Cerebrovascular segmentation from TOF using stochastic models

Hassouna, M. Sabry; Farag, Aly A.; Hushek, Stephen G.; Moriarty, Thomas

doi:10.1016/j.media.2004.11.009

Cited by 98 publications

(60 citation statements)

References 27 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Many segmenting strategies are available (for an overview, see (17)), including histogram-based techniques for TOF-MRA (18). However, segmentation of vessels with a diameter <5 pixels is considered nearly impossible based on overlapping signal intensity distributions with background tissue (11,19). Even though segmentation of collateral arteries may not be achievable, filtering was beneficial for SID analysis.…”

Section: Discussionmentioning

confidence: 99%

Automated multiscale vessel analysis for the quantification of MR angiography of peripheral arteriogenesis

Jaspers

Slenter

Leiner

et al. 2011

Magnetic Resonance Imaging

View full text Add to dashboard Cite

Purpose: To automatically analyze the time course of collateralization in a rat hindlimb ischemia model based on signal intensity distribution (SID). Materials and Methods:Time-of-flight magnetic resonance angiograms (TOF-MRA) were acquired in eight rats at 2, 7, and 21 days after unilateral femoral artery ligation. Analysis was performed on maximum intensity projections filtered with multiscale vessel enhancement filter. Differences in SID between ligated limb and a reference region were monitored over time and compared to manual collateral artery identification.Results: The differences in SID correlated well with the number of collateral arteries found with manual quantification. The time courses of ultrasmall (diameter (0.5 mm) and small (diameter %0.5 mm) collateral artery development could be differentiated, revealing that maturation of the collaterals and enlargement of their feeding arteries occurred mainly after the first week postligation.Conclusion: SID analysis performed on axial maximum intensity projections is easy to implement, fast, and objective and provides more insight in the time course of arteriogenesis than manual identification. COLLATERAL ARTERIES developing from preexistent arterioles can act as natural bypasses in patients with peripheral or coronary arterial occlusive disease (1). Currently, therapies stimulating the formation of collateral arteries (ie, arteriogenesis) and restoring the impaired blood supply are explored (2-4). Accurate evaluation of such novel treatments strongly depends on image acquisition and analysis techniques that can detect and quantify the vascular responses both in experimental and clinical settings. Previously, noncontrast-enhanced inflow magnetic resonance angiography (MRA) techniques such as time-of-flight (TOF) MRA have been successfully used to noninvasively quantify collateralization in small animal ischemic hindlimb models (5-8).Collaterals are usually identified and quantified manually on (rotating) maximum intensity projections (MIPs) using the Longland definition (9), which requires identification of the collateral stem, mid-, and reentry zone. Visualization of complete trajectories is challenging due to inflow artifacts, flow voids for in-plane arteries (10), and the ultrasmall collateral artery calibers ((0.5 mm). Moreover, manual quantification is time-consuming, observer-dependent, and provides little information on changes in vessel size and physiological impact. A physiologically more relevant measure would be the blood volume or flow of the collateral arteries. Wagner et al (5) proposed a method to quantify collateral flow based on the dependence between pixel signal intensity and the amount and speed of inflow of unsaturated blood spins. The signal intensity histogram of TOF angiograms may provide surrogate parameters for changes in vessel morphology, as the signal intensity distribution depends on vessel size (11). A problem to overcome is the small volume fraction of arteries in hindlimbs compared to the large volume of background tissue. ...

show abstract

Section: Discussionmentioning

confidence: 99%

Automated multiscale vessel analysis for the quantification of MR angiography of peripheral arteriogenesis

Jaspers

Slenter

Leiner

et al. 2011

Magnetic Resonance Imaging

View full text Add to dashboard Cite

show abstract

“…Hassouna et al [18] addressed the problem of detecting gaps in 3D vessel segmentation as a post-processing step in their intensity-based vessel segmentation technique. Markov Random Fields models were used to take the spatial information into account.…”

Section: State Of the Artmentioning

confidence: 99%

Automatic Correction of Gaps in Cerebrovascular Segmentations Extracted from 3D Time-of-Flight MRA Datasets

Forkert¹,

Schmidt-Richberg²,

Fiehler³

et al. 2012

Methods Inf Med

View full text Add to dashboard Cite

SummaryObjectives: Exact cerebrovascular segmentations are required for several applications in today's clinical routine. A major drawback of typical automatic segmentation methods is the occurrence of gaps within the segmentation. These gaps are typically located at small vessel structures exhibiting low intensities. Manual correction is very time-consuming and not suitable in clinical practice. This work presents a post-processing method for the automatic detection and closing of gaps in cerebrovascular segmentations. Methods: In this approach, the 3D centerline is calculated from an available vessel segmentation, which enables the detection of corresponding vessel endpoints. These endpoints are then used to detect possible connections to other 3D centerline voxels with a graphbased approach. After consistency check, rea- sonable detected paths are expanded to the vessel boundaries using a level set approach and combined with the initial segmentation. Results: For evaluation purposes, 100 gaps were artificially inserted at non-branching vessels and bifurcations in manual cerebrovascular segmentations derived from ten Time-of-Flight magnetic resonance angiography datasets. The results show that the presented method is capable of detecting 82% of the non-branching vessel gaps and 84% of the bifurcation gaps. The level set segmentation expands the detected connections with 0.42 mm accuracy compared to the initial segmentations. A further evaluation based on 10 real automatic segmentations from the same datasets shows that the proposed method detects 35 additional connections in average per dataset, whereas 92.7% were rated as correct by a medical expert. Conclusion: The presented approach can considerably improve the accuracy of cerebrovascular segmentations and of following analysis outcomes.

show abstract

“…They led, in particular to the definition of GaussianGaussianuniform [107] and normalRayleigh-2×normal [80] mixtures for timeofflight (TOF) MRA, and MaxwellGaussian [19], Maxwell Gaussianuniform [17] mixtures for phasecontrast (PC) MRA. In [18], a hybrid model, enables to choose between these two kinds of mixtures.…”

Section: Statistical Approachesmentioning

confidence: 99%

“…From an algorithmic point of view, segmentation improvements were also performed in con sidering of spatial information (i.e., statistical dependence) between neighbour voxels, by integrating Markov random fields (MRF) [38] in a postclassification correction step [80]. In other works, speed and phase information provided by PCMRA were fused and involved in a maximum a posterioriMRF framework to enhance vessel segmenta tion [17,18].…”

Section: Statistical Approachesmentioning

confidence: 99%