The clique potential of Markov random field in a random experiment for estimation of noise levels in 2D brain MRI

Osadebey, Michael; Bouguila, Nizar; Arnold, Douglas L.

doi:10.1002/ima.22065

Cited by 6 publications

(8 citation statements)

References 28 publications

(62 reference statements)

Supporting

Mentioning

Contrasting

Order By: Relevance

“…(14). At each iteration, the Markov random field energy expressed by the total clique potential is computed according to the formulation in Osadebey et al (2013). Using the mathematical model described by the plots in Figure 3, the stopping criterion E s is fixed at E s 5K bg U bg and E s 5 K fg U fg for background and foreground modes, respectively, where K bg and K fg are arbitrary constants.…”

Section: Methodsmentioning

confidence: 99%

See 1 more Smart Citation

Four‐neighborhood clique kernel: A general framework for Bayesian and variational techniques of noise reduction in magnetic resonance images of the brain

Osadebey

Bouguila

Arnold

2014

Int J Imaging Syst Tech

Self Cite

View full text Add to dashboard Cite

Several algorithms have been proposed in the literature for image denoising but none exhibit optimal performance for all range and types of noise and for all image acquisition modes. We describe a new general framework, built from four‐neighborhood clique system, for denoising medical images. The kernel quantifies smoothness energy of spatially continuous anatomical structures. Scalar and vector valued quantification of smoothness energy configures images for Bayesian and variational denoising modes, respectively. Within variational mode, the choice of norm adapts images for either total variation or Tikhonov technique. Our proposal has three significant contributions. First, it demonstrates that the four‐neighborhood clique kernel is a basic filter, in same class as Gaussian and wavelet filters, from which state‐of‐the‐art denoising algorithms are derived. Second, we formulate theoretical analysis, which connects and integrates Bayesian and variational techniques into a two‐layer structured denoising system. Third, our proposal reveals that the first layer of the new denoising system is a hitherto unknown form of Markov random field model referred to as single‐layer Markov random field (SLMRF). The new model denoises a specific type of medical image by minimizing energy subject to knowledge of mathematical model that describes relationship between the image smoothness energy and noise level but without reference to a classical prior model. SLMRF was applied to and evaluated on two real brain magnetic resonance imaging datasets acquired with different protocols. Comparative performance evaluation shows that our proposal is comparable to state‐of‐the‐art algorithms. SLMRF is simple and computationally efficient because it does not incorporate a regularization parameter. Furthermore, it preserves edges and its output is devoid of blurring and ringing artifacts associated with Gaussian‐based and wavelet‐based algorithms. The denoising system is potentially applicable to speckle reduction in ultrasound images and extendable to three‐layer structure that account for texture features in medical images. © 2014 Wiley Periodicals, Inc. Int J Imaging Syst Technol, 24, 224–238, 2014

show abstract

Section: Methodsmentioning

confidence: 99%

“…(14). At each iteration, the Markov random field energy is computed and the stopping criterion is fixed at ðE s 5 K b U b 521:5Þ based on knowledge of mathematical function describing relationship between MRF energy and noise level (Osadebey et al, 2013).…”

Section: B3 Double-layer Markov Randon Field: Rof Total Variation Amentioning

confidence: 99%

Four‐neighborhood clique kernel: A general framework for Bayesian and variational techniques of noise reduction in magnetic resonance images of the brain

Osadebey

Bouguila

Arnold

2014

Int J Imaging Syst Tech

Self Cite

View full text Add to dashboard Cite

show abstract

“…Perceptual weight assigned to each quality factor is derived from the results of the experiment reported in [36]. The experiment models a MRI slice as a Markov random field (MRF) [16], [59], [27] but without reference to a prior model image.…”

Section: Perceptual Weightsmentioning

confidence: 99%

“…Figure 1 and Fig. 2 shows how a simple phantom obtained from McGill University BrainWeb Simulated Database [9] can be used to demonstrate the experiment in [36]. The original image of a MRI slice, assumed to be free of noise, is shown in Fig.…”

Section: Perceptual Weightsmentioning

confidence: 99%

No‐reference quality measure in brain MRI images using binary operations, texture and set analysis

Osadebey

Pedersen

Arnold

et al. 2017

IET Image Processing

Self Cite

View full text Add to dashboard Cite

We propose a new application-specific post-acquisition quality evaluation method for brain MRI images. The domain of a MRI slice is regarded as the universal set. Four feature images; grayscale, local entropy, local contrast and local standard deviation are extracted from the slice and transformed into the binary domain. Each feature image is regarded as a set enclosed by the universal set. Four qualities attribute; lightness, contrast, sharpness and texture details are described by four different combinations of the feature sets. In an ideal MRI slice the four feature sets are identically equal. The degree of distortion in real MRI slice is quantified by the fidelity between the sets that describe a quality attribute. Noise is the fifth quality attribute and it is described by the slice Euler number region property. The total quality score is the weighted sum of the five quality scores. Our proposed method addresses the current challenges in image quality evaluation. It is simple, easy-to-use and easy-to-understand. Incorporation of binary transformation in the proposed method reduces computational as well as operational complexity of the algorithm. We provide experimental results that demonstrate the efficacy of our proposed method on good quality images and on common distortions in MRI images of the brain.

show abstract

“…The first method requires an estimate of the noise level in the slice. The contribution in [67] adopt the theoretical principles stated in Eq. 22, Eq.…”

Section: Bayes Theoremmentioning

confidence: 99%

Bayesian framework inspired no-reference region-of-interest quality measure for brain MRI images

et al. 2017

Self Cite

View full text Add to dashboard Cite

We describe a new post-acquisition quality assessment method for brain MRI images. It is based on Bayesian theory. Entropy is the event of interest. The low and the high entropy regions of a slice are partitions of image quality sample space. Three separate observations of image quality are contrast, standard deviation and details feature images of the original slice. Two separate quality measures are defined as the posterior probability of an entropy region given corresponding region in an observation of image quality. Prior belief in each entropy region is determined by the normalized total clique potential (TCP) energy of the slice. For TCP below predefined threshold the prior probability is determined from a quality model described by normal distribution of percentage composition of low and high entropy regions. The quality model was built from 250 MRI volume data. The data was made available by the Alzheimers disease neuroimaging initiative (ADNI). For TCP above threshold the prior is computed using a TCP-Noise level model. Total and global quality measures are weighted sums of quality measures from each regions and observations, respectively. Experimental results demonstrate good correlation with subjective opinions of radiologists for different types and levels of quality distortion.

show abstract

The clique potential of Markov random field in a random experiment for estimation of noise levels in 2D brain MRI

Cited by 6 publications

References 28 publications

Four‐neighborhood clique kernel: A general framework for Bayesian and variational techniques of noise reduction in magnetic resonance images of the brain

Four‐neighborhood clique kernel: A general framework for Bayesian and variational techniques of noise reduction in magnetic resonance images of the brain

No‐reference quality measure in brain MRI images using binary operations, texture and set analysis

Bayesian framework inspired no-reference region-of-interest quality measure for brain MRI images

Contact Info

Product

Resources

About