Search

Background Lung region segmentation is an important stage of automated image-based approaches for the diagnosis of respiratory diseases. Manual methods executed by experts are considered the gold standard, but it is time consuming and the accuracy is dependent on radiologists’ experience. Automated methods are relatively fast and reproducible with potential to facilitate physician interpretation of images. However, these benefits are possible only after overcoming several challenges. The traditional methods that are formulated as a three-stage segmentation demonstrate promising results on normal CT data but perform poorly in the presence of pathological features and variations in image quality attributes. The implementation of deep learning methods that can demonstrate superior performance over traditional methods is dependent on the quantity, quality, cost and the time it takes to generate training data. Thus, efficient and clinically relevant automated segmentation method is desired for the diagnosis of respiratory diseases. Methods We implement each of the three stages of traditional methods using deep learning methods trained on five different configurations of training data with ground truths obtained from the 3D Image Reconstruction for Comparison of Algorithm Database (3DIRCAD) and the Interstitial Lung Diseases (ILD) database. The data was augmented with the Lung Image Database Consortium (LIDC-IDRI) image collection and a realistic phantom. A convolutional neural network (CNN) at the preprocessing stage classifies the input into lung and none lung regions. The processing stage was implemented using a CNN-based U-net while the postprocessing stage utilize another U-net and CNN for contour refinement and filtering out false positives, respectively. Results The performance of the proposed method was evaluated on 1230 and 1100 CT slices from the 3DIRCAD and ILD databases. We investigate the performance of the proposed method on five configurations of training data and three configurations of the segmentation system; three-stage segmentation and three-stage segmentation without a CNN classifier and contrast enhancement, respectively. The Dice-score recorded by the proposed method range from 0.76 to 0.95. Conclusion The clinical relevance and segmentation accuracy of deep learning models can improve though deep learning-based three-stage segmentation, image quality evaluation and enhancement as well as augmenting the training data with large volume of cheap and quality training data. We propose a new and novel deep learning-based method of contour refinement.

Data mining is an extensive area of research involving pattern discovery and feature extraction which is applied in various critical domains. In clinical aspect, data mining has emerged to assist the clinicians in early detection, diagnosis, and prevention of diseases. Advances in computational methods have led to implementation of machine learning in multi-modal clinical image analysis. One recent method is online learning where data become available in a sequential order, thus sequentially updating the best predictor for the future data at each step, as opposed to batch learning techniques which generate the best predictor by learning the entire data set at once.

Noise-based quality evaluation of MRI images is highly desired in noise-dominant environments. Current noise-based MRI quality evaluation methods have drawbacks which limit their effective performance. Traditional full-reference methods such as SNR and most of the model-based techniques cannot provide perceptual quality metrics required for accurate diagnosis, treatment and monitoring of diseases. Although techniques based on the Moran coefficients are perceptual quality metrics, they are full-reference methods and will be ineffective in applications where the reference image is not available. Furthermore, the predicted quality scores are difficult to interpret because their quality indices are not standardized. In this paper, we propose a new no-reference perceptual quality evaluation method for grayscale images such as MRI images. Our approach is formulated to mimic how humans perceive an image. It transforms noise level into a standardized perceptual quality score. Global Moran statistics is combined with local indicators of spatial autocorrelation in the form of local Moran statistics. Quality score is predicted from perceptually weighted combination of clustered and random pixels. Performance evaluation, comparative performance evaluation and validation by human observers, shows that the proposed method will be a useful tool in the evaluation of retrospectively acquired MRI images and the evaluation of noise reduction algorithms. © 2019 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (http://creativecommons.org/licenses/by/4.0/).

Background: Rician noise, bias felds and blur are the common distortions that degrade MRI images during acquisition. Blur is unique in comparison to Rician noise and bias felds because it can be introduced into an image beyond the acquisition stage such as postacquisition processing and the manifestation of pathological condi tions. Most current blur assessment algorithms are designed and validated on con sumer electronics such as television, video and mobile appliances. The few algorithms dedicated to medical images either requires a reference image or incorporate manual approach. For these reasons it is difcult to compare quality measures from diferent images and images with diferent contents. Furthermore, they will not be suitable in environments where large volumes of images are processed. In this report we propose a new blind blur assessment method for diferent types of MRI images and for diferent applications including automated environments. Methods: Two copies of the test image are generated. Edge map is extracted by sepa rately convolving each copy of the test image with two parallel diference of Gaussian flters. At the start of the multiscale representation, the initial output of the flters are equal. In subsequent scales of the multiscale representation, each flter is tuned to diferent operating parameters over the same fxed range of Gaussian scales. The flters are termed low and high energy flters based on their characteristics to successively attenuate and highlight edges over the range of multiscale representation. Quality score is predicted from the distance between the normalized mean of the edge maps at the fnal output of the flters. Results: The proposed method was evaluated on cardiac and brain MRI images. Performance evaluation shows that the quality index has very good correlation with human perception and will be suitable for application in routine clinical practice and clinical research. © The Author(s) 2018. This article is distributed under the terms of the Creative Commons Attribution 4.0 International License (http://creativecommons.org/licenses/by/4.0/)

