Search

A leakage-free and disposable biochip for deoxyribonucleic acid (DNA) separation and detection is developed in this study. The biochip comprises of 50 mm long polydimethylsiloxane (PDMS) microchannel and copper electrodes engraved on flame retardant (FR-4) printed circuit board (PCB) substrate. An inhibitor made from photocurable diacrylate bisphenol-A polymer (DABA) was used to establish a permanent bonding between PDMS and PCB substrates. Pull-off test experiments resulted in average strength of 287.4 kPa and a standard deviation of ± 23.8 kPa. These results are comparable to recent studies on leak-free microfluidic applications. Meanwhile, the leakage test showed that the biochip could withstand a pressure of more than 190 kPa, which is sufficiently high for most DNA measurements. Finally, experiments performed on two different DNA samples indicated that the proposed design could accurately detect DNA fragments with current sensitivity higher than 100 nA, under electric field strength of 20 V/cm. The new design effectively seals the biochip, thus preventing leakage of liquid from the sensor matrix. Together with its biologically and electrochemically inert characteristic, it opens up possibilities in building a truly portable biosensing device.

A sudden blow or jolt to the human brain called traumatic brain injury (TBI) is one of the most common injuries recorded in the health insurance claim. Generally, computed tomography (CT) or magnetic resonance imaging (MRI) is required to identify the trauma’s severity. Unfortunately, CT and MRI equipment are bulky, expensive, and not always available, limiting their use in TBI detection. Therefore, as an alternative, this study presents a novel classification architecture that can classify non-severe TBI patients from healthy subjects by using resting-state electroencephalogram (EEG) as the input. The proposed architecture employs a convolutional neural network (CNN), and error-correcting output codes support vector machine (ECOC-SVM) to perform automated feature extraction and multi-class classification. In this architecture, complex feature selection and extraction steps are avoided. The proposed architecture attained a high-performance classification accuracy of 99.76%, potentially being used as a classification approach to preventing healthcare insurance fraud. The proposed method is compared to existing studies in the literature. The outcome from the comparisons indicates that the proposed method has outperformed the benchmarked methods by presenting the highest classification accuracy and precision.

Working memory (WM) encompasses crucial cognitive processes or abilities to retain and manipulate temporary information for immediate execution of complex cognitive tasks in daily functioning such as reasoning and decision-making. The WM of individuals sustaining traumatic brain injury (TBI) was commonly compromised, especially in the domain of WM. The current study investigated the brain responses of WM in a group of participants with mild-moderate TBI compared to their healthy counterparts employing functional magnetic resonance imaging. All consented participants (healthy: n = 26 and TBI: n = 15) performed two variations of the n-back WM task with four load conditions (0-, 1-, 2-, and 3-back). The respective within-group effects showed a right hemisphere-dominance activation and slower reaction in performance for the TBI group. Random-effects analysis revealed activation difference between the two groups in the right occipital lobe in the guided n-back with cues, and in the bilateral occipital lobe, superior parietal region, and cingulate cortices in the n-back without cues. The left middle frontal gyrus was implicated in the load-dependent processing of WM in both groups. Further group analysis identified that the notable activation changes in the frontal gyri and anterior cingulate cortex are according to low and high loads. Though relatively smaller in scale, this study was eminent as it clarified the neural alterations in WM in the mild-moderate TBI group compared to healthy controls. It confirmed the robustness of the phenomenon in TBI with the reproducibility of the results in a heterogeneous non-Western sample.

Background: This paper presents an improved radar-based imaging system for breast cancer detection that features p-slot ultrawideband antennae in a 32-array set-up. The improved reconstruction algorithm incorporates the phase coherence factor (PCF) into the conventional delay and sum (DAS) beamforming algorithm, thus effectively suppressing noise arising from the side- and gratinglobe interferences. Methods: The system is tested by using several breast models fabricated from chemical mixtures formulated on the basis of realistic human tissues. Each model is placed in a hemispherical breast radome that was fabricated from polylactide material and surrounded by 32 p-slot antennae mounted in four concentric layers. These antennae are connected to an 8.5 GHz vector network analyser through two 16-channel multiplexers that automatically switch different combinations of transmitter and receiver pairs in a sequential manner. Results: The system can accurately detect 5 mm tumours in a complex and homogeneously dense 3D breast model with an average signal-to-clutter ratio and full-width half-maximum of 7.0 dB and 2.3 mm, respectively. These values are more competitive than the values of other beamforming algorithms, even with contrasts as low as 1:2. Conclusion: The proposed PCF-weighted DAS is the best-performing algorithm amongst the tested beamforming techniques. This research paves the way for a clinical trial involving human subjects. Our laboratory is planning such a trial as part of future work.

