Search

Knee-joint sounds or vibroarthrographic (VAG) signals contain diagnostic information related to the roughness, softening, breakdown, or the state of lubrication of the articular cartilage surfaces. Objective analysis of VAG signals provides features for pattern analysis, classification, and noninvasive diagnosis of knee-joint pathology of various types. We propose parameters related to signal variability for the analysis of VAG signals, including an adaptive turns count and the variance of the mean-squared value computed during extension, flexion, and a full swing cycle of the leg, for the purpose of classification as normal or abnormal, that is, screening. With a database of 89 VAG signals, screening efficiency of up to 0.8570 was achieved, in terms of the area under the receiver operating characteristics curve, using a neural network classifier based on radial-basis functions, with all of the six proposed features. Using techniques for feature selection, the turns counts for the flexion and extension parts of the VAG signals were chosen as the top two features, leading to an improved screening efficiency of 0.9174. The proposed methods could lead to objective criteria for improved selection of patients for clinical procedures and reduce healthcare costs.

Malignant breast tumors typically appear in mammograms with rough, spiculated, or microlobulated contours, whereas most benign masses have smooth, round, oval, or macrolobulated contours. Several studies have shown that shape factors that incorporate differences as above can provide high accuracies in distinguishing between malignant tumors and benign masses based upon their contours only. However, global measures of roughness, such as compactness, are less effective than specially designed features based upon spicularity and concavity. We propose a method to derive polygonal models of contours that preserve spicules and details of diagnostic importance. We show that an index of spiculation derived from the turning functions of the polygonal models obtained by the proposed method yields better classification accuracy than a similar measure derived using a previously published method. The methods were tested with a set of 111 contours of 65 benign masses and 46 malignant tumors. A high classification accuracy of 0.94 in terms of the area under the receiver operating characteristics curve was obtained.

Fractures with partial collapse of vertebral bodies are generically referred to as “vertebral compression fractures” or VCFs. VCFs can have different etiologies comprising trauma, bone failure related to osteoporosis, or metastatic cancer affecting bone. VCFs related to osteoporosis (benign fractures) and to cancer (malignant fractures) are commonly found in the elderly population. In the clinical setting, the differentiation between benign and malignant fractures is complex and difficult. This paper presents a study aimed at developing a system for computer-aided diagnosis to help in the differentiation between malignant and benign VCFs in magnetic resonance imaging (MRI). We used T1-weighted MRI of the lumbar spine in the sagittal plane. Images from 47 consecutive patients (31 women, 16 men, mean age 63 years) were studied, including 19 malignant fractures and 54 benign fractures. Spectral and fractal features were extracted from manually segmented images of 73 vertebral bodies with VCFs. The classification of malignant vs. benign VCFs was performed using the k-nearest neighbor classifier with the Euclidean distance. Results obtained show that combinations of features derived from Fourier and wavelet transforms, together with the fractal dimension, were able to obtain correct classification rate up to 94.7% with area under the receiver operating characteristic curve up to 0.95.

PurposeVertebral compression fractures (VCFs) result in partial collapse of vertebral bodies. They usually are nontraumatic or occur with low-energy trauma in the elderly secondary to different etiologies, such as insufficiency fractures of bone fragility in osteoporosis (benign fractures) or vertebral metastasis (malignant fractures). Our study aims to classify VCFs in T1-weighted magnetic resonance images (MRI). MethodsWe used the median sagittal planes of lumbar spine MRIs from 63 patients (38 women and 25 men) previously diagnosed with VCFs. The lumbar vertebral bodies were manually segmented and statistical features of gray levels were computed from the histogram. We also extracted texture and shape features to analyze the contours of the vertebral bodies. In total, 102 lumbar VCFs (53 benign and 49 malignant) and 89 normal lumbar vertebral bodies were analyzed. The k-nearest-neighbor method, a neural network with radial basis functions, and a naive Bayes classifier were used with feature selection. We compared the classification obtained by these classifiers with the final diagnosis of each case, including biopsy for the malignant fractures and clinical and laboratory follow up for the benign fractures. ResultsThe results obtained show an area under the receiver operating characteristic curve of 0.97 in distinguishing between normal and fractured vertebral bodies, and 0.92 in discriminating between benign and malignant fractures. ConclusionsThe proposed classification methods based on shape, texture, and statistical features have provided high accuracy and may assist in the diagnosis of VCFs. Vertebral compression fractures (VCFs) involve partial collapse of vertebral bodies.Statistics of gray levels, texture features, and shape features were computed.Classification of normal vertebral bodies versus VCFs was performed.Area under the receiver operating characteristic curve AUROC=0.97 was obtained.AUROC=0.92 was obtained in classification of benign versus malignant VCFs.

