Showing papers presented at "Biomedical Engineering International Conference in 2012"

Automatic segmentation and degree identification in burn color images

Abstract: When burn injury occurs, the most important step is to provide treatment to the injury immediately by identifying degree of the burn which can only be diagnosed by specialists. However, specialists for burn trauma are still inadequate for some local hospitals. Hence, the invention of an automatic system that is able to help evaluating the burn would be extremely beneficial to those hospitals. The aim of this work is to develop an automatic system with the ability of providing the first assessment to burn injury from burn color images. The method used in this work can be divided into 2 parts, i.e., burn image segmentation and degree of burn identification. Burn image segmentation employs the Cr-transformation, Luv-transformation and fuzzy c-means clustering technique to separate the burn wound area from healthy skin and then mathematical morphology is applied to reduce segmentation errors. The segmentation algorithm performance is evaluated by the positive predictive value (PPV) and the sensitivity (S). Burn degree identification uses h-transformation and texture analysis to extract feature vectors and the support vector machine (SVM) is applied to identify the degree of burn. The classification results are compared with that of Bayes and K-nearest neighbor classifiers. The experimental results show that our proposed segmentation algorithm yields good results for the burn color images. The PPV and S are about 0.92 and 0.84, respectively. Degree of burn identification experiments show that SVM yields the best results of 89.29 % correct classification on the validation sets of the 4-fold cross validation. SVM also yields 75.33 % correct classification on the blind test experiment.

Human falling detection algorithm using back propagation neural network

Abstract: A fall monitor system is necessary to reduce the rate of fall fatalities in elderly people. As an accelerometer has been smaller and inexpensive, it has been becoming widely used in motion detection fields. This paper proposes the falling detection algorithm based on back propagation neural network to detect the fall of elderly people. In the experiment, a tri-axial accelerometer was attached to waists of five healthy and young people. In order to evaluate the performance of the fall detection, five young people were asked to simulate four daily-life activities and four falls; walking, jumping, flopping on bed, rising from bed, front fall, back fall, left fall and right fall. The experimental results show that the proposed algorithm can potentially distinguish the falling activities from the other daily-life activities.

Removal of blink from EEG by Empirical Mode Decomposition (EMD)

Abstract: The electroencephalographic signals (EEG) are rather weak and contaminated with different artifacts that have biological and external sources. Among these artifacts, blinks and eye movements are the most common of them. In this paper, we introduce a new method, Empirical Mode Decomposition (EMD), for removal of blink contamination from EEG signal. The proposed method is compared to a fourth order Butterworth high-pass filtering with cutoff frequency at 2 Hz. The performance index of our experiment is mean square error (MSE) between bands of pure EEG and corrected EEG. Results obtained from the analysis of contaminated EEG signal show that EMD method outperforms the high pass filtering for elimination of blink contamination from EEG. However, EMD could not be applied on-line.

Directional eye movement detection system for virtual keyboard controller

Abstract: Several researches concerning electrooculography interface for Human Computer Interface (HCI) have been developed in recent years. For applications of disabled person such as lock-in, and Motor Neuron disease, a simple and effective technology for communication is necessary. Eye blink is defined as a selection command in existing research. Problem of current research is occurred when user blinks his eye involuntarily. To resolve this problem, in this paper, we develop a new electrooculography based system for typing words via virtual keyboard by using voltage threshold algorithm. EOG signal with different direction of eye movement in horizontal and vertical directions are detected. EOG signal is measured by two channels with six electrodes. Measurement circuit consists of three major processes: instrument amplifier, filter and signal conditioning amplifier processes. These circuits filter noise out, pass frequencies in ranges of EOG signal and then amplify the signal. The voltage threshold algorithm is then used to classify the EOG signal. Selection command is defined by closing eye in a short period of use to avoid eye blink involuntary. To test the performance of method, typing rate and accuracy are measured. Typing rate on virtual keyboard 25.94 seconds/letter and its accuracy is 95.2%. The results show the feasibility of proposed method.

