No full-text available
To read the full-text of this research,
you can request a copy directly from the authors.
Alzheimer's detection at early stage using local measures on MRI: A comparative study on local measures
Abstract
Alzheimer's disease (AD) is a Dementia among older people which causes neurological degradation. Mild Cognitive Impairment [1] (MCI) is a condition which could progress and then become AD but is not explicitly visible in one's behavior. This paper presents a strategic approach for recognizing MCI at early stage using Magnetic Resonance Imaging (MRI). Initially Grey Matter (GM) is segmented and Local Patterns is extracted from it This study explores the ability of Local Patterns to classify between Normal, Mild Cognitive Impairment (MCI) and AD. This study is based on the fact that GM volume loss in the MCI group compared to Normal Aging and AD is greater and reports the classification accuracy of various Local Patterns. Local Graph Structure shows greater accuracy compared to other Local Patterns.
... The machine learning algorithms in brain imaging try to learn from the structural changes of the brain, such as reduced complexity and decreased size. For example, Nayaki and Varghese [54] used MRI scans from the Alzheimer's Disease Neuroimaging Initiative (ADNI) database and tested local patterns of grey matter and classified them using SVM. The task of local patterns was to encode the loss of grey matter volume for MCI and AD when compared to CN. ...
The utilisation of magnetic resonance imaging (MRI) images for the automated detection of Alzheimer’s disease has garnered significant attention in recent years. This interest stems from the progress made in machine learning techniques and the possible application of such methods in the field of diagnostics. This study aims to evaluate the performance of 16 histogram-based image texture descriptors and features extracted from 18 pre-trained convolutional neural networks in characterising brain patterns observed in 2D slices of MRI images. The primary objective is to determine the most effective feature types for this task. The characteristics were taken from the magnetic resonance imaging (MRI) dataset given by the Alzheimer’s Disease Neuroimaging Initiative (ADNI). The study involved the calculation of features on 2D axial, coronal, and sagittal slices, followed by classification using five binary machine learning algorithms. The objective was to differentiate between individuals with normal cognitive function and those diagnosed with Alzheimer’s disease. The proposed methodology additionally facilitated the identification of specific brain areas to be selected for each axis, in order to achieve optimal accuracy. This involved determining the matching feature and classifier combinations.
... PCA is applied with the extracted feature to do the feature reduction and the SVM classifier is used to study the proposed method with ADNI (Alzheimer's Disease Neuroimaging Initiative) dataset [9]. Sankara Nayaki and Abraham varghese [10] studied the local measures on MRI based on the Grey Matter (GM) loss. Different types of local patterns were extracted like Local Binary pattern, Local Quinary pattern [11], Dominant Local Binary pattern and Local Graph Structure (LGS) [12]. ...
Alzheimer’s Disease (AD) is a serious neurodegenerative and progressive disease. It annihilates memory power and other mental functionalities. Development of Computer aided diagnosis will provide a second opinion to the radiologist in the detection and diagnosis processes. So, herein a novel method has been developed to study the brain MR image to classify the normal and Alzheimer’s. The proposed method consists of two stages. In the first stage image pre-processing like image enhancement, skull stripping and region of interest extraction has been performed. In the later stage, different graph structure based methods like local graph structure, extended local graph structure and hybrid local graph structure has been experimented and reported. The proposed method is experimented using publically available database OASIS and the average accuracy is observed to be 74.84%. This proposed method will floor the approach for developing devices for detection of AD in the future. Alzheimer’s Disease (AD) is an progressive neuro degenerative disease.
... MRI scans successfully show the difference between grey and white matter. Nayaki and Varghese proposed a mild cognitive impairment (MCI) recognition approach using MRI scans [8], where grey matter (GM) was segmented and local patterns was extracted from it. However, a perfect local pattern could not be identified in their work. ...