This study focuses on detecting and classifying neovascularization caused by proliferative diabetic retinopathy (PDR) in retinal fundus images. Image processing methods were applied to detect retinal vessels. A fractal analysis approach based on the box-counting method was used to quantify vascular patterns in normal and abnormal cases showing neovascularization near the optic disk (NVD). Ten images including five normal cases and five neovascularization cases were analyzed. The mean fractal dimension obtained for the NVD cases was 1.66 compared to the mean value of 1.52 for the normal cases. The statistical significance of the difference was high, with a p-value of 0.0088. The results show promise for use in detecting neovascularization in retinal images caused by PDR.

Fractures with partial collapse of vertebral bodies are referred to as "vertebral compression fractures" (VCFs). VCFs can have different etiologies comprising trauma, bone failure related to osteoporosis (benign), or metastatic cancer (malignant). This paper presents a study aimed to help in the classification of VCFs in magnetic resonance imaging (MRI). We used T1-weighted MRI of the lumbar spine in the sagittal plane. Images from 63 patients diagnosed with at least one VCF were studied. Shape features as normalized compactness, convex deficiency, Hu's moments and Fourier-descriptor-based feature were extracted from 102 lumbar vertebral bodies with VCFs and 106 normal lumbar vertebral bodies. Classification was performed using different classifiers with different feature combinations. Results obtained show that shape analysis was able to correctly distinguish between normal and fractured vertebral bodies with a correct classification rate up to 90.58% and between benign and malignant with a correct classification rate up to 80.39%.

Graphical abstractDisplay Omitted HighlightsSymbolic entropy, approximate entropy, and fuzzy entropy are used to measure the irregularity of knee vibroarthrographic signals.Mean, standard deviation, and root-mean-squared value parameters of the signal envelope amplitude are computed to characterize the signal fluctuations.Gender-dependent statistics of entropy and envelope amplitude parameters are discussed. Background and objectiveInjury of knee joint cartilage may result in pathological vibrations between the articular surfaces during extension and flexion motions. The aim of this paper is to analyze and quantify vibroarthrographic (VAG) signal irregularity associated with articular cartilage degeneration and injury in the patellofemoral joint. MethodsThe symbolic entropy (SyEn), approximate entropy (ApEn), fuzzy entropy (FuzzyEn), and the mean, standard deviation, and root-mean-squared (RMS) values of the envelope amplitude, were utilized to quantify the signal fluctuations associated with articular cartilage pathology of the patellofemoral joint. The quadratic discriminant analysis (QDA), generalized logistic regression analysis (GLRA), and support vector machine (SVM) methods were used to perform signal pattern classifications. ResultsThe experimental results showed that the patients with cartilage pathology (CP) possess larger SyEn and ApEn, but smaller FuzzyEn, over the statistical significance level of the Wilcoxon rank-sum test (p

We propose the use of fuzzy membership functions to analyze images of diffuse pulmonary diseases (DPDs) based on fractal and texture features. The features were extracted from preprocessed regions of interest (ROIs) selected from high-resolution computed tomography images. The ROIs represent five different patterns of DPDs and normal lung tissue. A Gaussian mixture model (GMM) was constructed for each feature, with six Gaussians modeling the six patterns. Feature selection was performed and the GMMs of the five significant features were used. From the GMMs, fuzzy membership functions were obtained by a probability-possibility transformation and further statistical analysis was performed. An average classification accuracy of 63.5% was obtained for the six classes. For four of the six classes, the classification accuracy was superior to 65%, and the best classification accuracy was 75.5% for one class. The use of fuzzy membership functions to assist in pattern classification is an alternative to deterministic approaches to explore strategies for medical diagnosis.