Diabetes diagnosis by direct measurement from urine odor using electronic nose

Abstract: Diabetes is one of the most common chronic diseases and can occur at any age. To avoid many side effects of diabetes on health, early detection of diabetes is very necessary. In this paper, we propose an alternative method to detect diabetes based on direct measurement of urine odor by using an electronic nose (E-nose). Artificial urines, used to simulated a situation of diabetes patient, were produced by adding glucose into the pure urine samples. Eight commercial chemical gas sensors were used as the sensing elements of our e-nose. Principal components analysis (PCA) and cluster analysis (CA) methods were employed for data analysis. The PCA and CA results show that the proposed technique was able to identify the glucose concentration in urine. In the future, e-nose can be a potential tool for diabetes diagnosis by healthcare personnel as well as home users.

A simply fall-detection algorithm using accelerometers on a smartphone

Abstract: Every year, several thousands of elderly people experience with falling accident. Falling is then a main problem about healthiness of elders. This paper tries to find out a simple algorithm to detect a fall. With less calculation, the device can quickly distinguish between a fall and a normal activity of daily living (ADL). As the smartphone technology is currently in very advance, it includes several sensors to come along. The sensors building in the smartphone are very useful in every field of measurements even in medical engineering. The tri-axial accelerometer is one sensor available on the smartphone and one application is to use for fall detection. From the study, the simple algorithm can be applied for fall detection by observing any change of x-, y-, or z-acceleration 10g within time limited obtaining from ADLs in terms of lying down. The advantages of using the smartphone as a fall detector are that it can alarm or call out for help. It is also getting cheap, widely used, and comfortable to use or mount.

Real-time monitoring glucose by used microwave antenna apply to biosensor

Abstract: In this paper we investigate the electromagnetic field interaction with a glucose aqueous solution using a microwave antenna (U-shape) to evaluate the glucose concentration and vary temperature. The glucose concentration vary from 10–40 mg/ml compare with DI(de-ionized) water, the operating frequency of about 1–2.5GHz. The change of the glucose concentration is directly related to the change of the reflection coefficient due to electromagnetic interaction between the dielectric wave and the glucose aqueous solution. A glucose biosensor using microwave antenna (U-shape) provides a unique approach for glucose monitoring. The antenna is designed have various formed and test by comsol program. The principles of Comsol is finite element for dispersion of electromagnetic waves from a real experiment to measure the concentration of glucose solution.

Sensor pillow system: Monitoring cardio-respiratory and posture movements during sleep

Abstract: In this paper, we have developed a sleep monitoring and gesture recognition system for patient based on polysomnography which will be useful for patient communication with healthcare personals and/or relatives. In particular, we present the sensor pillow system that employs wireless networks based on low-cost ZigBee technology and a sensor array of force sensitive resistors (FSR) based on polymer thick film (PTF) device for classifying and recognizing sleep posture. This paper also proposes a simple motion model that explains the change of the head pressure distribution. In addition, we can detect some physiological parameters during the sleep stages and wakefulness as well as record cardio-respiratory activity as related to different physiological factors. The integration of the sensor system and wireless technology with a computer software could make this healthcare monitoring system a commercial product valuable for point-of-care applications.

The design and construction of infusion pump calibrator

Abstract: The purpose of this research was to study on the design and construction of infusion pump calibrator This research has adopted the principle of physics, electronics and microcontroller The designed and constructed the infusion pump calibrator was composed of three main parts: 1)the control part consisting of micro switches for selected the operating mode, 2) signals detection part composing of load cells, pressure sensor for measuring flow rate and occlusion alarm respectively and 3) the processing and display part comprising of microcontroller AVR ATmega1280 with C Language program and LCD sequentially The results of functional testing were compared with standard infusion devices analyzer Metron Lagu showed that the average percentage error of flow rate measuring and occlusion pressure tests were 045 percent and 021 percent successively

