Computer-aided diabetic retinopathy detection using trace transforms on digital fundus images.

Diabetic retinopathy (DR) is a leading cause of vision loss among diabetic patients in developed countries. Early detection of occurrence of DR can greatly help in effective treatment. Unfortunately, symptoms of DR do not show up till an advanced stage. To counter this, regular screening for DR is essential in diabetic patients. Due to lack of enough skilled medical professionals, this task can become tedious as the number of images to be screened becomes high with regular screening of diabetic patients. An automated DR screening system can help in early diagnosis without the need for a large number of medical professionals. To improve detection, several pattern recognition techniques are being developed. In our study, we used trace transforms to model a human visual system which would replicate the way a human observer views an image. To classify features extracted using this technique, we used support vector machine (SVM) with quadratic, polynomial, radial basis function kernels and probabilistic neural network (PNN). Genetic algorithm (GA) was used to fine tune classification parameters. We obtained an accuracy of 99.41 and 99.12% with PNN-GA and SVM quadratic kernels, respectively.

Automated diagnosis of mammogram images of breast cancer using discrete wavelet transform and spherical wavelet transform features: a comparative study.

Mammograms are one of the most widely used techniques for preliminary screening of breast cancers. There is great demand for early detection and diagnosis of breast cancer using mammograms. Texture based feature extraction techniques are widely used for mammographic image analysis. In specific, wavelets are a popular choice for texture analysis of these images. Though discrete wavelets have been used extensively for this purpose, spherical wavelets have rarely been used for Computer-Aided Diagnosis (CAD) of breast cancer using mammograms. In this work, a comparison of the performance between the features of Discrete Wavelet Transform (DWT) and Spherical Wavelet Transform (SWT) based on the classification results of normal, benign and malignant stage was studied. Classification was performed using Linear Discriminant Classifier (LDC), Quadratic Discriminant Classifier (QDC), Nearest Mean Classifier (NMC), Support Vector Machines (SVM) and Parzen Classifier (ParzenC). We have obtained a maximum classification accuracy of 81.73% for DWT and 88.80% for SWT features using SVM classifier.

Application of higher order spectra for accurate delineation of atrial arrhythmia.

The electrocardiogram (ECG) is being commonly used as a diagnostic tool to distinguish different types of atrial tachyarrhythmias. The inherent complexity and mechanistic and clinical inter-relationships often brings about diagnostic difficulties to treating physicians and primary health care professionals creating frequent misdiagnoses and cross classifications using visual criteria. The current paper presents a methodology for ECG based pattern analysis for detection of atrial flutter, atrial fibrillation and normal sinus rhythm beats. ECG is an inherently non-linear and non-stationary signal; its variation may contain indicators of current disease, or warnings about impending cardiac diseases. Routinely used time domain and frequency domain methods will not be able to capture the hidden information present in the ECG beats. In the present study, we have used non-linear features of higher order spectra (HOS) to differentiate the normal, atrial fibrillation and atrial flutter ECG beats. The bispectrum features were subjected to independent component analysis (ICA) for data reduction. The ICA coefficients were subsequently subjected to K-nearest-neighbor (KNN), classification and regression tree (CART) and neural network (NN) classifiers to evaluate the best automated classifier. We have obtained an average accuracy of 97.65%, sensitivity and specificity of 98.75% and 99.53% respectively using ten-fold cross validation. Overall, the results show that application of higher order spectra statistics is useful for the classification of atrial tachyarrhythmias with reasonably high accuracies. Further validation of the proposed technique will yield acceptable results for clinical implementation.

Automated detection of optic disk in retinal fundus images using intuitionistic fuzzy histon segmentation.

The human eye is one of the most sophisticated organs, with perfectly interrelated retina, pupil, iris cornea, lens, and optic nerve. Automatic retinal image analysis is emerging as an important screening tool for early detection of eye diseases. Uncontrolled diabetic retinopathy (DR) and glaucoma may lead to blindness. The identification of retinal anatomical regions is a prerequisite for the computer-aided diagnosis of several retinal diseases. The manual examination of optic disk (OD) is a standard procedure used for detecting different stages of DR and glaucoma. In this article, a novel automated, reliable, and efficient OD localization and segmentation method using digital fundus images is proposed. General-purpose edge detection algorithms often fail to segment the OD due to fuzzy boundaries, inconsistent image contrast, or missing edge features. This article proposes a novel and probably the first method using the Attanassov intuitionistic fuzzy histon (A-IFSH)-based segmentation to detect OD in retinal fundus images. OD pixel intensity and column-wise neighborhood operation are employed to locate and isolate the OD. The method has been evaluated on 100 images comprising 30 normal, 39 glaucomatous, and 31 DR images. Our proposed method has yielded precision of 0.93, recall of 0.91, F-score of 0.92, and mean segmentation accuracy of 93.4%. We have also compared the performance of our proposed method with the Otsu and gradient vector flow (GVF) snake methods. Overall, our result shows the superiority of proposed fuzzy segmentation technique over other two segmentation methods.

