MyMedR

Displaying publications 101 - 120 of 421 in total

Abstract:

Sort:

Fulltext Automatic COVID-19 Detection Using Exemplar Hybrid Deep Features with X-ray Images

Barua PD, Muhammad Gowdh NF, Rahmat K, Ramli N, Ng WL, Chan WY, et al.

Int J Environ Res Public Health, 2021 07 29;18(15).
PMID: 34360343 DOI: 10.3390/ijerph18158052

COVID-19 and pneumonia detection using medical images is a topic of immense interest in medical and healthcare research. Various advanced medical imaging and machine learning techniques have been presented to detect these respiratory disorders accurately. In this work, we have proposed a novel COVID-19 detection system using an exemplar and hybrid fused deep feature generator with X-ray images. The proposed Exemplar COVID-19FclNet9 comprises three basic steps: exemplar deep feature generation, iterative feature selection and classification. The novelty of this work is the feature extraction using three pre-trained convolutional neural networks (CNNs) in the presented feature extraction phase. The common aspects of these pre-trained CNNs are that they have three fully connected layers, and these networks are AlexNet, VGG16 and VGG19. The fully connected layer of these networks is used to generate deep features using an exemplar structure, and a nine-feature generation method is obtained. The loss values of these feature extractors are computed, and the best three extractors are selected. The features of the top three fully connected features are merged. An iterative selector is used to select the most informative features. The chosen features are classified using a support vector machine (SVM) classifier. The proposed COVID-19FclNet9 applied nine deep feature extraction methods by using three deep networks together. The most appropriate deep feature generation model selection and iterative feature selection have been employed to utilise their advantages together. By using these techniques, the image classification ability of the used three deep networks has been improved. The presented model is developed using four X-ray image corpora (DB1, DB2, DB3 and DB4) with two, three and four classes. The proposed Exemplar COVID-19FclNet9 achieved a classification accuracy of 97.60%, 89.96%, 98.84% and 99.64% using the SVM classifier with 10-fold cross-validation for four datasets, respectively. Our developed Exemplar COVID-19FclNet9 model has achieved high classification accuracy for all four databases and may be deployed for clinical application.
Automatic colonic polyp detection using integration of modified deep residual convolutional neural network and ensemble learning approaches

Liew WS, Tang TB, Lin CH, Lu CK

Comput Methods Programs Biomed, 2021 Jul;206:106114.
PMID: 33984661 DOI: 10.1016/j.cmpb.2021.106114

BACKGROUND AND OBJECTIVE: The increased incidence of colorectal cancer (CRC) and its mortality rate have attracted interest in the use of artificial intelligence (AI) based computer-aided diagnosis (CAD) tools to detect polyps at an early stage. Although these CAD tools have thus far achieved a good accuracy level to detect polyps, they still have room to improve further (e.g. sensitivity). Therefore, a new CAD tool is developed in this study to detect colonic polyps accurately.
METHODS: In this paper, we propose a novel approach to distinguish colonic polyps by integrating several techniques, including a modified deep residual network, principal component analysis and AdaBoost ensemble learning. A powerful deep residual network architecture, ResNet-50, was investigated to reduce the computational time by altering its architecture. To keep the interference to a minimum, median filter, image thresholding, contrast enhancement, and normalisation techniques were exploited on the endoscopic images to train the classification model. Three publicly available datasets, i.e., Kvasir, ETIS-LaribPolypDB, and CVC-ClinicDB, were merged to train the model, which included images with and without polyps.
RESULTS: The proposed approach trained with a combination of three datasets achieved Matthews Correlation Coefficient (MCC) of 0.9819 with accuracy, sensitivity, precision, and specificity of 99.10%, 98.82%, 99.37%, and 99.38%, respectively.
CONCLUSIONS: These results show that our method could repeatedly classify endoscopic images automatically and could be used to effectively develop computer-aided diagnostic tools for early CRC detection.
Fulltext Deep learning architectures analysis for age-related macular degeneration segmentation on optical coherence tomography scans