Development of a wireless electronic shoe for walking abnormalities detection

Abstract: This work portrays the development of wireless smart shoe for gait analysis. Force sensitive resistors (FSRs) and resistive bend sensor were employed by mounting onto a shoe for detection of pressure distribution beneath the foot during normal and abnormal walking. In the present study, we have been interested in 3 walking postures including normal walking, tiptoe walking and dragging foot walking. Wireless sensor network (WSN) based on ZigBee technology was employed in this work for data communication. The Principal component analysis (PCA) was used for pattern recognition in the analysis part. It was shown that the smart shoe was successful to classify between normal gait pattern and some abnormal gait patterns.

Counting number of points for acne vulgaris using UV Fluorescence and image processing

Abstract: This paper presents counting number of points for the P.acne vulgaris using UV Fluorescence and image processing. This proposed method uses a process of image processing as follows. Cropping a UV image is to select a region of interest and, then the cropped image is resized for a suitable size and it (or the color image) is converted to a gray image. Quality of this gray image will be improved for image enhancement using adaptive histogram equalization. Finally, extended maxima transform, i.e. the regional maxima of the H-maxima transform, is used for counting number of the P.acne points. The experimental results show that the accuracy, sensitivity and precision are approximately at 83.75%, 98.22% and 85.04%, respectively. It saves time to analysis of the acne points.

Effect of experimental factors on the properties of PEDOT-NaPSS galvanostatically deposited from an aqueous micellar media for invasive electrodes

Abstract: Poly(3,4-ethylenedioxythiophene) (PEDOT) coat for invasive electrodes has been generated considerable interest in terms of electrical properties and biocompatibility. Adequate formulations and methodology are necessary to deposit polymer reproducibly on macro and micro-electrodes. Electrochemical properties of PEDOT as interface material were assessed with respect to different experimental factors. PEDOT/NaPSS was polymerized from a sodium polystyrenesulfonate (NaPSS) dispersion. The dispersion was characterized by cyclic voltammetry and its critical micellar concentration (CMC) value is reported, according to our best knowledge, for the first time for PEDOT deposition. The polymer obtained was studied using Raman spectroscopy, impedance and cyclic voltammetry measurement. PEDOT-NaPSS coat shows high charge injection (Qinj), low interfacial impedance (Z) and superficial chemical stability (Raman). This study provides a simple, constant and reproducible framework to coat invasive electrodes.

Wireless intelligent fall detection and movement classification using fuzzy logic

Abstract: Global population aging leads to increased interests in preventive healthcare technology. As falls are the most common cause of injury or death in old persons, fall detection and movement classification is one of the key topics in this research area. In this paper we propose a simple wireless intelligent system prototype for fall detection and movement classification for real-time monitoring of the elderly. The portable sensor unit acquires data from a triaxial accelerometer and sends the data wirelessly to a computer using Zigbee technology. Alternative to classic methods, the movement data is analyzed using a fuzzy inference system. The system is designed to distinguish between four movement types: standing, sitting, forward fall, and backward fall. Its classification accuracy is investigated using experimental data. It is observed that the system performs well with high sensitivity and excellent specificity. Additionally, the system is applicable for monitoring rehabilitative patients and is extendable to a larger class of movements and postures.

A wireless networked smart-shoe system for monitoring human locomotion

Abstract: This paper presents the development of a low cost wireless data shoe system for monitoring human locomotion. The sensor unit consists of 3 force sensing resistors (denoted by FSR located at ball, lateral border and heel) and 3-axis acceleration sensor (ADXL335). Pressure and acceleration data were sampled at 10 Hz, which is sufficient for various activities such as sitting, standing, walking and possibly for running. The data from these sensors were sent to a base station (via ZigBee wireless network) connected to a personal computer. Experimental results show a clear discrimination of patterns between static and dynamic postures. In addition, we can detect some gait phase from walking sample data which can be used for analysis of healthy gait behavior.