Pectoral muscle segmentation: a review.

Mammograms are X-ray images of breasts which are used to detect breast cancer. The pectoral muscle is a mass of tissue on which the breast rests. During routine mammographic screenings, in medio-lateral oblique (MLO) views, the pectoral muscle turns up in the mammograms along with the breast tissues. The pectoral muscle has to be segmented from the mammogram for an effective automated computer aided diagnosis (CAD). This is due to the fact that pectoral muscles have pixel intensities and texture similar to that of breast tissues which can result in awry CAD results. As a result, a lot of effort has been put into the segmentation of pectoral muscles and finding its contour with the breast tissues. To the best of our knowledge, currently there is no definitive literature available which provides a comprehensive review about the current state of research in this area of pectoral muscle segmentation. We try to address this shortcoming by providing a comprehensive review of research papers in this area. A conscious effort has been made to avoid deviating into the area of automated breast cancer detection.

Computer-aided breast cancer detection using mammograms: a review.

The American Cancer Society (ACS) recommends women aged 40 and above to have a mammogram every year and calls it a gold standard for breast cancer detection. Early detection of breast cancer can improve survival rates to a great extent. Inter-observer and intra-observer errors occur frequently in analysis of medical images, given the high variability between interpretations of different radiologists. Also, the sensitivity of mammographic screening varies with image quality and expertise of the radiologist. So, there is no golden standard for the screening process. To offset this variability and to standardize the diagnostic procedures, efforts are being made to develop automated techniques for diagnosis and grading of breast cancer images. A few papers have documented the general trend of computer-aided diagnosis of breast cancer, making a broad study of the several techniques involved. But, there is no definitive documentation focusing on the mathematical techniques used in breast cancer detection. This review aims at providing an overview about recent advances and developments in the field of Computer-Aided Diagnosis (CAD) of breast cancer using mammograms, specifically focusing on the mathematical aspects of the same, aiming to act as a mathematical primer for intermediates and experts in the field.

Compressed sampling for heart rate monitoring.

For the first time compressed sampling (CS) has been applied to heart rate (HR) measurements. The signals can be reconstructed from samples far below the Nyquist rate with negligible small errors, a sampling reduction of 8 has been demonstrated in the paper. As a result, the bitrate of the CS sampler is half when compared to a normal sampler. A lower bitrate leads to a reduction in power consumption for HR measurement devices.

Automated oral cancer identification using histopathological images: a hybrid feature extraction paradigm.

Oral cancer (OC) is the sixth most common cancer in the world. In India it is the most common malignant neoplasm. Histopathological images have widely been used in the differential diagnosis of normal, oral precancerous (oral sub-mucous fibrosis (OSF)) and cancer lesions. However, this technique is limited by subjective interpretations and less accurate diagnosis. The objective of this work is to improve the classification accuracy based on textural features in the development of a computer assisted screening of OSF. The approach introduced here is to grade the histopathological tissue sections into normal, OSF without Dysplasia (OSFWD) and OSF with Dysplasia (OSFD), which would help the oral onco-pathologists to screen the subjects rapidly. The biopsy sections are stained with H&E. The optical density of the pixels in the light microscopic images is recorded and represented as matrix quantized as integers from 0 to 255 for each fundamental color (Red, Green, Blue), resulting in a M×N×3 matrix of integers. Depending on either normal or OSF condition, the image has various granular structures which are self similar patterns at different scales termed "texture". We have extracted these textural changes using Higher Order Spectra (HOS), Local Binary Pattern (LBP), and Laws Texture Energy (LTE) from the histopathological images (normal, OSFWD and OSFD). These feature vectors were fed to five different classifiers: Decision Tree (DT), Sugeno Fuzzy, Gaussian Mixture Model (GMM), K-Nearest Neighbor (K-NN), Radial Basis Probabilistic Neural Network (RBPNN) to select the best classifier. Our results show that combination of texture and HOS features coupled with Fuzzy classifier resulted in 95.7% accuracy, sensitivity and specificity of 94.5% and 98.8% respectively. Finally, we have proposed a novel integrated index called Oral Malignancy Index (OMI) using the HOS, LBP, LTE features, to diagnose benign or malignant tissues using just one number. We hope that this OMI can help the clinicians in making a faster and more objective detection of benign/malignant oral lesions.

Systems engineering principles for the design of biomedical signal processing systems.

