TGIF: Topological Gap In-Fill for Vascular Networks

Schneider, Matthias; Hirsch, Sven; Weber, Bruno; Székely, Gábor; Menze, Bjoern

doi:10.1007/978-3-319-10470-6_12

Cited by 11 publications

(12 citation statements)

References 12 publications

(36 reference statements)

Supporting

Mentioning

Contrasting

Order By: Relevance

“…To generate synthetic data, we follow the method of Schneider et al (2012) which implements a simulator of a vascular tree that follows a generative process inspired by the biology of angiogenesis. This approach, described in Schneider et al (2012), has initially been developed to complement missing elements of a vascular tree, a common problem in µCT imaging of the vascular bed (Schneider et al, 2014). We now use this generator to simulate physiologically plausible vascular trees that we can use in training our CNN algorithms.…”

Section: Synthetic Data For Transfer Learningmentioning

confidence: 99%

“…However, assembling a properly labeled dataset of 3-D curvilinear structures, such as vessels and vessel features, takes a lot of human effort and time, which turns out to be the bottleneck for most medical applications. To overcome this problem, we generate synthetic data based on the method proposed in Schneider et al (2012Schneider et al ( , 2014. A brief description of this process has already been presented in section 2.3.…”

Section: Synthetic Datasetmentioning

confidence: 99%

“…Manually annotating vessels, centerlines, and bifurcations requires many hours of work and expertise. To this end, we demonstrate the successful use of simulation based frameworks (Szczerba and Székely, 2005;Schneider et al, 2012Schneider et al, , 2014) that can be used for generating synthetic data with accurate labels (see section 2.3) for pre-training our networks, rendering the training of our supervised classification algorithm feasible. The transfer learning approach turns out to be a critical component for training CNNs in a wide range of angiography tasks and applications ranging from CT micrographs to TOF MRA.…”

Section: Introductionmentioning

confidence: 99%

“…Our simulator enforces the Murray's law during the tree generation process.Further constraints, cos(φ l ) = the bifurcation angles of the left (φ l ) and right (φ r ) vessel extension elements respectively. This corresponds corresponds to the optimal position of the branching point − → P b with respect to a minimum volume principle, another constraint enforced in the simulator fromSchneider et al (2012Schneider et al ( , 2014.…”

mentioning

confidence: 99%

See 3 more Smart Citations

DeepVesselNet: Vessel Segmentation, Centerline Prediction, and Bifurcation Detection in 3-D Angiographic Volumes

et al. 2020

Self Cite

View full text Add to dashboard Cite

We present DeepVesselNet, an architecture tailored to the challenges faced when extracting vessel trees and networks and corresponding features in 3-D angiographic volumes using deep learning. We discuss the problems of low execution speed and high memory requirements associated with full 3-D networks, high-class imbalance arising from the low percentage (<3%) of vessel voxels, and unavailability of accurately annotated 3-D training data—and offer solutions as the building blocks of DeepVesselNet. First, we formulate 2-D orthogonal cross-hair filters which make use of 3-D context information at a reduced computational burden. Second, we introduce a class balancing cross-entropy loss function with false-positive rate correction to handle the high-class imbalance and high false positive rate problems associated with existing loss functions. Finally, we generate a synthetic dataset using a computational angiogenesis model capable of simulating vascular tree growth under physiological constraints on local network structure and topology and use these data for transfer learning. We demonstrate the performance on a range of angiographic volumes at different spatial scales including clinical MRA data of the human brain, as well as CTA microscopy scans of the rat brain. Our results show that cross-hair filters achieve over 23% improvement in speed, lower memory footprint, lower network complexity which prevents overfitting and comparable accuracy that does not differ from full 3-D filters. Our class balancing metric is crucial for training the network, and transfer learning with synthetic data is an efficient, robust, and very generalizable approach leading to a network that excels in a variety of angiography segmentation tasks. We observe that sub-sampling and max pooling layers may lead to a drop in performance in tasks that involve voxel-sized structures. To this end, the DeepVesselNet architecture does not use any form of sub-sampling layer and works well for vessel segmentation, centerline prediction, and bifurcation detection. We make our synthetic training data publicly available, fostering future research, and serving as one of the first public datasets for brain vessel tree segmentation and analysis.

