Automated lesion detection and lesion quantitation in MR images using autoassociative memory.

Complex-valued Hopfield Associative Memory (CHAM) can store multi-valued patterns. But CHAM stores not only given training patterns but also many spurious patterns, such as their rotated patterns, at the same time. These rotated patterns and spurious patterns reduce the noise robustness of the CHAM. In the present work, we propose Complex-valued Bidirectional Auto-Associative Memory (CBAAM) as a model of auto-associative memory which improves the noise robustness. CBAAM consists of two layers. Although the structure of CBAAM is a Bidirectional Associative Memory (BAM), CBAAM works as an auto-associative memory, because the one layer is a visible layer and the other one is an invisible layer. The visible layer consists of complex-valued neurons and can process multi-valued patterns. The invisible layer consists of real-valued neurons and can reduce pseudo-memory such as rotated patterns. Thus, CBAAM has strong noise robustness. In the computer simulations, we show that the noise robustness of CBAAM highly exceeds that of CHAM. Especially, we find that CBAAM maintains high noise robustness independent of the resolution factor.

The use of autoassociative memory to store nonorthogonal nonrandom data is discussed. The basic model of autoassociative memory examined is the Hopfield network. Hopfield networks are content-addressable memories processing all the emergent properties. A description of associative memory and a proof that it can also be used as a content-addressable memory, without any further computation than standard, are given. The results of simulations demonstrate an improved behavior when associative memory is used as content-addressable memory instead of Hopfield networks. The final model, a combination of Hopfield networks and associative memory, is discussed in detail, and a model of shared content-addressable memory (SCAM) based on this model is discussed. The problem of storing the same pattern twice and at the same time is addressed, and a solution is proposed to the problem of associating two patterns of activity in an autoassociative memory. >

We propose an evolutionary computation (EC) based constrained optimization approach for parameter tuning of an extended autoassociative memory. Being motivated to evaluate the capacity of the conventional autoassociative memory model and to go beyond the bound, we developed a series of extended models which have more parameters to increase the degree of flexibility. Meanwhile, optimization of these parameters has also become more difficult to maximize the performance of such models. By the way, we developed a new EC-based constrained optimization method in which all the constraints can be handled effectively by using the so-called “feasibilization operations” in a previous study. Now, we attempt to apply it to the optimization problem of the autoassociative memory.

BackgroundAccurate and precise detection of brain lesions on MR images (MRI) is paramount for accurately relating lesion location to impaired behavior. In this paper, we present a novel method to automatically detect brain lesions from a T1-weighted 3D MRI. The proposed method combines the advantages of both unsupervised and supervised methods.MethodsFirst, unsupervised methods perform a unified segmentation normalization to warp images from the native space into a standard space and to generate probability maps for different tissue types, e.g., gray matter, white matter and fluid. This allows us to construct an initial lesion probability map by comparing the normalized MRI to healthy control subjects. Then, we perform non-rigid and reversible atlas-based registration to refine the probability maps of gray matter, white matter, external CSF, ventricle, and lesions. These probability maps are combined with the normalized MRI to construct three types of features, with which we use supervised methods to train three support vector machine (SVM) classifiers for a combined classifier. Finally, the combined classifier is used to accomplish lesion detection.ResultsWe tested this method using T1-weighted MRIs from 60 in-house stroke patients. Using leave-one-out cross validation, the proposed method can achieve an average Dice coefficient of 73.1 % when compared to lesion maps hand-delineated by trained neurologists. Furthermore, we tested the proposed method on the T1-weighted MRIs in the MICCAI BRATS 2012 dataset. The proposed method can achieve an average Dice coefficient of 66.5 % in comparison to the expert annotated tumor maps provided in MICCAI BRATS 2012 dataset. In addition, on these two test datasets, the proposed method shows competitive performance to three state-of-the-art methods, including Stamatakis et al., Seghier et al., and Sanjuan et al.ConclusionsIn this paper, we introduced a novel automated procedure for lesion detection from T1-weighted MRIs by combining both an unsupervised and a supervised component. In the unsupervised component, we proposed a method to identify lesioned hemisphere to help normalize the patient MRI with lesions and initialize/refine a lesion probability map. In the supervised component, we extracted three different-order statistical features from both the tissue/lesion probability maps obtained from the unsupervised component and the original MRI intensity. Three support vector machine classifiers are then trained for the three features respectively and combined for final voxel-based lesion classification.

A Comprehensive Study on the Applications of Artificial Intelligence for the Medical Diagnosis and Prognosis of Asthma

Chapter 4 - Lattice Algebra Approach to Endmember Determination in Hyperspectral Imagery

Multiple Sclerosis (MS) causes the central nervous system to malfunction due to inflammation surrounding nerve cells. Detection of MS at an early stage is very important to prevent progressive MS attacks. Clinical findings, cerebrospinal fluid examinations, the evoked potentials, magnetic resonance imaging (MRI) findings have an important role in the diagnosis and follow‐up of MS. However, many of the findings on MRI may indicate brain disorders other than MS. In addition, the clinical practices accepted by physicians for MS detection are very limited. In this study, a Mask R‐CNN based method in two dataset is proposed for the automatic detection of MS lesions on magnetic resonance scans. We also improved the ROI detection stage with RPN in the Mask R‐CNN to easily adapt for different lesion sizes. MS lesions in different sizes in the dataset are successfully detected with 84.90% Dice similarity rate and 87.03% precision rates using the proposed method. In addition, volumetric overlap error and lesion‐wise true positive rate are obtained as 12.97% and 73.75%, respectively. Moreover, performance tests of the use of different numbers of GPU hardware structures are also performed and the evaluation of its effects on processing speed is performed on experimental studies..