Alsaih K, Yusoff MZ, Tang TB, Faye I, Mériaudeau F

Comput Methods Programs Biomed, 2020 Oct;195:105566.
PMID: 32504911 DOI: 10.1016/j.cmpb.2020.105566

BACKGROUND AND OBJECTIVES: Aged people usually are more to be diagnosed with retinal diseases in developed countries. Retinal capillaries leakage into the retina swells and causes an acute vision loss, which is called age-related macular degeneration (AMD). The disease can not be adequately diagnosed solely using fundus images as depth information is not available. The variations in retina volume assist in monitoring ophthalmological abnormalities. Therefore, high-fidelity AMD segmentation in optical coherence tomography (OCT) imaging modality has raised the attention of researchers as well as those of the medical doctors. Many methods across the years encompassing machine learning approaches and convolutional neural networks (CNN) strategies have been proposed for object detection and image segmentation.
METHODS: In this paper, we analyze four wide-spread deep learning models designed for the segmentation of three retinal fluids outputting dense predictions in the RETOUCH challenge data. We aim to demonstrate how a patch-based approach could push the performance for each method. Besides, we also evaluate the methods using the OPTIMA challenge dataset for generalizing network performance. The analysis is driven into two sections: the comparison between the four approaches and the significance of patching the images.
RESULTS: The performance of networks trained on the RETOUCH dataset is higher than human performance. The analysis further generalized the performance of the best network obtained by fine-tuning it and achieved a mean Dice similarity coefficient (DSC) of 0.85. Out of the three types of fluids, intraretinal fluid (IRF) is more recognized, and the highest DSC value of 0.922 is achieved using Spectralis dataset. Additionally, the highest average DSC score is 0.84, which is achieved by PaDeeplabv3+ model using Cirrus dataset.
CONCLUSIONS: The proposed method segments the three fluids in the retina with high DSC value. Fine-tuning the networks trained on the RETOUCH dataset makes the network perform better and faster than training from scratch. Enriching the networks with inputting a variety of shapes by extracting patches helped to segment the fluids better than using a full image.
Mitotic nuclei analysis in breast cancer histopathology images using deep ensemble classifier

Sohail A, Khan A, Nisar H, Tabassum S, Zameer A

Med Image Anal, 2021 08;72:102121.
PMID: 34139665 DOI: 10.1016/j.media.2021.102121

Mitotic nuclei estimation in breast tumour samples has a prognostic significance in analysing tumour aggressiveness and grading system. The automated assessment of mitotic nuclei is challenging because of their high similarity with non-mitotic nuclei and heteromorphic appearance. In this work, we have proposed a new Deep Convolutional Neural Network (CNN) based Heterogeneous Ensemble technique "DHE-Mit-Classifier" for analysis of mitotic nuclei in breast histopathology images. The proposed technique in the first step detects candidate mitotic patches from the histopathological biopsy regions, whereas, in the second step, these patches are classified into mitotic and non-mitotic nuclei using the proposed DHE-Mit-Classifier. For the development of a heterogeneous ensemble, five different deep CNNs are designed and used as base-classifiers. These deep CNNs have varying architectural designs to capture the structural, textural, and morphological properties of the mitotic nuclei. The developed base-classifiers exploit different ideas, including (i) region homogeneity and feature invariance, (ii) asymmetric split-transform-merge, (iii) dilated convolution based multi-scale transformation, (iv) spatial and channel attention, and (v) residual learning. Multi-layer-perceptron is used as a meta-classifier to develop a robust and accurate classifier for providing the final decision. The performance of the proposed ensemble "DHE-Mit-Classifier" is evaluated against state-of-the-art CNNs. The performance evaluation on the test set suggests the superiority of the proposed ensemble with an F-score (0.77), recall (0.71), precision (0.83), and area under the precision-recall curve (0.80). The good generalisation of the proposed ensemble with a considerably high F-score and precision suggests its potential use in the development of an assistance tool for pathologists.
Fulltext Neovascularization Detection and Localization in Fundus Images Using Deep Learning