Our study aimed to develop a system for computer-aided diagnosis of vertebral compression fractures (VCFs) using magnetic resonance imaging (MRI), to help in the differentiation between malignant and benign VCFs. Lumbar spine MRI was used to acquire T1-weighted images in the sagittal plane. Images from 63 consecutive patients (38 women, 25 men, mean age 62.25 ± 14.13 years) with at least one VCF diagnosis were studied. Contrast and texture features were extracted from manually segmented images of 103 vertebral bodies with VCFs. The classification of malignant vs. benign VCFs was performed using the k-nearest neighbor (KNN) classifier with the Euclidean distance. Using a KNN classifier with k=3, feature selection, and 10-fold cross-validation, we obtained a value of the area under the receiver operating characteristic curve of 0.913.

This paper presents results of statistical analysis of fractal and texture features obtained from images of diffuse pulmonary diseases (DPDs). The features were extracted from preprocessed regions of interest (ROIs) selected from high-resolution computed tomography images. The ROIs represent five different patterns of DPDs and normal lung tissues. A Gaussian mixture model (GMM) was constructed for each feature, including all patterns. For each GMM, the six classes were identified and compared with the radiological classification of the corresponding ROIs. In 78.5% of the features, the GMM provides, for at least one class, a correct classification of at least 60%. The GMM approach facilitates detailed statistical analysis of the characteristics of each feature and assists in the development of classification strategies.

In this paper, we derive forward autoregressive models to describe the stochastic process underlying stride interval series related to idiopathic Parkinson's disease. The parameters of the autoregressive model that specify pole locations in the complex z-plane were used as dominant features for the separation of gait series of healthy subjects and patients with Parkinson's disease. Based on the autoregressive parameters, linear discriminant analysis and support vector machines can provide classification accurate rates over 74% and area larger than 0.8 under the receiver operating characteristic curve. The results obtained show that the autoregressive model parameters could be useful for classification of stride series.

This chapter establishes links between the characteristics of certain epochs in a number of biomedical signals and the corresponding physiological or pathological events in the biomedical systems of concern. Derivative-based operators may be applied to detect QRS complexes in the ECG signal as well as the dicrotic notch in the carotid pulse signal. The utility of correlation and spectral density functions in the detection of rhythms and events in EEG signals are also demonstrated. Event detection is an important step that is required before we may attempt to analyze the corresponding waves or wavelets in more detail. After a specific wave of interest has been detected, isolated, and extracted, methods targeted to the expected characteristics of the event may be applied for specifically focused analysis of the corresponding physiological or pathological event. Analysis of the event is then not hindered or obscured by other events or artifacts in the acquired signal.

This chapter develops methods to study the dynamic characteristics of nonstationary and multicomponent biomedical signals and proposes schemes to apply the well-established Fourier transform and autoregressive-modeling techniques to analyze and parameterize nonstationary and multicomponent signals. Fixed or adaptive segmentation of the signals into quasistationary segments is one approach to facilitate the analysis of such signals using traditional techniques. Several approaches for segmentation have been presented in the chapter. Adaptive segmentation facilitates not only the identification of distinct and separate events at unknown time instants in the given signal, but also the characterization of events of variable duration using the same number of parameters. The chapter also presents advanced and recently developed techniques such as wavelets and time-frequency distributions that permit the analysis of nonstationary and multicomponent biomedical signals without segmentation, and blind source separation techniques that can help separate mixed and overlapping components. The results obtained using the techniques studied in this chapter provide advantages in pattern classification as well as efficient representation and analysis of signals.

We propose methods for the analysis and classification of breast masses in mammograms via fractal analysis of signatures of their contours. Fractal dimension, on its own or in conjunction with other shape factors such as fractional concavity and spiculation index, resulted in the classification of mammographic masses as benign or malignant with overall accuracies of up to 0.93 with a database of 111 contours, as indicated by the area under the receiver operating characteristics curve. Furthermore, fractal dimension was able to classify atypical spiculated benign masses more accurately than other shape factors.

We present a procedure based on one-dimensional cepstral filters in the Radon domain to extract texture elements or textons from images with (quasi-) periodic or ordered texture. With this approach, no assumption is required on the homogeneity of the texton. By applying the cepstral filter in the Radon domain, the difficulties associated with two-dimensional cepstral analysis and phase unwrapping are obviated. The necessity of a weighting function as a preprocessing step and details of wavelet extraction in the Radon domain are discussed. The method should facilitate structural analysis of ordered texture and the constituent textons.