Magnetic Resonance Imaging (MRI) plays a significant role in the current characterization and diagnosis of multiple sclerosis (MS) in radiological imaging. However, early detection of MS lesions from MRI still remains a challenging problem. In the present work, an information theoretic approach to cluster the voxels in MS lesions for automatic segmentation of lesions of various sizes in multi-contrast (T1, T2, PD-weighted) MR images, is applied. For accurate detection of MS lesions of various sizes, the skull-stripped brain data are rescaled and histogram manipulated prior to mapping the multi-contrast data to pseudo-color images. For automated segmentation of multiple sclerosis (MS) lesions in multi-contrast MRI, the improved jump method (IJM) clustering method has been enhanced via edge suppression for improved segmentation of white matter (WM), gray matter (GM), cerebrospinal fluid (CSF) and MS lesions if present. From this preliminary clustering, a pseudo-color to grayscale conversion is designed to equalize the intensities of the normal brain tissues, leaving the MS lesions as outliers. Binary discrete and 8-bit fuzzy labels are then assigned to segment the MS lesions throughout the full brain. For validation of the proposed method, three brains, with mild, moderate and severe hyperintense MS lesions labeled as ground truth, were selected. The MS lesions of mild, moderate and severe categories were detected with a sensitivity of 80%, and 96%, and 94%, and with the corresponding Dice similarity coefficient (DSC) of 0.5175, 0.8739, and 0.8266 respectively. The MS lesions can also be clearly visualized in a transparent pseudo-color computer rendered 3D brain.

An autoassociative memory network is constructed by storing character pattern vectors whose components consist of a small positive number ε and 1−ε. Although its connection weights and threshold values can not be determined only by this storing condition, it is proved that the output function of the network is contractive in a region around each stored pattern, if ε is sufficiently small. This implies that the region is a domain of attraction in the network. The shape of the region is clarified in our contraction mapping analysis. In addition to this region, larger domains of attraction are also found. Any noisy pattern vector in such domains, which may have real valued components, can be recognized as one of the stored patterns. Moreover, an autoassociative memory model having large domains of attraction is proposed. This model has symmetric connection weights and is successfully applied to character pattern recognition.KeywordsConnection WeightLarge DomainSmall Positive NumberReference PatternCharacter PatternThese keywords were added by machine and not by the authors. This process is experimental and the keywords may be updated as the learning algorithm improves.

Multiple Sclerosis (MS) is a common neurological disorder in recent years. The diagnosis process of the disease starts with the accurate and precise detection of lesions from MR images. In addition, important achievements are achieved with computer aided decision support systems, which are used as an auxiliary secondary tool in the detection of MS. In this study, we present a web-based decision support system (DeepMSWeb) developed via deep learning for the detection of MS lesions on a publicly-available dataset. Mask R-CNN architecture, one of the deep learning models, is used in the infrastructure of DeepMSWeb, and the developed web application has a flexible and user-friendly interface. In addition, experimental studies are carried out with DeepMSWeb on the dataset consisting of MR images for the detection of MS lesions, and the detection accuracy of the application is supported by similarity measurement metrics. Radiologists who have experienced DeepMsWeb are confirmed that DeepMSWeb can be used as a decision support system for the detection of MS lesions. In addition, it is evaluated that DeepMs Web can be used in different screen sizes, is easy to use and is a fast as decision support tool.

We propose a face recognition scheme based on an auto-associative memory (AM) model. Two kinds of AM models are compared, namely, pseudo-inverse memory and radial basis function (RBF) network, and we found that RBF based associative memory is much more efficient. To capture substantial facial features and reduce computational complexity, we use a wavelet transform (WT) to decompose face images and choose the lowest resolution subband coefficients for face representation. Results indicate that the modular scheme yields accurate recognition on the widely used XM2VTS face database and Olivetti Research Laboratory (ORL) face database.

An auto-associative memory based method can be used to detect abnormal tissue in MR images automatically. An abnormal image vector is tested by application of an autocorrelation matrix that consists of normal image vectors. After the novelty filtering, the output corresponding to the abnormality is produced. The preliminary result and analysis showed that the method is feasible and effective.

Reduced perfusion in white matter lesions in multiple sclerosis

An automated tool for detection of FLAIR-hyperintense white-matter lesions in Multiple Sclerosis

Although depression frequently occurs in multiple sclerosis (MS), an association with cerebral pathology is unclear. This sets MS apart from other common neurological disorders. The reasons for this are explored. The literature on depression and magnetic resonance imaging (MRI) is reviewed and methodological issues are critically evaluated. Failure to demonstrate cerebral correlates of depression is in part a function of poor study design. However, the diffuse nature of cerebral demyelination creates difficulties in image analysis peculiar to MS. More research using valid psychiatric assessment procedures, high field strength MRI and automated lesion detection is needed to resolve the issue. It is premature to reject psychosocial causes at this stage.