Simulator sickness in immersive virtual environment

Abstract: The simulator sickness in immersive virtual environment was studied for future questionnaire improvement. The top four sickness scores are general discomfort, eyestrain, difficulty concentrating, and fatigue. These experimental results suggest future questionnaire for immersive virtual environment to concentrate on problems about eyes and seeing. Furthermore, the least four simulator sickness are burping, increase salivation, sweating, and stomach awareness. These sickness symptoms might be reduced to shorten future simulator sickness questionnaire.

A single tooth segmentation using structural orientations and statistical textures

Abstract: Segmentation of a single tooth is important for extracting tooth's features from x-ray dental images. PCA-stacked Gabor is a method that can help in enhancement of more details of tooth texture and bone. The orientation of Gabor banks correspond to the angle of tooth's orientation. In this paper, we propose a segmentation method based on an estimation of the angle of tooth's orientation in the image. The intensity projection and peak detection method are used in PCA-stacked Gabor method in order to improve the detail of tooth texture and bone. The tooth's contour is then segmented using active contour method without edge. In the experimental results, average mean square error (AMSE) is reduced to 11.763 compared active contour without edge (ACWE) method of 15.762 and the conventional method of 12.296, respectively.

Fabrication of novel shark collagen-pectin scaffolds for tissue engineering

Abstract: Collagen-pectin scaffolds comprised of collagen derived from the skins of brown banded bamboo sharks (Chiloscyllium punctatum) and citrus pectin were fabricated for tissue engineering. Crosslinking of the scaffolds was performed by pregelation of pectin using calcium sulfate and chemical crosslinking of collagen using carbodiimide. The surface morphology of the scaffolds as determined by scanning electron microscopy (SEM) showed a porous structure with an average pore size of 134.53 ± 52.44 μm with interconnecting pores. The mechanical properties of the scaffolds were a hardness of 0.591±0.135 N, and a springiness of 0.958±0.022. Compatibility of the scaffolds was confirmed using an in vitro cytotoxicity test on C2C12 myoblast cells. It was concluded that these collagen-pectin scaffolds from a novel source of collagen could be suitable for applications in tissue engineering.

Low cost and customized plantar pressure analyzer for foot pressure image in rehabilitation foot clinic

Abstract: Plantar pressure image is generally accepted as an effective tool to analyze a foot pathology and disorder in foot care investigation for a long time. In this research, we propose a preliminary study on the design and construction of low cost and customized plantar pressure analyzer for foot pressure image in a general rehabilitation foot clinic. A single bare foot loading platform is designed to measure a spread of plantar pressure inwhich combined of 72 force sensitive sensor in a form of fix array 6×12 matrix. An appropriate GUI software is specially designed in order to fit with a clinician user true requirement in both 2D and 3D pressure distribution analysis. Finally, database system for storage and retrieving in follow-up process also well develop. Not only an interesting output obtained from this study, this research also guide us to develop more customized and appropriated foot care scanner system for using in general hospital in the future.

EEG-based mental fatigue prediction for driving application

Abstract: Mental fatigue prediction using the electroencephalogram (EEG) has widely been studied. EEG definitely changes when one feels fatigue. However, the challenge is that the accurate results of fatigue prediction are from how to select the EEG interval of interest for real-time prediction. This paper proposes a novel method for efficiently selecting the EEG signal during fatigue period. Eye-blinking (EB) signs detected via the electrooculogram (EOG) are employed as the marker. The EEG band powers are further extracted as the features. The results illustrate that the proposed marker is possible to be efficiently used to predict the mental fatigue state in real-time.

Natural interaction on 3D medical image viewer software

Abstract: A common interaction with medical 2D and 3D images is to apply windows, icons, menus and pointers (WIMP) paradigm; however, WIMP might not be convenient in some situation, such as in the operating room where sterilization becomes an issue. Thus, in this project, the controller-free natural interaction on 3D medical image viewer software through a Kinect camera will be developed to replace WIMP such that surgeons can view medical images in the operating room without contacting mouse and keyboard. The interaction will include both body and figure gestures. The 3D medical image viewer software will be based on the Visualization ToolKit (VTK) library to visualize 3D image data in the DICOM (Digital Imaging and Communications in Medicine) format. Moreover, suitable controller patterns will be designed according to practical use in the operating room.

