubes algorithm [ 44 ] and uploaded on MorphoNet. The VJI values for the image were computed and uploaded as a quantitative property of each individual 3D cell of the image for each segmentation pipeline.…”

Section: Resultsmentioning

confidence: 99%

Benchmarking of deep learning algorithms for 3D instance segmentation of confocal image datasets

et al. 2022

View full text Add to dashboard Cite

Segmenting three-dimensional (3D) microscopy images is essential for understanding phenomena like morphogenesis, cell division, cellular growth, and genetic expression patterns. Recently, deep learning (DL) pipelines have been developed, which claim to provide high accuracy segmentation of cellular images and are increasingly considered as the state of the art for image segmentation problems. However, it remains difficult to define their relative performances as the concurrent diversity and lack of uniform evaluation strategies makes it difficult to know how their results compare. In this paper, we first made an inventory of the available DL methods for 3D cell segmentation. We next implemented and quantitatively compared a number of representative DL pipelines, alongside a highly efficient non-DL method named MARS. The DL methods were trained on a common dataset of 3D cellular confocal microscopy images. Their segmentation accuracies were also tested in the presence of different image artifacts. A specific method for segmentation quality evaluation was adopted, which isolates segmentation errors due to under- or oversegmentation. This is complemented with a 3D visualization strategy for interactive exploration of segmentation quality. Our analysis shows that the DL pipelines have different levels of accuracy. Two of them, which are end-to-end 3D and were originally designed for cell boundary detection, show high performance and offer clear advantages in terms of adaptability to new data.

show abstract

Section: Resultsmentioning

confidence: 99%

Benchmarking of deep learning algorithms for 3D instance segmentation of confocal image datasets

et al. 2022

View full text Add to dashboard Cite

show abstract

“…Here we have developed a MophoNet based method for interactive 3D visualization of segmentation quality measures, by projecting VJI values on 3D meshes created from 3D confocal images (Fig 13). The 3D meshes were created by the marching cubes algorithm [47] and uploaded on MorphoNet. The VJI values for the image were computed and uploaded as a quantitative property of each individual 3D cell of the image for each segmentation pipeline.…”

Section: Resultsmentioning

confidence: 99%

“…In this work, we have developed a MophoNet based method for interactive 3D visualization of segmentation quality measures, by projecting VJI values on 3D meshes created from 3D confocal images (Figure 11). The 3D meshes were created by the marching cubes algorithm (Feltell & Bai, 2010) and uploaded on MorphoNet.…”

Section: Effect Of Image Noisementioning

confidence: 99%

Benchmarking of deep learning algorithms for 3D instance segmentation of confocal image datasets

Kar

Petit

Refahi

et al. 2021

Preprint

View full text Add to dashboard Cite

Segmenting three dimensional microscopy images is essential for understanding phenomena like morphogenesis, cell division, cellular growth and genetic expression patterns. Recently, deep learning (DL) pipelines have been developed which claim to provide high accuracy segmentation of cellular images and are increasingly considered as the state-of-the-art for image segmentation problems. However, it remains difficult to define their relative performance as the concurrent diversity and lack of uniform evaluation strategies makes it difficult to know how their results compare. In this paper, we first made an inventory of the available DL methods for 3D segmentation. We next implemented and quantitatively compared a number of representative DL pipelines, alongside a highly efficient non-DL method named MARS. The DL methods were trained on a common dataset of 3D cellular confocal microscopy images. Their segmentation accuracies were also tested in the presence of different image artefacts. A new method for segmentation quality evaluation was adopted which isolates segmentation errors due to under/over segmentation. This is complemented with new visualisation strategies that make interactive exploration of segmentation quality possible. Our analysis shows that the DL pipelines have very different levels of accuracy. Two of them show high performance, and offer clear advantages in terms of adaptability to new data.

show abstract

“…However, MC and its improvements have a real-time and object of interest unawareness problem because it essentially need traverse each cell in the procedure of triangles generation on the large scale data sets, and the isosurfaces of all objects are equally extracted. This problem even becomes increasingly serious along with the increasingly massive volumetric data sets produced by recent technical breakthroughs in new imaging modalities such as CT, MRI, and other 3-D scanning technologies [7], [8].…”

Section: Introductionmentioning

confidence: 99%

Improved marching cubes using novel adjacent lookup table and random sampling for medical object-specific 3D visualization

Wang¹,

Niu²,

Tan³

et al. 2014

JSW

View full text Add to dashboard Cite

The marching cubes (MC) algorithm is a widely used routine to extract isosurfaces from volumetric data set. This method suffers from the exhaustive accessing cubes sequentially and the failure of directly separating the isosurfaces. To address this, an improved MC algorithm called AlutMC is proposed to track connected surfaces by combining a new adjacent lookup table and random sampling technique. AlutMC has a natural ability to separate the isosurfaces related to different components, which can efficiently segment 3D objects without extra expenditure and helpful for visualizing medical objects. Experimental results on three test data sets show the computational efficiency in identifying isosurfaces. Additionally, it is simple and efficient, not requiring complex data structure, and can be used as is into the MC and its improvements because it is quite cost-effective and self-contained.

show abstract

A New Marching Cubes Algorithm for Interactive Level Set with Application to MR Image Segmentation

Cited by 3 publications

References 12 publications

Benchmarking of deep learning algorithms for 3D instance segmentation of confocal image datasets

Benchmarking of deep learning algorithms for 3D instance segmentation of confocal image datasets

Benchmarking of deep learning algorithms for 3D instance segmentation of confocal image datasets

Improved marching cubes using novel adjacent lookup table and random sampling for medical object-specific 3D visualization

Contact Info

Product

Resources

About