One major problem encountered during the manufacture of crystalline silicon solar cells is microcrack. State-of-the-art system that uses the photoluminescence technique for microcrack detection suffers from one major drawback—this method is sensitive to microcrack as well as noise. Solving this problem leads to a design of an imaging system based on light transflection principle. In this case, an incident light enters a solar cell and is diffusely scattered in multiple directions due to the crystalline material and surface roughness of a solar cell. As a result, an area in the vicinity of the incident light is illuminated from beneath the surface. The presence of microcrack along the transmission path of the transflected light causes an abrupt change in the illumination intensity, thereby creating a shadow representation of the defect. Compared with photoluminescence imaging, the proposed system produces much cleaner images since it is less susceptible to artifacts in solar cell such as grain boundaries, surface structure, dislocations, cosmetic scratches, etc. This simplifies the detection process as complicated algorithms are avoided. Hence, such a design is ideally suited for an automated production environment and rapid in-line inspection process.

Encoding Image and video can be performed by different techniques. One of the popular technique is known as fractal-image-compression (FIC) method. Although it has several advantages like simplicity, resolution independent, and fast decoding, it has one main disadvantage, i.e. the compression time. This attributes to the requirement of conducting intensive similarity searches in order to map the image’s blocks together. The way of performing the search is a key point in the encoding speed. Therefore, this paper introduces a dynamic full-search approach for FIC method. The dynamic search is based on initiating the search from the nearest domain block to the range block required to be encoded, and the search window is expanded till an desirable matched domain block is detected or the search window covers the entire image. Although the proposed dynamic-search method looks simple, it is more efficient than the standard full search method which generally initiates the search, for each range block, from a static position (i.e. generally the top-left corner of the image and down to the bottom-right corner) regardless of its location. As a compromise between the image quality and the encoding time, a matching threshold equal to 500 has been used. For evaluation, the proposed method is compared with the traditional full-search method in terms of number of searches performed, and the quality of the decompressed image. For a 265 $\times$ 256 grayscale Lena image, the presented method is able to significantly decrease the search amount by 23 % compared to the traditional method, and also improve the image quality by 0.5 dB.

Both the multi-step and singe -step solutions to UWB inverse scattering problem are first discussed in this chapter. In the former, the CDF-Broyden update using the refined Polak-Rebiere's non -quadratic approximation will be focused. Meanwhile, the enhanced delay and sum (EDAS) algorithm, which uses the coherence factor and multiply in pair will be the principal topic of the later technique. In both cases the dispersive properties of the media are modelled using Debye first-order system, while the conformal mapping is deployed to accurately map curved geometries. The second part of the chapter focuses on an UWB radar data acquisition system, which is specifically developed to assess the performance of these algorithms experimentally. The hardware consists of custom-designed wideband antennas mounted on arrays and connected to a vector network analyser (VNA) via a 16-channel multiplexer switch unit. This set-up allows measurements be performed in the time domain automatically. Both limited-view and full-view geometries are used for measurements.

Traumatic brain injury (TBI) is one of the injuries that can bring serious consequences if medical attention has been delayed. Commonly, analysis of computed tomography (CT) or magnetic resonance imaging (MRI) is required to determine the severity of a moderate TBI patient. However, due to the rising number of TBI patients these days, employing the CT scan or MRI scan to every potential patient is not only expensive, but also time consuming. Therefore, in this paper, we investigate the possibility of using electroencephalography (EEG) with computational intelligence as an alternative approach to detect the severity of moderate TBI patients. EEG procedure is much cheaper than CT or MRI. Although EEG does not have high spatial resolutions as compared with CT and MRI, it has high temporal resolutions. The analysis and prediction of moderate TBI from EEG using conventional computational intelligence approaches are tedious as they normally involve complex preprocessing, feature extraction, or feature selection of the signal. Thus, we propose an approach that uses convolutional neural network (CNN) to automatically classify healthy subjects and moderate TBI patients. The input to this computational intelligence system is the resting-state eye-closed EEG, without undergoing preprocessing and feature selection. The EEG dataset used includes 15 healthy volunteers and 15 moderate TBI patients, which is acquired at the Hospital Universiti Sains Malaysia, Kelantan, Malaysia. The performance of the proposed method has been compared with four other existing methods. With the average classification accuracy of 72.46%, the proposed method outperforms the other four methods. This result indicates that the proposed method has the potential to be used as a preliminary screening for moderate TBI, for selection of the patients for further diagnosis and treatment planning.