An ultrasound imaging system prototype for raw data acquisition

Abstract: This paper presents a hardware design of an ultrasound imaging system for transmitting pulses and acquiring echo signals to/from individual elements of an array transducer. Since most of the ultrasound imaging systems in the market provide only image files after reconstruction, the presented hardware can be controlled by a computer in order to transmit pulses in specific patterns to transducer elements and in order to access the data from individual transducer elements. It benefits the research and development of new algorithms for beamforming, image reconstruction, and other applications, such as elastography and ultrasonic transmission tomography.

Artifactual component classification from MEG data using support vector machine

Abstract: Recently, an independent component analysis (ICA) has been proven to be an effective method for removing artifacts and noise in multi-channel physiological measures. ICA can extract independent component (IC) which was directly regarded as artifacts. In this paper, we propose an automatic method for classifying physiological artifacts from magnetoencephalogram (MEG) data. The artifactual ICs were classified based on support vector machine (SVM) algorithm. The following parameters: kurtosis (K), probability density (PD), central moment of frequency (CMoF), spectral entropy (SpecEn), and fractal dimension (FD) were used as input vector of SVM. The proposed method showed the average classification rates of 99.18%, 92.33%, and 98.15% for cardiac (EKG), ocular (EOG), and high-amplitude changes (HAM), respectively.

Detection of Mycobacterium tuberculosis by using loop-mediated isothermal amplification combined with a lateral flow dipstick biosensor

Abstract: Tuberculosis is a persistent problem in the developing world and the biggest cause of mortality. Loop-mediated isothermal amplification (LAMP) allows DNA to be amplified rapidly at a constant temperature. Here a LAMP method was combined with a chromatographic lateral-flow dipstick (LFD) to detect IS6110 gene of M. tuberculosis rapidly and specifically. The reaction was optimized at 65 °C for 90 min and amplified DNA hybridized to an FITC-labeled oligonucleotide probe for 5 min was detected at LFD test line 5 min after application. Excluding for the step of DNA extraction, test results could be generated within 1 h 40 min. In addition to advantages of short assay time, confirmation of amplicon identity by hybridization and elimination of electrophoresis with carcinogenic ethidium bromide, the LAMP-LFD was more sensitive than an existing PCR assay for detection of M. tuberculosis.

Design and development of touch screen based Continuous Passive Motion device for knee rehabilitation

Abstract: Continuous Passive Motion (CPM) device is commonly used for knee rehabilitation to recover the range of motion or to lessen edema and swelling of the knee following injuries or surgeries. The objectives of this work are to design and develop a touch screen based CPM device for knee rehabilitation. The device is capable of flexing/extending the knee joint in the range of motion from 0 degrees to 120 degrees with the speed of 30 degrees/minute to 150 degrees/minute. The device consists of a touch screen input/graphic color LCD, a microcontroller, a motor driver, a DC motor, an encoder, two safety sensors, an original point sensor and a mechanical part. The developed CPM device gives better benefits associated with its conventional counterparts since it has a portable size with a smooth motion of the motor system and offers the same anatomical adjustability. With the advantages of a touch screen based controller, the device also provides a user friendly interface for the controller.

On the development of EEG based emotion classification

Abstract: This paper proposes an investigation on classification of the positive and negative emotions via the use of electroencephalogram (EEG). EEG bandpowers are extracted as the feature of interest. Two simple decision rules to classify positive and negative emotions are proposed, i.e. 1) using both the left and right frontal information and 2) using only one side of the left or right frontal information. First decision reports low accuracy while the second decision rule can achieve higher accuracy between 80 to 90%. This can be concluded that the proposed method is possible for the real-time emotion classification in neuroeconomics.