We present region-based filters for noise removal in three-dimensional (3D) images in the (x, y, z) domain. The filters start with region growing at each voxel to determine a context-dependent region in 3D. Appropriate selection of region-growing criteria provides regions that approximate 3D objects or features present in the image. Local statistics are then used to filter noise. The method permits adaptive noise removal without degradation of edges, surfaces, or shapes of objects. Results of 3D region-based mean, median, and local linear minimum mean-squared error (LLMMSE) filtering are shown, along with results of two-dimensional (2D) region-based filtering on a slice-by-slice basis, as well as 2D and 3D fixed-neighborhood filtering. Results of 3D region-based filtering possess lower mean-squared errors (MSE) than the results of fixed-neighborhood filters in 3D or 2D. Furthermore, results of 3D region-based filtering possess MSE lower than or equal to the MSE of the results of 2D region-based filtering ...

Decomposition of local field potential (LFP) signals with different oscillatory rhythms is useful for analysis of various neuronal activities in mice. In this paper, we first removed the power-line interference with high signal fidelity by using a notch filter with infinite impulse response. Next, we applied the ensemble empirical mode decomposition (EEMD) method to separate the LFP signal into low-frequency, Delta, Theta, Beta, Gamma, Ripple, and high-frequency oscillations, in the form of different intrinsic mode functions (IMFs). The LFP signal components with different frequency bands were identified and then reconstructed from the IMFs within the same frequency range by analyzing their power spectral ratios (PSRs). Then, normalized autocorrelation functions of the resting respiratory signal and the reconstructed Delta oscillations were computed to estimate the corresponding power spectral densities by means of the Fourier transform. The results of LFP signal decomposition and oscillatory rhythm reconstruction demonstrated the effectiveness of the EEMD and PSR analysis methods. The coherence analysis results indicate that the primary periodicity peak of the Delta LFP component is definitely linked to that of resting respiration in an awake mouse. Our major contribution is to establish a novel LFP signal separation and identification procedure by combining the EEMD method with appropriate parameter setting and the power spectral analysis approach.

The purpose of this article is to propose the use of fractal and texture analysis for computer-aided diagnosis (CAD) of diffuse pulmonary diseases (DPDs) in high-resolution computed tomography (HRCT) images. We propose multiple techniques to extract features from preprocessed regions of interest (ROIs) selected to represent five radiographic patterns useful in the differential diagnosis of DPDs, as well as normal cases. First-order statistics of gray-level distribution, Haralick's and Laws' texture features, statistical information extracted from the ROIs' discrete Fourier transforms, and their fractal dimension values were used as attributes. The features were used as inputs for a k-nearest neighbor classifier (k=5). With a dataset of 3252 ROIs, correct classification rates of up to 82.62% were achieved.

We present a novel method to detect asymmetry in mammograms based upon bilateral analysis of the spatial distribution of density within paired mammographic strips. Various differential measures of spatial correlation of gray-scale values were computed with reference to the position of the nipple for a set of 128 pairs of mammograms from the Digital Database for Screening Mammography (DDSM). Features were selected by stepwise logistic regression and the leave-one-patient-out method was used for cross-validation of results. An area under the receiver operating characteristic curve of 0.87 (SE = 0.08) was achieved by using an artificial neural network classifier with radial basis functions.

Changes in retinal vessel width can be indicative of the presence of several diseases, such as diabetic retinopathy, retinopathy of prematurity (ROP), and hypertension. Accurate detection and measurement of such changes could help in computer-aided diagnosis. An increase in venular thickness is a sign of plus disease, which warrants treatment of ROP. We present image processing methods for detection, tracking, and measurement of the width of the major temporal arcade (MTA), which is the thickest branch of the venular vessels, in retinal fundus images of preterm infants. The methods include the use of Gabor filters for the detection of the blood vessels, as well as morphological image processing for tracking and measurement of the width of the MTA. The results indicate a statistically significant difference in vessel width of normal cases as compared to cases diagnosed with plus disease (p = 0.015).