Alzheimer's (AD) is the most common form of dementia that negatively affects memory, thinking, behavior, and intellectual abilities in one's daily life. The risk of AD increases with age 65 and older although it may appear in people who are in their 40s or 50s. AD has no current cure. There are many ongoing research studies to understanding and treating this disease better. AD, a slowly progressive brain disease, begins well before clinical symptoms emerge. A diagnosis of AD is usually made by primary care physicians based on patient's medical and family history, psychiatric history, and history of cognitive and behavioral changes. Magnetic resonance imaging (MRI) is the most common imaging modality to use to identify brain changes. In this study, MRI scans of 10 individuals without AD and 10 individuals with AD are compared and analyzed based on brain volume. Measurements included the occipital lobe and hippocampal.
Alzheimer’s disease (AD) impairs thinking and cognition while further lowering one’s quality of life. Data mining techniques are gaining traction in the medical profession for computer-assisted diagnosis, particularly in the early detection of AD. The use of structural magnetic resonance imaging (MRI) to diagnose AD is a common successful procedure. Nonlinear registration by subjects and brain tissue segmentation are two time-consuming processes that are crucial to the success of computer-aided diagnostic approaches based on structural MRI data. Several techniques displayed remarkable precision but were tested on several imaging methods with pathologically untested datasets, making it impossible to make a reasonable distinction. The proposed work aims to conduct a comparative analysis of data mining methods. The suggested method has resulted in a more optimal feature subset with better estimation accuracy.KeywordsAlzheimer’s diseaseBrainMRIFeature extractionClassificationPCAData mining
Dementia is the loss of cognitive functioning and behavioural abilities to some extent. This is a major neurocognitive disorder which is a group of symptoms caused by other different conditions. Alzheimer’s disease has considered as the most common type of dementia. Apart from that, vascular dementia (VD), Lewy body dementia (DLB), frontotemporal dementia (FTD), Parkinson’s disease dementia, normal pressure hydrocephalus (NPH), Creutzfeldt–Jakob disease and syphilis are under Dementia. The cure for this disease is yet to be found, hence recognizing the disease in early stages and delaying the progress is a highly important fact. So, the investigation of this disease will remain as an open challenge. The aim of this paper is to review biomarkers and selected machine-learning techniques that can be segregated into early detection of dementia. Various machine-learning techniques such as artificial neural networks, decision trees and support vector machine will be discussed in this paper to find a better approach to identify Dementia in early stages. Especially this paper is consisting of following sections: (i) A brief description of Dementia and each type and the global statistics; (ii) A review of various type of medical techniques to identify dementia (MRI, CT, SPECT, fMRI, PET, EEG and CSF); (iii) Pre-processing signals; (iv) A review of machine-learning techniques.
Alzheimer’s disease (AD) is the most common cause of dementia among people aged 60 years and older. Mild Cognitive impairment (MCI) is a pre-dementia condition that has been shown to have a high likelihood of progression to AD. In this prospective study evaluate the accuracy of the GM and CSF volumetry to help distinguish between patients with AD and MCI and subjects with elderly controls. This study we explored the ability of BP-ANN identify the structural changes of Grey Matter (GM), White Matter (WM) and Cerebrospinal fluid (CSF) in different groups using real MR images. The proposed approach employs morphological operations used for skull stripping and gabor filter for feature extraction. In these results we report a statistically significant trend towards accelerated GM volume loss in the MCI group compared to the NCI and AD from the MCI. We report the results of the classification accuracies on both training and test images are up to 96%.
In this correspondence, a completed modeling of the local binary pattern (LBP) operator is proposed and an associated completed LBP (CLBP) scheme is developed for texture classification. A local region is represented by its center pixel and a local difference sign-magnitude transform (LDSMT). The center pixels represent the image gray level and they are converted into a binary code, namely CLBP-Center (CLBP_C), by global thresholding. LDSMT decomposes the image local differences into two complementary components: the signs and the magnitudes, and two operators, namely CLBP-Sign (CLBP_S) and CLBP-Magnitude (CLBP_M), are proposed to code them. The traditional LBP is equivalent to the CLBP_S part of CLBP, and we show that CLBP_S preserves more information of the local structure than CLBP_M, which explains why the simple LBP operator can extract the texture features reasonably well. By combining CLBP_S, CLBP_M, and CLBP_C features into joint or hybrid distributions, significant improvement can be made for rotation invariant texture classification.