Systems engineering aims to produce reliable systems which function according to specification. In this paper we follow a systems engineering approach to design a biomedical signal processing system. We discuss requirements capturing, specification definition, implementation and testing of a classification system. These steps are executed as formal as possible. The requirements, which motivate the system design, are based on diabetes research. The main requirement for the classification system is to be a reliable component of a machine which controls diabetes. Reliability is very important, because uncontrolled diabetes may lead to hyperglycaemia (raised blood sugar) and over a period of time may cause serious damage to many of the body systems, especially the nerves and blood vessels. In a second step, these requirements are refined into a formal CSP‖ B model. The formal model expresses the system functionality in a clear and semantically strong way. Subsequently, the proven system model was translated into an implementation. This implementation was tested with use cases and failure cases. Formal modeling and automated model checking gave us deep insight in the system functionality. This insight enabled us to create a reliable and trustworthy implementation. With extensive tests we established trust in the reliability of the implementation.

Analysis and automatic identification of sleep stages using higher order spectra.

Electroencephalogram (EEG) signals are widely used to study the activity of the brain, such as to determine sleep stages. These EEG signals are nonlinear and non-stationary in nature. It is difficult to perform sleep staging by visual interpretation and linear techniques. Thus, we use a nonlinear technique, higher order spectra (HOS), to extract hidden information in the sleep EEG signal. In this study, unique bispectrum and bicoherence plots for various sleep stages were proposed. These can be used as visual aid for various diagnostics application. A number of HOS based features were extracted from these plots during the various sleep stages (Wakefulness, Rapid Eye Movement (REM), Stage 1-4 Non-REM) and they were found to be statistically significant with p-value lower than 0.001 using ANOVA test. These features were fed to a Gaussian mixture model (GMM) classifier for automatic identification. Our results indicate that the proposed system is able to identify sleep stages with an accuracy of 88.7%.

Automatic identification of epileptic and background EEG signals using frequency domain parameters.

The analysis of electroencephalograms continues to be a problem due to our limited understanding of the signal origin. This limited understanding leads to ill-defined models, which in turn make it hard to design effective evaluation methods. Despite these shortcomings, electroencephalogram analysis is a valuable tool in the evaluation of neurological disorders and the evaluation of overall cerebral activity. We compared different model based power spectral density estimation methods and different classification methods. Specifically, we used the autoregressive moving average as well as from Yule-Walker and Burg's methods, to extract the power density spectrum from representative signal samples. Local maxima and minima were detected from these spectra. In this paper, the locations of these extrema are used as input to different classifiers. The three classifiers we used were: Gaussian mixture model, artificial neural network, and support vector machine. The classification results are documented with confusion matrices and compared with receiver operating characteristic curves. We found that Burg's method for spectrum estimation together with a support vector machine classifier yields the best classification results. This combination reaches a classification rate of 93.33%, the sensitivity is 98.33% and the specificy is 96.67%.

Cardiac health diagnosis using higher order spectra and support vector machine.

The Electrocardiogram (ECG) is an important bio-signal representing the sum total of millions of cardiac cell depolarization potentials. It contains important insight into the state of health and nature of the disease afflicting the heart. Heart rate variability (HRV) refers to the regulation of the sinoatrial node, the natural pacemaker of the heart by the sympathetic and parasympathetic branches of the autonomic nervous system. The HRV signal can be used as a base signal to observe the heart's functioning. These signals are non-linear and non-stationary in nature. So, higher order spectral (HOS) analysis, which is more suitable for non-linear systems and is robust to noise, was used. An automated intelligent system for the identification of cardiac health is very useful in healthcare technology. In this work, we have extracted seven features from the heart rate signals using HOS and fed them to a support vector machine (SVM) for classification. Our performance evaluation protocol uses 330 subjects consisting of five different kinds of cardiac disease conditions. We demonstrate a sensitivity of 90% for the classifier with a specificity of 87.93%. Our system is ready to run on larger data sets.

Cardiac health diagnosis using data fusion of cardiovascular and haemodynamic signals.

The electrocardiogram (ECG) is a representative signal containing useful information about the condition of the heart. The shape and size of the P-QRS-T wave, the r-r interval, etc., may help to identify the nature of disease afflicting the heart. However, human observer cannot directly monitor these subtle details and it is difficult to evaluate the cardiac health using ECG alone. Hence, the fusion of ECG, blood pressure, saturated oxygen content and respiratory data for achieving improved clinical diagnosis of patients in cardiac care units. In this study, a computer based analysis and display of the heterogeneous signals for the detection of life threatening states is demonstrated using fuzzy logic based data fusion. And to evaluate the severity of the disease a new parameter, deterioration index is proposed and results are tabulated for various cases.

Computer-based classification of eye diseases.