Two common defects encountered during manufacturing of crystalline silicon solar cells are microcrack and dark spot or dark region. The microcrack in particular is a major threat to module performance since it is responsible for most PV failures and other types of damage in the field. On the other hand, dark region in which one cell or part of the cell appears darker under UV illumination is mainly responsible for PV reduced efficiency, and eventually lost of performance. Therefore, one key challenge for solar cell manufacturers is to remove defective cells from further processing. Recently, few researchers have investigated deep learning as an alternative approach for defect detection in solar cell manufacturing. The results are quite encouraging. This paper evaluates the convolutional neural network based on heavy-weighted You Only Look Once (YOLO) version 4 or YOLOv4 and the tiny version of this algorithm referred here as Tiny-YOLOv4. Experimental results suggest that the multi-class YOLOv4 is the best model in term of mean average precision (mAP) and prediction time, averaging at 98.8% and 62.9 ms respectively. Meanwhile an improved Tiny-YOLOv4 with Spatial Pyramid Pooling scheme resulted in mAP of 91.0% and runtime of 28.2 ms. Even though the tiny-weighted YOLOv4 performs slightly lower compared to its heavy-weighted counterpart, however the runtime of the former is 2.2 order much faster than the later.

Extraction of an optimal region of interest (ROI) is crucial in many image processing applications, such as estimation of the point spread function (PSF) and blind deconvolution (BD). Although the amount of publications on PSF and BD is quite extensive; however, the work on ROI estimation has not received much attention. Existing methods which used heuristic models are not only time-consuming but also computationally expensive. In this paper, we proposed a new ROI retrieval scheme based on image partitioning and entropy measurement feedback. This method has low computation cost since it contains no matrix operations. Comprehensive experiments on real and synthetic datasets revealed that the proposed method is competitive when compared with existing search techniques, averaging at 26.1 dB, 0.46 and 1.44 on peak signal-to-noise ratio, universal image quality index and error ratio scales, respectively. On average, the proposed method takes less than 10 s to retrieve the ROI which is significantly fas...

Traumatic brain injury (TBI) needs to be identified faster, so that suitable treatment can be planned properly. Normally, the severity of TBI is evaluated through the study from computed tomography (CT) or magnetic resonance imaging (MRI). Unfortunately, the number of CT scanners and MRI scanners is limited. Therefore, it is impractical to directly do CT or MRI scan to all patients without screening. Thus, this research investigates a method for screening moderate TBI patient. Data from resting state 63-channels electroencephalography is used in this work. Power of the signal is extracted from alpha, beta, theta and gamma frequency bands. This work utilizes a support vector machine, which is one of machine learning approaches, to identify moderate TBI patients. From the experimental results, it is shown that the average power from alpha or theta band gives the best accuracy score, which is at 70.83%.

Traumatic brain injury (TBI) is one of the traumas that requires immediate identification and treatment. Delayed in receiving medical attention can bring severe consequences. Conventional way of identifying TBI are from the computed tomography (CT) or magnetic resonance imaging (MRI). However, CT scan or MRI is hard to be performed to all patients as there were limited resources in the hospitals. As an early screening for TBI, EEG is a suitable substitution. Yet, it is time consuming for doctors or expert to perform EEG analysis manually for each patient. Therefore, this study suggests a machine learning method to identify moderate TBI patients. In this study, two type of features were analyzed. First is the raw EEG signal of alpha, beta, theta and gamma frequency bands. The second type of features are the power spectral density (PSD) of the respective frequency bands. The extracted features were used to train a support vector machine (SVM) separately to investigate the most informative frequency band features for identification of moderate TBI. From the result, it is found that alpha band features give the best identification accuracy which is at 87.5% for the raw features and 83.33 % for the PSD.