This paper proposes a novel approach to extract image features for texture classification. The proposed features are robust to image rotation, less sensitive to histogram equalization and noise. It comprises of two sets of features: dominant local binary patterns (DLBP) in a texture image and the supplementary features extracted by using the circularly symmetric Gabor filter responses. The dominant local binary pattern method makes use of the most frequently occurred patterns to capture descriptive textural information, while the Gabor-based features aim at supplying additional global textural information to the DLBP features. Through experiments, the proposed approach has been intensively evaluated by applying a large number of classification tests to histogram-equalized, randomly rotated and noise corrupted images in Outex, Brodatz, Meastex, and CUReT texture image databases. Our method has also been compared with six published texture features in the experiments. It is experimentally demonstrated that the proposed method achieves the highest classification accuracy in various texture databases and image conditions.
We describe a fully automated method for model-based tissue classification of magnetic resonance (MR) images of the brain. The method interleaves classification with estimation of the model parameters, improving the classification at each iteration. The algorithm is able to segment single- and multispectral MR images, corrects for MR signal inhomogeneities, and incorporates contextual information by means of Markov random Fields (MRF's). A digital brain atlas containing prior expectations about the spatial location of tissue classes is used to initialize the algorithm. This makes the method fully automated and therefore it provides objective and reproducible segmentations. We have validated the technique on simulated as well as on real MR images of the brain.
Local binary pattern (LBP) is sensitive to noise. Local ternary pattern (LTP) partially solves this problem by encoding the small pixel difference into a third state. The small pixel difference may be easily overwhelmed by noise. Thus, it is difficult to precisely determine its sign and magnitude. In this paper, we propose the concept of uncertain state to encode the small pixel difference. We do not care its sign and magnitude, and encode it as both 0 and 1 with equal probability. The proposed Relaxed LTP is tested on the CMU-PIE database, the extended Yale B database and the O2FN mobile face database. Superior performance is demonstrated compared with LBP and LTP.
The field of medical imaging gains its importance with increase in the need of automated and efficient diagnosis in a short period of time. Other than that, medical image retrieval system is to provide a tool for radiologists to retrieve the images similar to query image in content. Magnetic resonance imaging (MRI) is an imaging technique that has played an important role in neuroscience research for studying brain images. Classification is an important part in retrieval system in order to distinguish between normal patients and those who have the possibility of having abnormalities or tumor. In this paper, we have obtained the feature related to MRI images using discrete wavelet transformation. An advanced kernel based techniques such as Support Vector Machine (SVM) for the classification of volume of MRI data as normal and abnormal will be deployed.
The problem associated with Illumination variation is one of the major problems in image processing, pattern recognition, medical image, etc; hence there is a need to handle and deal with such variations. This paper presents a novel and efficient algorithm for images illumination normalization called Eimad-Housam Technique (EHT). EHT features are derived from a general definition of texture in a local graph neighborhood. The experiment results show the effectiveness of proposed algorithm
Inhibitory control and cognitive flexibility are two key executive functions that develop in childhood and adolescence, increasing one's capacity to respond dynamically to changing external demands and refrain from impulsive behaviours. These gains evolve in concert with significant brain development. MRI studies have identified numerous frontal and cingulate cortical areas associated with performance on inhibition tasks, but less is known about the involvement of the underlying anatomical connectivity, namely white matter. Here we used diffusion tensor imaging (DTI) to examine correlations between a DTI derived parameter, fractional anisotropy (FA) of white matter, and performance on the NEPSY-II Inhibition test (Naming, Inhibition and Switching conditions) in 49 healthy children aged 5-16 years, (20 females; 29 males). First, whole brain voxel-based analysis revealed several clusters in the frontal projections of the corpus callosum, where higher fractional anisotropy (FA) was associated with worse inhibitory performance, as well as several clusters in posterior brain regions and one in the brainstem where higher FA was associated with better cognitive flexibility (in the Switching task), suggesting a dichotomous relationship between FA and these two aspects of cognitive control. Tractography through these clusters identified several white matter tracts, which were then manual traced in native space. Pearson's correlations confirmed associations between higher FA of frontal projections of the corpus callosum with poorer inhibitory performance (independent of age), though associations with Switching were not significant. Post-hoc evaluation suggested that FA of orbital and anterior frontal projections of the corpus callosum also mediates performance differences across conditions, which may reflect differences in self-monitoring or strategy use. These findings suggest a link between the development of inhibition and cognitive control with the underlying role of white matter, and may help to identify deviations of neurobiology in adolescent psychopathology.