show abstract

Section: Synthetic Data For Transfer Learningmentioning

confidence: 99%

Section: Synthetic Datasetmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

mentioning

confidence: 99%

See 2 more Smart Citations

DeepVesselNet: Vessel Segmentation, Centerline Prediction, and Bifurcation Detection in 3-D Angiographic Volumes

et al. 2020

Self Cite

View full text Add to dashboard Cite

show abstract

“…This high level strategy is complementary to lower level ones like in [11], where only local image features communicate to fill the gaps. In the same vein, the simultaneous reconstruction and separation of multiple interwoven tree structures using high-level representation of the trees was also proposed in [1], and a physiologically motivated strategy based on a simplified angiogenesis model was proposed in [12] to correct the vascular connectivity. Closer to our strategy, [3] derived intensity based information within a tensor model to perform the robust segmentation of tubular structures.…”

Section: Introductionmentioning

confidence: 99%

Filling Large Discontinuities in 3D Vascular Networks Using Skeleton- and Intensity-Based Information

Bates

Risser

Irving

et al. 2015

Lecture Notes in Computer Science

View full text Add to dashboard Cite

Abstract. Segmentation of vasculature is a common task in many areas of medical imaging, but complex morphology and weak signal often lead to incomplete segmentations. In this paper, we present a new gap filling strategy for 3D vascular networks. The novelty of our approach is to combine both skeleton-and intensity-based information to fill large discontinuities. Our approach also does not make any hypothesis on the network topology, which is particularly important for tumour vasculature due to the chaotic arrangement of vessels within tumours. Synthetic results show that using intensity-based information, in addition to skeleton-based information, can make the detection of large discontinuities more robust. Our strategy is also shown to outperform a classic gap filling strategy on 3D Micro-CT images of preclinical tumour models.

show abstract

A 3D‐1D coupled blood flow and oxygen transport model to generate microvascular networks

Köppl¹,

Vidotto²,

Wohlmuth

2020

Numer Methods Biomed Eng

View full text Add to dashboard Cite

In this work, we introduce an algorithmic approach to generate microvascular networks starting from larger vessels that can be reconstructed without noticeable segmentation errors. Contrary to larger vessels, the reconstruction of fine-scale components of microvascular networks shows significant segmentation errors, and an accurate mapping is time and cost intense. Thus there is a need for fast and reliable reconstruction algorithms yielding surrogate networks having similar stochastic properties as the original ones. The microvascular networks are constructed in a marching way by adding vessels to the outlets of the vascular tree from the previous step. To optimise the structure of the vascular trees, we use Murray's law to determine the radii of the vessels and bifurcation angles. In each step, we compute the local gradient of the partial pressure of oxygen and adapt the orientation of the new vessels to this gradient. At the same time, we use the partial pressure of oxygen to check whether the considered tissue block is supplied sufficiently with oxygen. Computing the partial pressure of oxygen, we use a 3D-1D coupled model for blood flow and oxygen transport. To decrease the complexity of a fully coupled 3D model, we reduce the blood vessel network to a 1D graph structure and use a bi-directional coupling with the tissue which is described by a 3D homogeneous porous medium. The resulting surrogate networks are analysed with respect to morphological and physiological aspects. K E Y W O R D S 3D-1D coupled flow models, blood flow simulations, dimensionally reduced models, flows in porous media, oxygen transport, vascular growth

show abstract

TGIF: Topological Gap In-Fill for Vascular Networks

Cited by 11 publications

References 12 publications

DeepVesselNet: Vessel Segmentation, Centerline Prediction, and Bifurcation Detection in 3-D Angiographic Volumes

DeepVesselNet: Vessel Segmentation, Centerline Prediction, and Bifurcation Detection in 3-D Angiographic Volumes

Filling Large Discontinuities in 3D Vascular Networks Using Skeleton- and Intensity-Based Information

A 3D‐1D coupled blood flow and oxygen transport model to generate microvascular networks

Contact Info

Product

Resources

About