Tang MCS, Teoh SS, Ibrahim H, Embong Z

Sensors (Basel), 2021 Aug 06;21(16).
PMID: 34450766 DOI: 10.3390/s21165327

Proliferative Diabetic Retinopathy (PDR) is a severe retinal disease that threatens diabetic patients. It is characterized by neovascularization in the retina and the optic disk. PDR clinical features contain highly intense retinal neovascularization and fibrous spreads, leading to visual distortion if not controlled. Different image processing techniques have been proposed to detect and diagnose neovascularization from fundus images. Recently, deep learning methods are getting popular in neovascularization detection due to artificial intelligence advancement in biomedical image processing. This paper presents a semantic segmentation convolutional neural network architecture for neovascularization detection. First, image pre-processing steps were applied to enhance the fundus images. Then, the images were divided into small patches, forming a training set, a validation set, and a testing set. A semantic segmentation convolutional neural network was designed and trained to detect the neovascularization regions on the images. Finally, the network was tested using the testing set for performance evaluation. The proposed model is entirely automated in detecting and localizing neovascularization lesions, which is not possible with previously published methods. Evaluation results showed that the model could achieve accuracy, sensitivity, specificity, precision, Jaccard similarity, and Dice similarity of 0.9948, 0.8772, 0.9976, 0.8696, 0.7643, and 0.8466, respectively. We demonstrated that this model could outperform other convolutional neural network models in neovascularization detection.
Fulltext Identification of Autism Subtypes Based on Wavelet Coherence of BOLD FMRI Signals Using Convolutional Neural Network

Al-Hiyali MI, Yahya N, Faye I, Hussein AF

Sensors (Basel), 2021 Aug 04;21(16).
PMID: 34450699 DOI: 10.3390/s21165256

The functional connectivity (FC) patterns of resting-state functional magnetic resonance imaging (rs-fMRI) play an essential role in the development of autism spectrum disorders (ASD) classification models. There are available methods in literature that have used FC patterns as inputs for binary classification models, but the results barely reach an accuracy of 80%. Additionally, the generalizability across multiple sites of the models has not been investigated. Due to the lack of ASD subtypes identification model, the multi-class classification is proposed in the present study. This study aims to develop automated identification of autism spectrum disorder (ASD) subtypes using convolutional neural networks (CNN) using dynamic FC as its inputs. The rs-fMRI dataset used in this study consists of 144 individuals from 8 independent sites, labeled based on three ASD subtypes, namely autistic disorder (ASD), Asperger's disorder (APD), and pervasive developmental disorder not otherwise specified (PDD-NOS). The blood-oxygen-level-dependent (BOLD) signals from 116 brain nodes of automated anatomical labeling (AAL) atlas are used, where the top-ranked node is determined based on one-way analysis of variance (ANOVA) of the power spectral density (PSD) values. Based on the statistical analysis of the PSD values of 3-level ASD and normal control (NC), putamen_R is obtained as the top-ranked node and used for the wavelet coherence computation. With good resolution in time and frequency domain, scalograms of wavelet coherence between the top-ranked node and the rest of the nodes are used as dynamic FC feature input to the convolutional neural networks (CNN). The dynamic FC patterns of wavelet coherence scalogram represent phase synchronization between the pairs of BOLD signals. Classification algorithms are developed using CNN and the wavelet coherence scalograms for binary and multi-class identification were trained and tested using cross-validation and leave-one-out techniques. Results of binary classification (ASD vs. NC) and multi-class classification (ASD vs. APD vs. PDD-NOS vs. NC) yielded, respectively, 89.8% accuracy and 82.1% macro-average accuracy, respectively. Findings from this study have illustrated the good potential of wavelet coherence technique in representing dynamic FC between brain nodes and open possibilities for its application in computer aided diagnosis of other neuropsychiatric disorders, such as depression or schizophrenia.
Fulltext Neural networks for forecasting daily reservoir inflows

Karimi-Googhari, Shahram, Huang, Yuk Feng, Abdul Halim B. Ghazali, Lee, Teang Shui