Convolution neural network (CNN) presents high robustness in computer vision applications. In state-of-the-art methods, CNN is being used in EEG processing for various classification and problem solving. To enable EEG to fit in the CNN architecture, data conversion of EEG has to be done. The ways of data conversion need to be investigated in order to fully utilize the information. From the study, it was found that the simplest way of re-arranging the signal is by creating a two dimensional matrix of channels versus time points. There are approaches that compute Pearson correlation coefficients and fit them into a two dimensional matrix to represent the input signal. There are also methods which extract frequency components and fit them in matrix structure as channels versus frequency components, such as power spectral density. Other approaches includes graph representation and wavelet components.

This paper reports the technical design of millimeter wave (mm-Wave) low noise amplifier (LNA). The fast-growing wireless communication market and the necessity of a higher data rate certainly drive the radio frequency (RF) system towards the mm-Wave technology. The 28 GHz channel band possesses a large amount of raw bandwidth space with a spectrum allocation that exceeds the bandwidth of 1 GHz. The mm-Wave cellular wireless system is developed to address with a worldwide wireless bandwidth shortage. The 28 GHz LNA is used to amplify a very low signal at 28 GHz and minimize the present of unwanted additional noises. The performances of single and cascaded LNAs are analyzed and presented. The footprint design of LNA is fabricated on ROGERS 4003 C with permittivity of 3.38. The off-the-shelf commercial MMIC LNA is mounted and matched to 50 Ω. The fabricated LNA is analyzed based on the output power levels and signal gains. The minimum and maximum gains for a single LNA are recorded at 11 dB and 14.83 dB respectively. 22.05 dB is verified as the lowest gain for the cascaded LNA while the highest gain is indicated at 28.83 dB when the RF signal input power is quoted at −74.54 dBm.

Lab-on-chip (LOC) is recognised as one of the most affordable solutions for integrating electronics and fluidics devices. In this field, bonding plays a vital role because it provides the means for attaching multiple components onto a substrate, transforming them into a microfluidic circuit. Bonding is an integral step, especially when designing a device that is free from leakage and eventual clogging. A comprehensive review of the latest irreversible bonding technologies is discussed in this paper, in which the focus is on the layered microfluidic systems with large sensor arrays. This review covers microfluidic devices fabricated from a rigid-type glass–fibre-printed circuit board and a thermoplastic flexible printed circuit with 186 references whose development date back three decades ago. The bonding techniques are organised into the following four groups: (a) adhesive bonding, (b) thermal and solvent bonding, (c) surface modification and dry bonding and (d) photoresist groups. Other techniques are available beyond these groupings, but they can be classified into the nearest group to facilitate the discussion. This paper will benefit researchers and practitioners aiming to develop polymer-based LOC devices.

This paper presents a high-speed architecture for fractal image compression.Bit-size reduction approach is used for speedup purpose.Real-time implementation of FIC based on full-search has been achieved.Parallel and pipeline implementation have been utilized to increase encoding speed. Fractal image compression (FIC) is a very popular coding technique use in image/video applications due to its simplicity and superior performance. The major drawback with FIC is that it is very time consuming algorithm, especially when a full search is attempted. Hence, it is very challenging to achieve a real-time operation if this algorithm is implemented on general processors. In this paper, a new parallel architecture with bit-width reduction scheme is implemented. The hardware is synthesized on Altera Cyclone II FPGA whose architecture is optimized at circuit level in order to achieve a real-time operation. The performance of the proposed architecture is evaluated in terms of runtime, peak-signal-to-noise-ratio (PSNR) and compression efficiency. On average the speedup of 3 was attainable through a bit-width reduction while the PSNR was maintained at acceptable level. Empirical results demonstrate that this firmware is competitive when compared to other existing hardware with PSNR averaging at 29.9źdB, 5.82% compression efficiency and a runtime equivalent to video speed of 101 frames per second (fps).

