Mehmet Engin

ECG beat classification using neuro-fuzzy network

In this paper we have studied the application on the fuzzy-hybrid neural network for electrocardiogram (ECG) beat classification. Instead of original ECG beat, we have used; autoregressive model coefficients, higher-order cumulant and wavelet transform variances as features. Tested with MIT/BIH arrhytmia database, we observe significant performance enhancement using proposed method.

Early prostate cancer diagnosis by using artificial neural networks and support vector machines

The aim of this study is to design a classifier based expert system for early diagnosis of the organ in constraint phase to reach informed decision making without biopsy by using some selected features. The other purpose is to investigate a relationship between BMI (body mass index), smoking factor, and prostate cancer. The data used in this study were collected from 300 men (100: prostate adenocarcinoma, 200: chronic prostatism or benign prostatic hyperplasia). Weight, height, BMI, PSA (prostate specific antigen), Free PSA, age, prostate volume, density, smoking, systolic, diastolic, pulse, and Gleason score features were used and independent sample t-test was applied for feature selection. In order to classify related data, we have used following classifiers; scaled conjugate gradient (SCG), Broyden-Fletcher-Goldfarb-Shanno (BFGS), and Levenberg-Marquardt (LM) training algorithms of artificial neural networks (ANN) and linear, polynomial, and radial based kernel functions of support vector machine (SVM). It was determined that smoking is a factor increases the prostate cancer risk whereas BMI is not affected the prostate cancer. Since PSA, volume, density, and smoking features were to be statistically significant, they were chosen for classification. The proposed system was designed with polynomial based kernel function, which had the best performance (accuracy: 79%). In Turkish Family Health System, family physician to whom patients are applied firstly, would contribute to extract the risk map of illness and direct patients to correct treatments by using expert system such proposed.

The classification of human tremor signals using artificial neural network

Tremor is an involuntary movement characterized by regular or irregular oscillations of one or several body segments. Physiological and pathological tremor in motor control can be defined as roughly sinusoidal movements with particular amplitude and frequency profiles. The electrophysiological analysis of human tremor has a long tradition. Tremor time series belongs to stochastic signals. This because the mechanism of generating them is so complex and exposed to so many uncontrollable influence that mathematical equations describing them contain random quantities. In this study, we concerned with tremor classification for the purpose of medical diagnosis. Accelerometer based tremor signals belong to Parkinsonian, essential, and healthy subjects were considered for this aim. Following features were extracted from tremor signals for classification by artificial neural network (ANN); linear prediction coefficients, wavelet transform detail coefficients, wavelet transform based entropy and variance, power ratio, and higher-order cumulants. Scaled-conjugate (SCG) and BFGS (Broyden-Fletcher-Goldfarb-Shanno) gradient learning algorithms were used. Despite BFGS algorithm had more sensitivity value (92.27%), SCG algorithm had more specificity value (89.01%). According to overall performance, BFGS algorithm (91.02%) was better than SCG algorithm (88.48%).

Wavelet Transformation Based Watermarking Technique for Human Electrocardiogram (ECG)

Nowadays, watermarking has become a technology of choice for a broad range of multimedia copyright protection applications. Watermarks have also been used to embed prespecified data in biomedical signals. Thus, the watermarked biomedical signals being transmitted through communication are resistant to some attacks. This paper investigates discrete wavelet transform based watermarking technique for signal integrity verification in an Electrocardiogram (ECG) coming from four ECG classes for monitoring application of cardiovascular diseases. The proposed technique is evaluated under different noisy conditions for different wavelet functions. Daubechies (db2) wavelet function based technique performs better than those of Biorthogonal (bior5.5) wavelet function. For the beat-to-beat applications, all performance results belonging to four ECG classes are highly moderate.