The field of medical imaging gains its importance with increase in the need of automated and efficient diagnosis in a short period of time. Other than that, medical image retrieval system is to provide a tool for radiologists to retrieve the images similar to query image in content. Magnetic resonance imaging (MRI) is an imaging technique that has played an important role in neuroscience research for studying brain images. Classification is an important part in retrieval system in order to distinguish between normal patients and those who have the possibility of having abnormalities or tumor. In this paper, we have obtained the feature related to MRI images using discrete wavelet transformation. An advanced kernel based techniques such as Support Vector Machine (SVM) for the classification of volume of MRI data as normal and abnormal will be deployed.
The way that Alzheimer's disease (AD) invades brain is to destroy its fundamental elements, i.e. neurons. The phenomenon of neuron destruction reflects volume changes on brain tissues such as gray matter, white matter and cerebro-spinal fluid. In the AD-related research, the volumetric analysis of hippocampus is the most extensive study. However, the segmentation and identification of the hippocampus are highly complicated and time-consuming. Therefore, we designed a MRI-based classification framework to distinguish AD's patients from normal individuals. First, 3-D volumetric features and 2-D shape features were extracted from MRI data. Afterward, Principle component analysis (PCA) was utilized to decrease the dimensions of feature space. Finally, a SVM classifier was trained for AD classification. With the proposed framework, the classification accuracy is improved from 64% or 72%, by only using 3-D volumetric features or 2-D shape features, to 84% by using both features.
Local energy pattern (LEP), a statistical histogrambased representation, is proposed for texture classification. We use normalized local oriented energies to generate local feature vectors which describe the local structures distinctively and are less sensitive to imaging conditions. Then, each local feature vector is quantized by self-adaptive quantization thresholds which are determined in the learning stage using histogram specification, and the quantized local feature vector is transformed to a number by N-nary coding which helps to preserve more structure information during vector quantization. Finally, the frequency histogram is used as the representation feature. The performance is benchmarked by material categorization on KTHTIPS and KTH-TIPS2-a databases. Our method is compared with typical statistical approaches such as basic image features (BIF), local binary pattern (LBP), local ternary pattern (LTP), completed LBP (CLBP), Weber local descriptor (WLD) and VZ algorithms (VZ-MR8 and VZ-Joint). The results show that our method is superior to other methods on KTH-TIPS2-a database, and achieving competitive performance on KTH-TIPS database. Furthermore, we extend the representation from static image to dynamic texture, and achieving favorable recognition results on UCLA dynamic texture database.
We collected MRI diffusion tensor imaging data from 80 younger (20-32 years) and 63 older (60-71 years) healthy adults. Tract-based spatial statistics (TBSS) analysis revealed that white matter integrity, as indicated by decreased fractional anisotropy (FA), was disrupted in numerous structures in older compared to younger adults. These regions displayed five distinct region-specific patterns of age-related differences in other diffusivity properties: (1) increases in both radial and mean diffusivity; (2) increases in radial diffusivity; (3) no differences in parameters other than FA; (4) a decrease in axial and an increase in radial diffusivity; and (5) a decrease in axial and mean diffusivity. These patterns suggest different biological underpinnings of age-related decline in FA, such as demyelination, Wallerian degeneration, gliosis, and severe fiber loss, and may represent stages in a cascade of age-related degeneration in white matter microstructure. This first simultaneous description of age-related differences in FA, mean, axial, and radial diffusivity requires histological and functional validation as well as analyses of intermediate age groups and longitudinal samples.
Detection of prodromal Alzheimer's disease via pattern classification of magnetic resonance imaging
- Sach Miriam
- Winkler Gerhard
The Local binary pattern approach to texture analysis-Extensions and Applications
- Topi Manenppa
Global and local gray matter loss in mild cognitive impairment and Alzheimer's disease
- Keon Chung