Magnetic resonance imaging (MRI) system images are important components in the development of drugs because it can reveal the underlying pathology in diseases. Unfortunately, the processes of image acquisition, storage, transmission, processing, and analysis can influence image quality with the risk of compromising the reliability of MRI-based data. Therefore, it is necessary to monitor image quality throughout the different stages of the imaging workflow. This report describes a new approach to evaluate the quality of an MRI slice in multi-center clinical trials. The design philosophy assumes that an MRI slice, such as all natural images, possess statistical properties that can describe different levels of contrast degradation. A unique set of pixel configuration is assigned to each possible level of contrast-distorted MRI slice. Invocation of the central limit theorem results in two separate Gaussian distributions. The central limit theorem says that the mean and standard deviation of pixel configuration assigned to each possible level of contrast degradation will follow a normal distribution. The mean of each normal distribution corresponds to the mean and standard deviation of the underlying ideal image. Quality prediction processes for a test image can be summarized into four steps. The first step extracts local contrast feature image from the test image. The second step computes the mean and standard deviation of the feature image. The third step separately standardizes each normal distribution using the mean and standard deviation computed from the feature image. This gives two separate z-scores. The fourth step predicts the lightness contrast quality score and the texture contrast quality score from cumulative distribution function of the appropriate normal distribution. The proposed method was evaluated objectively on brain and cardiac MRI volume data using four different types and levels of degradation. The four types of degradation are Rician noise, circular blur, motion blur, and intensity nonuniformity also known as bias fields. Objective evaluation was validated using a proposed variation of difference of mean opinion scores. Results from performance evaluation show that the proposed method will be suitable to monitor and standardize image quality throughout the different stages of imaging workflow in large clinical trials. MATLAB implementation of the proposed objective quality evaluation method can be downloaded from (https://github.com/ezimic/Image-Quality-Evaluation)., This report describes a new approach to evaluate the quality of an MRI cardiac or brain slice in multi-center clinical trials.

The authors propose a new application-specific, post-acquisition quality evaluation method for brain magnetic resonance imaging (MRI) images. The domain of a MRI slice is regarded as universal set. Four feature images; greyscale, 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 feature sets. In an ideal MRI slice, the four feature sets are identically equal. Degree of distortion in real MRI slice is quantified by fidelity between the sets that describe a quality attribute. Noise is the fifth quality attribute and is described by the slice Euler number region property. Total quality score is the weighted sum of the five quality scores. The authors' proposed method addresses 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 and operational complexity of the algorithm. They provide experimental results that demonstrate efficacy of their proposed method on good quality images and on common distortions in MRI images of the brain. © The Institution of Engineering and Technology 2017. This is the authors’ accepted and refereed manuscript to the article.

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

Effective performance of many image processing and image analysis algorithms is strongly dependent on accurate estimation of noise level. We exploit the simplicity and similarity of statistics of human anatomy among different subjects to develop new noise level estimation algorithm for magnetic resonance images of brain. Objects of the experiment are noise-free 3D brain MRI of 422 subjects. There are 21 slices for each subject. For each slice, total clique potential (TCP) of Markov random field, computed from local clique potential, is indexed by 200 different levels of noise. The sample space is the set of TCP-noise level data of each slice. The random variable is the set of indices of noise level of TCP in each element of sample space that is closest in numerical value to TCP measured from a test MRI slice. Noise level is estimated from the mean and variance of the random variable. We also report the formulation of a generalized mathematical model describing relationship between TCP and Rician noise level in brain MRI images. Our proposal can operate in the absence of signals in the background and significantly reduce modeling errors inherent in strong parametric assumptions adopted by some of the current algorithms. © 2013 Wiley Periodicals, Inc. Int J Imaging Syst Technol, 23, 304–413, 2013

Researchers in the field of computer vision and medical imaging acknowledge that segmentation trivializes the task of bias field correction, and bias field distorts only the intensity but not the spatial attributes of an image. We exploit this knowledge to develop a new algorithm that corrects bias field in magnetic resonance images of brain. K-means algorithm generates regions of interest (ROIs) equal to the number of anatomic structures in the test image. Anatomic structural map is derived by combining entropy image from the output of multiscale edge detector with output of Otsu threshold. Spatial information from anatomic structural map aids detection of outliers in each ROI. Outliers are assigned to their appropriate tissue classes. The task of Intensity inhomogeneity correction is completed by rescaling intensity levels of voxels in each tissue class to conform to statistics of uncorrupted voxels. Performance evaluation demonstrates efficiency of our proposal for different characteristics of bias field.

Traditionally, the fields of anthropology and biomedical engineering are two diverse areas of science. Over the last few years various subfields of biomedical engineering have been trending towards personalized medicine. Presently there are a limited number of models that neuroscientists use to evaluate theory and solve application problems, thus depersonalizing medicine. This chapter introduces a novel integration of how anthropology can lend to and improve personalized neuromedicine through the study of physical charactersistics, the relationship of races, and gender. By integrating anthropometric and craniofacial data, future head models should accomdate race, gender, age, and size to better approximate personalized medicine on a wide-scale approach.

Borrowing from the fields of paleoanthropology and cranial anthropology we mathematically transform a universal head model to reflect the size of an average female of other ethnic groups. Our goal is to make deformable generic head models readily available for individuals who have not been medically imaged. We concluded that elliptical and exponential functions must be implemented to reshape the template in the frontal, occipital and parietal lobes. We have identified a few parametric values, namely the exponents and major-to-minor axis ratios to generate realistic human shapes.

We built a realistically-shaped head volume conductor model using the finite element method (FEM). We lofted Bezier splines together to reduce the computational boundary complexity while preserving the geometrical structure. We based our spline reduction on coupling axis-length profile slopes of major and minor angular distances. We found that the optimal ratio of points per spline to the number of slices per tissue ranged from 1 to 1.5. We supported the structure reduction with an adaptive mesh, using quadratic-order tetrahedral basis elements with the highest resolution at the subdomain boundaries to maintain the geometric integrity.