Fuzzy-Hybrid Neural Network Based ECG Beat Recognition Using Three Different Types of Feature Sets

This paper represents the application on the fuzzy-hybrid neural network for electrocardiographic (ECG) beat recognition and classification. We proposed the approach to ECG beat recognition that is based on QRS complex classification. Instead of the original QRS waveform, we have used three different types of QRS features sets. The main objective of this work was to develop a technique which was less sensitive to the morphological variation of the QRS waveform. Linear predictive coefficients (LPC), third-order cumulant-based auto regressive coefficient (AR), and the variance of the wavelet transform detail coefficients of the isolated QRS complexes are used as the features. It will be shown that the wavelet transform based approach is better than the other techniques. The main properties of the proposed method are simplicity, moderate recognition rate, and fast computation time.

Portable heart rate monitoring system

Telemedicine is producing a great impact in the monitoring of patients located in non-clinical environments such as homes, gymnasiums, schools, remote military bases, ships, and rural area. A number of applications, ranging from data collection to chronic patient monitoring, and even to the control of therapeutic procedures, are being implemented in many parts of the world. As part of this growing trend, this paper explains the design of a portable heart rate monitoring system. A prototype system consists of analog data acquisition and a module which has a memory for recording heart rate values and corresponding time sequences. This module connected to computer via serial data communication protocol. At the computer side, we use interface software which is enable to graphically display the recorded heart rate values.

The evaluation of EEG response to photic stimulation in normal and diseased subjects

In this paper, our aim is to determine two photic stimulation frequencies, which would represent normal and diseased subjects, separately. Following features were extracted for this aim; linear prediction coefficients (LPC), subband wavelet entropy (SWE), subband wavelet variance (SWV), and relative power (RP). After extracting related features, analysis of variance (ANOVA) statistical test was used for the statistical evaluation of these features. According to the obtained results, wavelet transform-based entropy gave the best results to determine the representing stimulation frequencies. As a result, 29 Hz stimulation frequency was determined as the most representative frequency for normal subjects, whereas 8 Hz stimulation frequency was determined as the most representative frequency for diseased subjects.

ECG-Late Potential Extraction Using Averaged Singular—Values of Third-Order Cumulant (TOC) Based Bispectrum

The bispectrum can suppress Gaussian activity and extracts signals arising from non-linear process. We used bispectral analysis to detect diagnostically important low level signals, that are masked by background electrocardiogram (ECG) activity. The detecting performance is tested by Z-statistical test.

Deep tissue near-infrared imaging for vascular network analysis

Subcutaneous vein network plays important roles to maintain microcirculation that is related to some diagnostic aspects. Despite developments of optical imaging technologies, still the difficulties about deep skin vascular imaging have been continued. On the other hand, since hemoglobin concentration of human blood has key role in the veins imaging by optical manner, the used wavelength in vascular imaging, must be chosen considering absorption of hemoglobin. In this research, we constructed a near infrared (NIR) light source because of lower absorption of hemoglobin in this optical region. To obtain vascular image, reflectance geometry was used. Next, from recorded images, vascular network analysis, such as calculation of width of vascular of interest and complexity of selected region were implemented. By comparing with other modalities, we observed that proposed imaging system has great advantages including nonionized radiation, moderate penetration depth of 0.5–3mm and diameter of 1mm, cost-effective a...

A multiresolution approach for enhancement and denoising of microscopy images

In order to overcome blurring due to microscope optics in fluorescence microscopy, we propose a wavelet transform-based non-iterative blind deconvolution method. In our proposed deconvolution algorithm, we used wavelet-based denoising algorithms. We compared discrete wavelet transform (DWT) and wavelet packet transform (WPT) structures as denoising algorithms. WPT-based algorithm resulted in less error than the DWT-based algorithm. Minimum error was obtained for coif5 wavelet type. We compared our denoising methods with several standard denoising methods. Also, we compared our proposed deconvolution algorithm with several standard deconvolution methods. Our proposed wavelet transform-based deconvolution method resulted in the least error compared to other methods. To test the efficacy of our deconvolution method on cell images, we proposed a wavelet entropy-based non-reference image quality (contrast enhancement) metric. We tested our proposed metric by increasing blurring ratio both for noiseless and noisy images. Our metric is useful for evaluating image quality in terms of deblurring.