Spike and epileptic seizure detection using wavelet packet transform based on approximate entropy and energy with artificial neural network

Abstract: This paper proposes the method that can detect both spikes and epileptic seizure at the same time based on wavelet packet transform, approximate entropy and energy, and artificial neural network. First, the EEG signals are decomposed into 4 levels, 16 frequency sub-bands, using Daubechies for mother wavelet to distinguish the usable signal. Then the approximate entropy and energy features are extracted for each sub-band to form the feature vector. Finally, the constructed feature vector is used as an input to the artificial neural network to classify the EEG signals into 6 types of spike, epileptic seizure, eye closed, eye opened, body movement, and normal signal. The experimental results show that the proposed method identified and classified the EEG signal with average sensitivity of 76.55%, specificity of 81.3%, and accuracy of 89.47%.

Breath monitoring based on the optical electronic nose system

Abstract: The identification of volatile organic compounds (VOCs) presented in exhaled human breath can provide direct evidence of the health status of patients. Electronic nose (e-nose) has been known as an efficient technique for odor classification of numerous products, living and non-living objects and environment. In this work, an optical e-nose system that consists of organic thin film gas sensor and optical detection system was investigated on its potential for biomedical applications. The gas sensor was prepared based on metal-porphyrin compounds that are the optically active to various basic VOC. In addition, low-cost light emitting diodes (LEDs) array and a CMOS photo-detector were integrated to analyze the color change in the sensing material. The sensing materials were tested with various common VOCs, for examples, alcohols, acetone, ammonia and water. Pattern recognition, namely principal component analysis (PCA), clearly distinguished the different concentration of ammonia vapor, and discriminated ammonia from acetone vapor. Moreover, optical e-nose successfully detected the change of exhaled breath such as before and after eating. Consequently, this system is expected to be further developed as a breath monitoring system.

A new content-based medical image retrieval system based on wavelet transform and multidimensional Wald-Wolfowitz runs test

Abstract: Recently, one of the authors proposed a new similarity measure, called weighted multidimensional Wald and Wolfowitz (MWW) runs test, for the content-based color image retrieval system. The algorithm outperforms conventional similarity measures for comparing two color images. In this paper, we propose a new content-based medical image retrieval system based on discrete wavelet transform (DWT) symlet and the weighted MWW runs test. The DWT is used to extracted texture features of the medical images. The weighted MWW runs test is used to compare distributions of texture features of two medical images. Our experiments were performed on 1,000 medical images from image retrieval in medical applications (IRMA). The experimental results show promisingly efficient to retrieve the medical images.

Estimation of longitudinal muscle motion using a primal-dual algorithm

Abstract: Ultrasonography is a convenient and widely-used technique to view the longitudinal muscle motion as it is non-invasive and real-time. However, the estimation of longitudinal muscle motion, in sense of dense deformation field, was paid little attention to previously. In this report a primal-dual algorithm was used to estimate the motion of gastrocnemius muscle (GM) in longitudinal direction. Then a color encoding scheme and the mean velocities (mm/s) of frame-wise motion field were adopted to show the synchronous activities. The preliminary results, on 600 frames of ultrasound muscle images in motion, demonstrated that the proposed estimation method are helpful to study of the contraction of muscle in longitudinal direction, providing motion details both spatially and temporally.

Improvements in ultrasound elastography using dynamic focusing

Abstract: Many ultrasound elastograms are reconstructed from pre- and post-compression radiofrequency (RF) signals after beamforming using a single or a few focusing depths. This might limit the accuracy of the displacement estimates for elastography in the area far from the focusing depths. This article studies the improvement of the elastograms reconstructed from the signals beamformed using a dynamic focusing scheme compared with a single focusing scheme. The results show that the elastograms reconstructed using the signals from the dynamic focusing scheme have higher signal-to-noise ratio (SNR e ) and contrast-to-noise ratio (CNR e ) than those from the single focusing scheme.