Biometric is an important field that enables identification of an individual to access their sensitive information and asset. In recent years, electroencephalography- (EEG-) based biometrics have been popularly explored by researchers because EEG is able to distinct between two individuals. The literature reviews have shown that convolutional neural network (CNN) is one of the classification approaches that can avoid the complex stages of preprocessing, feature extraction, and feature selection. Therefore, CNN is suggested to be one of the efficient classifiers for biometric identification. Conventionally, input to CNN can be in image or matrix form. The objective of this paper is to explore the arrangement of EEG for CNN input to investigate the most suitable input arrangement of EEG towards the performance of EEG-based identification. EEG datasets that are used in this paper are resting state eyes open (REO) and resting state eyes close (REC) EEG. Six types of data arrangement are compared in this paper. They are matrix of amplitude versus time, matrix of energy versus time, matrix of amplitude versus time for rearranged channels, image of amplitude versus time, image of energy versus time, and image of amplitude versus time for rearranged channels. It was found that the matrix of amplitude versus time for each rearranged channels using the combination of REC and REO performed the best for biometric identification, achieving validation accuracy and test accuracy of 83.21% and 79.08%, respectively.

Electroencephalogram (EEG) is used to study the activity of human brain using instrument named electroen-cephalograph. The application of EEG is now widened to many fields due to its great temporal resolution and other advantages. To utilize the advantage of EEG, the correct setup need to be done for recording the EEG. This is to ensure the EEG recorded is informative enough to serve the targeting application. In this paper, a literature study has been done to explore parameters of EEG recording used for different application. The parameters are number of channels, sampling rate and recording duration. It was found that for medical applications, the number of channel used ranging from two to 100 channels, sampling rate of 100Hz to 3000Hz, and recording duration from 3000 seconds to 21600 seconds. For BCI and neuromarketing, the number of channels used is from one up to 256 channels, 100Hz to 1000Hz sampling rate, and recording duration from five seconds onwards and some researches used continuous recording.

By using instrument known as electroencephalograph, electroencephalogram (EEG) of the human brain can be obtained. EEG is a record of postsynaptic potentials, generated by neurons. Usually, EEG is used to study the activities inside the human brain. This study gives significant assistant in medical field, especially in diagnosing and planning treatment for the brain related diseases. With the advancement of technology, the usage of EEG is now not only limited within the medical field. EEG has been used for brain-machine-interface (BCT) and neuromarketing. Therefore, the aim of this literature survey is to see the current trend of the applications of EEG. This survey is done by observing the research articles published in two well-known databases, which are IEEExplore and ScienceDirect. From this literature survey, it is found that the researches on EEG are still growing, with the area of applications is expanding.

A new approach incorporating adaptive lighting intensity for micro-crack inspection of solar wafers with variable thickness is proposed. Wafer thickness is measured with a pair of laser displacement sensors and the lighting intensity is adaptively adjusted to normalize near infrared (NIR) transmission based on measured thickness. This technique enables the image contrast be maintained at relatively uniform intensity in response to the variation of the solar wafer thickness. An improved version of Niblack segmentation algorithm is developed for this application. Experimental results demonstrate the competitiveness of the proposed system compared with established techniques, and achieves better performance both visually and quantitatively. Meanwhile, the runtime is consistently less than 1 s, corresponding to a throughput rate of approximately 3600 wafers/h. These results suggest that the methods and procedures are suitable for online processing of solar wafers.

Automatic grading of Harumanis mangoes by its shape and weight analysis.The Fourier descriptor method was developed to grade the shape of Harumanis mango.The cylinder method was developed to grade the weight of Harumanis mango.Multi-classification using Support Vector Machine and Discriminant analysis. This paper presents the work on the use of visible imaging as a tool in grading the mangoes. A Fourier-descriptor method was applied on mango images acquired by a CCD camera, to grade the fruits by their shapes. The method was able to correctly classify 98.3% using DA and 100% using SVM. It is also possible to estimate the weight of the mangoes from their images by applying the Cylinder approximation analysis method. The scatter plot between the estimated and actual values of the weight shows high correlation, with R2 equal to 94.0%. The high prediction accuracy obtained shows that this simple formula is adequate for the prediction of fruit weight and volume (measured volume using the cylinder method). The correlation formula derived based on the collected data is determined as w=2.256V-157.7 where w is estimated weight in grams and V is estimated volume. Overall result for weight grading using our proposed method yields 95% accuracy.