Accurate detection of the major temporal arcade (MTA) in retinal fundus images could assist in localization of various anatomical features of the retina, such as the optic nerve head (ONH) and fovea, as well as in detection of certain types of retinopathy. In this paper, we present a novel automated tracking algorithm to obtain a skeleton image representing only the MTA, by detection of the vascular tree using Gabor filters, detection of the center of the ONH using phase portrait analysis, and morphological image processing. The methods were trained and tested using two independent sets of 20 retinal images each. The results were evaluated in terms of the mean distance to the closest point (MDPC) computed for each tracked MTA as compared to its corresponding hand-drawn trace. The test results indicate a low average MDCP error of approximately 2 pixels per MTA skeleton. The proposed algorithm should assist in detection and diagnosis of diseases such as proliferative diabetic retinopathy and retinopathy of prematurity, as well as in localization of the ONH and fovea.

Sounds generated due to rubbing of knee-joint surfaces may lead to a potential tool for noninvasive assessment of articular cartilage degeneration. In the work reported in the present paper, an attempt is made to perform computer-assisted auscultation of knee joints by auditory display (AD) of vibration signals (also known as vibroarthrographic or VAG signals) emitted during active movement of the leg. Two types of AD methods are considered: audification and sonification. In audification, the VAG signals are scaled in time and frequency using a time-frequency distribution to facilitate aural analysis. In sonification, the instantaneous mean frequency and envelope of the VAG signals are derived and used to synthesize sounds that are expected to facilitate more accurate diagnosis than the original signals by improving their aural quality. Auditory classification experiments were performed by two orthopedic surgeons with 37 VAG signals including 19 normal and 18 abnormal cases. Sensitivity values (correct detection of abnormality) of 31%, 44%, and 83%, and overall classification accuracies of 53%, 40%, and 57% were obtained with the direct playback, audification, and sonification methods, respectively. The corresponding d' scores were estimated to be 1.10. -0.36, and 0.55. The high sensitivity of the sonification method indicates that the technique could lead to improved detection of knee-joint abnormalities; however, additional work is required to improve its specificity and achieve better overall performance.

A novel de-noising method for improving the signal-to-noise ratio of knee-joint vibration signals (also known as vibro-arthrographic (VAG) signals) is proposed. The de-noising methods considered are based on signal decomposition techniques, such as wavelets, wavelet packets and the matching pursuit (MP) method. Performance evaluation with synthetic signals simulated with the characteristics expected of VAG signals indicates good de-noising results with the MP method. Statistical pattern classification of non-stationary signal features extracted from time-frequency distributions of 37 (19 normal and 18 abnormal) MP method-de-noised VAG signals shows a sensitivity of 83.3%, a specificity of 84.2% and an overall accuracy of 83.8%.

Homomorphic filtering using the complex cepstrum has been used in many applications of one-dimensional (1D) signal processing, notably speech and seismic signal analysis. Some work has been done in extending cepstral filtering to the 2D domain. An application that we are particularly interested in is 2D echo removal. We show in this paper that 1D cepstral filtering techniques may be applied to the projections of images in the Radon domain. The filtered projections may then be used to reconstruct an image without the echoes. Results with synthetic images are presented to illustrate the strengths of the method. The technique could find applications in analysis of textured images and images obtained via reconstruction from projections.

Interpretation of vibrations or sound signals emitted from the patellofemoral joint during movement of the knee, also known as vibroarthrography (VAG), could lead to a safe, objective, and non-invasive clinical tool for early detection, localisation, and quantification of articular cartilage disorders. In this study with a reasonably large database of VAG signals of 90 human knee joints (51 normal and 39 abnormal), a new technique for adaptive segmentation based on the recursive least squares lattice (RLSL) algorithm was developed to segment the non-stationary VAG signals into locally-stationary components; the stationary components were then modelled autoregressively, using the Burg-Lattice method. Logistic classification of the primary VAG signals into normal and abnormal signals (with no restriction on the type of cartilage pathology) using only the AR coefficients as discriminant features provided an accuracy of 68.9% with the leave-one-out method. When the abnormal signals were restricted to chondromalacia patella only, the classification accuracy rate increased to 84.5%. The effects of muscle contraction interference (MCI) on VAG signals were analysed using signals from 53 subjects (32 normal and 21 abnormal), and it was found that adaptive filtering of the MCI from the primary VAG signals did not improve the classification accuracy rate. The results indicate that VAG is a potential diagnostic tool for screening for chondromalacia patella.