Eye disorders among the elderly are a major health problem. With advancing age, the normal function of eye tissues decreases and there is an increased incidence of ocular pathology. The most common causes of age related eye disorder and visual impairment in the elderly are cataracts, iridocyclitis and corneal haze. Iridocyclitis is an inflammation of the iris (the colored part of the eye), while corneal haze is a complication of refractive surgery characterized by the cloudiness of the normally clear cornea. Computer-based intelligent system for classification of these eye diseases is very useful in diagnostics and disease management. This paper presents a comparison of three classification strategies to classify four kinds of eye data sets (three different kinds of eye diseases and a normal class). Our protocol uses three different kinds of classifiers: artificial neural network, fuzzy classifier and neuro-fuzzy classifier. Features are extracted from these raw images which are then fed to these classifiers. These classifiers are run on a database of 135 subjects using the cross-validation strategy. We demonstrate a sensitivity of more than 85% for these classifiers with the specificity of 100% and results are very promising.

Elman neural networks for dynamic modeling of epileptic EEG.

In this paper, autoregressive modeling technique and neural network based modeling techniques are used to model and simulate electroencephalogram (EEG) signals. EEG signal modeling is used as a tool to identify pathophysiological EEG changes potentially useful in clinical diagnosis. The normal, background and epileptic EEG signals are modeled and the dynamical properties of the actual and modeled signals are compared. Chaotic invariants like correlation dimension (D(2)), largest Lyapunov exponent (lambda(1), Hurst exponent (H) and Kolmogorov entropy (K) are used to characterize the dynamical properties of the actual and modeled signals. Our study showed that the dynamical properties of the EEG signal modeled using neural network (NN) techniques are very similar to that of the signal.

Data fusion of multimodal cardiovascular signals.

The electrocardiogram (ECG) is a representative signal containing useful information about the condition of the heart. The shape and size of the P-QRS-T wave, the R-R interval etc. may help to identify the nature of disease afflicting the heart. However, human observer can not directly monitor these subtle details. Hence, the fusion of ECG, blood pressure, saturated oxygen content and respiratory data for achieving improved clinical diagnosis of patients in cardiac care units. Therefore, computer based analysis and display, is highly useful in diagnostics. The study demonstrates the feasibility of fuzzy logic based data fusion of the heterogeneous signals for the detection of life threatening states. Important parameters are derived from multimodal data and rule based approaches have been used. Fuzzified region for various abnormality conditions have been obtained which demonstrate the efficacy of the approach in various test cases. Comprehensive pictures showing the condition of the patient in various states will help physician in making a timely assessment in an intensive care set up.

Analysis of cardiac signals using spatial filling index and time-frequency domain.

BACKGROUND: Analysis of heart rate variation (HRV) has become a popular noninvasive tool for assessing the activities of the autonomic nervous system (ANS). HRV analysis is based on the concept that fast fluctuations may specifically reflect changes of sympathetic and vagal activity. It shows that the structure generating the signal is not simply linear, but also involves nonlinear contributions. These signals are essentially non-stationary; may contain indicators of current disease, or even warnings about impending diseases. The indicators may be present at all times or may occur at random in the time scale. However, to study and pinpoint abnormalities in voluminous data collected over several hours is strenuous and time consuming. METHODS: This paper presents the spatial filling index and time-frequency analysis of heart rate variability signal for disease identification. Renyi's entropy is evaluated for the signal in the Wigner-Ville and Continuous Wavelet Transformation (CWT) domain. RESULTS: This Renyi's entropy gives lower 'p' value for scalogram than Wigner-Ville distribution and also, the contours of scalogram visually show the features of the diseases. And in the time-frequency analysis, the Renyi's entropy gives better result for scalogram than the Wigner-Ville distribution. CONCLUSION: Spatial filling index and Renyi's entropy has distinct regions for various diseases with an accuracy of more than 95%.

Simultaneous storage of patient information with medical images in the frequency domain.

Digital watermarking is a technique of hiding specific identification data for copyright authentication. Most of the medical images are compressed by joint photographic experts group (JPEG) standard for storage. The watermarking is adapted here for interleaving patient information with medical images during JPEG compression, to reduce storage and transmission overheads. The text data is encrypted before interleaving with images in the frequency domain to ensure greater security. The graphical signals are also interleaved with the image. The result of this work is tabulated for a specific example and also compared with the spatial domain interleaving.

Transmission and storage of medical images with patient information.

Digital watermarking is a technique of hiding specific identification data for copyright authentication. This technique is adapted here for interleaving patient information with medical images, to reduce storage and transmission overheads. The text data is encrypted before interleaving with images to ensure greater security. The graphical signals are interleaved with the image. Two types of error control-coding techniques are proposed to enhance reliability of transmission and storage of medical images interleaved with patient information. Transmission and storage scenarios are simulated with and without error control coding and a qualitative as well as quantitative interpretation of the reliability enhancement resulting from the use of various commonly used error control codes such as repetitive, and (7,4) Hamming code is provided.