Discrimination between malignant tumors and benign masses in mammograms can be difficult because of the diversity of tumor shape. In order to facilitate this discrimination, we propose a method of shape analysis based only on characterization of boundaries of tumors via parabolic modeling. The method relies on the observation that most benign masses possess smooth or macro-lobulated contours, while most malignant tumors have complex contours with spiculations, concavities, and micro-lobulations. The contours used in this work were drawn on digitized mammograms by an expert radiologist (JELD). Classification of 54 tumors as benign or malignant was achieved with an accuracy of 76 percent.

Detection of blood vessels in retinal fundus images is an important initial step in the development of systems for computer-aided diagnosis of pathologies of the eye. In this study, we perform multifeature analysis for the detection of blood vessels in retinal fundus images. The techniques implemented include multiscale vesselness measures and Gabor filters. The selection of an appropriate threshold is crucial for accurate detection of retinal blood vessels. We propose an adaptive threshold selection method for this purpose. We also propose a postprocessing technique for removal of false-positive pixels around the optic nerve head. Values of the area under the receiver operating characteristic curve of up to 0.9616 were obtained using the 20 test images of the DRIVE database.

Monitoring the openness of the major temporal arcade (MTA) and how it changes in the presence of various forms of retinopathy, such as retinopathy of prematurity (ROP), could assist in timely detection and effective treatment. We present image processing techniques, including Gabor filters and a form of the generalized Hough transform, for the detection and modeling of the MTA via a graphical user interface (GUI). An arcade angle was also obtained via the GUI. An area under the receiver operating characteristic curve of A z = 0.75 was obtained using the results of single- and dual-parabolic modeling in the discrimination of Stage 0 ROP from Stage 3 ROP; A z = 0.71 was obtained in screening for ROP (Stage 0 versus Stages 1, 2, and 3). The arcade angle provided similar results in terms of A z values.

We demonstrate methods for the detection of architectural distortion in prior mammograms of interval-cancer cases based on analysis of the orientation of breast tissue patterns in mammograms. We hypothesize that architectural distortion modifies the normal orientation of breast tissue patterns in mammographic images before the formation of masses or tumors. In the initial steps of our methods, the oriented structures in a given mammogram are analyzed using Gabor filters and phase portraits to detect node-like sites of radiating or intersecting tissue patterns. Each detected site is then characterized using the node value, fractal dimension, and a measure of angular dispersion specifically designed to represent spiculating patterns associated with architectural distortion. Our methods were tested with a database of 106 prior mammograms of 56 interval-cancer cases and 52 mammograms of 13 normal cases using the features developed for the characterization of architectural distortion, pattern classification via quadratic discriminant analysis, and validation with the leave-one-patient out procedure. According to the results of free-response receiver operating characteristic analysis, our methods have demonstrated the capability to detect architectural distortion in prior mammograms, taken 15 months (on the average) before clinical diagnosis of breast cancer, with a sensitivity of 80% at about five false positives per patient.

We demonstrate methods for the detection of architectural distortion in prior mammograms of interval-cancer cases based on analysis of the orientation of breast tissue patterns in mammograms. We hypothesize that architectural distortion modifies the normal orientation of breast tissue patterns in mammographic images before the formation of masses or tumors. In the initial steps of our methods, the oriented structures in a given mammogram are analyzed using Gabor filters and phase portraits to detect node-like sites of radiating or intersecting tissue patterns. Each detected site is then characterized using the node value, fractal dimension, and a measure of angular dispersion specifically designed to represent spiculating patterns associated with architectural distortion. Our methods were tested with a database of 106 prior mammograms of 56 interval-cancer cases and 52 mammograms of 13 normal cases using the features developed for the characterization of architectural distortion, pattern classification via quadratic discriminant analysis, and validation with the leave-one-patient out procedure. According to the results of free-response receiver operating characteristic analysis, our methods have demonstrated the capability to detect architectural distortion in prior mammograms, taken 15 months (on the average) before clinical diagnosis of breast cancer, with a sensitivity of 80% at about five false positives per patient.