Pertanika Journal of Science & Technology, 2010;18(1):-.
MyJurnal

Proper integrated management of a dam reservoir requires that all components of the water resource system be known. One of these components is the daily reservoir inflow which is the subject matter of this study, i.e. to establish predictions of what is coming in the next rainfall-runoff process over a catchment. The transformation of rainfall into runoff is an extremely complex, dynamic, and more of a non-linear process. The available six-year average daily rainfall data across the Sembrong dam catchment were computed using the well-known Theissen’s polygon method. Daily reservoir inflow data were extracted by applying the water balance model to the Sembrong dam reservoir. Modelling of relationship between rainfall and reservoir inflow data was done using feed-forward back-propagation neural networks. The final selected model has one hidden layer with 11 neurons in the hidden layer. The selected model was applied for an independent data series testing. Results in relation to specific climatic and hydrologic properties of a small tropical catchment suggested that the model is suitable to be used in forecasting the next day’s reservoir inflow. The efficiencies of the model Abtained indicated the validity of using the neural network for modelling reservoir inflow series.
Fulltext Grid base classifier in comparison to nonparametric methods in multiclass classification

Mohebpour, M.R., Adznan, B.J., Saripan, M.I.

Pertanika Journal of Science & Technology, 2010;18(1):-.
MyJurnal

In this paper, a new method known as Grid Base Classifier was proposed. This method carries the advantages of the two previous methods in order to improve the classification tasks. The problem with the current lazy algorithms is that they learn quickly, but classify very slowly. On the other hand, the eager algorithms classify quickly, but they learn very slowly. The two algorithms were compared, and the proposed algorithm was found to be able to both learn and classify quickly. The method was developed based on the grid structure which was done to create a powerful method for classification. In the current research, the new algorithm was tested and applied to the multiclass classification of two or more categories, which are important for handling problems related to practical classification. The new method was also compared with the Levenberg-Marquardt back-propagation neural network in the learning stage and the Condensed nearest neighbour in the generalization stage to examine the performance of the model. The results from the artificial and real-world data sets (from UCI Repository) showed that the new method could improve both the efficiency and accuracy of pattern classification.
Fulltext Intelligent Monitoring interfaces for Coal fired Power Plant Boiler Trips: a review

Nistah, N. N. M., Samyudia, Y., Alnaimi, F. B. I., Motalebi, F.

Pertanika Journal of Science & Technology, 2014;22(2):593-601.
MyJurnal

A major source of contemporary power is a Coal-fired Power Plant. These power plants have the capacity to continuously supply electricity to almost 500,000 residential and business units. An essential component of a Coal-fired Power plant is automation. A feature of this automation is an Intelligent System developed for the Power Plant. These Intelligent Systems have different configurations and design. This research studies the various Intelligent Monitoring Interfaces developed for Coal-fired Power Plant Trips, their advantages, disadvantages and proposes a new Intelligent Monitoring Interface that would alleviate the disadvantages of the existing systems. Current systems that use Neural Network models are investigated. The improved Intelligent Monitoring Interface as proposed in this paper is a modification of the existing monitoring system for the Coal-fired Power Plant Boiler Trips. It is expected to improve the overall system by implementing remote accessibility and interactability between the plant operator and the control system interface. The interface will also assist the operator by providing guidelines to troubleshoot the identified trips and the remote server application will allow data collected to be viewed anytime, anywhere.
Statistical optimization of the phytoremediation of arsenic by Ludwigia octovalvis- in a pilot reed bed using response surface methodology (RSM) versus an artificial neural network (ANN)

Titah HS, Halmi MIEB, Abdullah SRS, Hasan HA, Idris M, Anuar N

Int J Phytoremediation, 2018 Jun 07;20(7):721-729.
PMID: 29723047 DOI: 10.1080/15226514.2017.1413337

In this study, the removal of arsenic (As) by plant, Ludwigia octovalvis, in a pilot reed bed was optimized. A Box-Behnken design was employed including a comparative analysis of both Response Surface Methodology (RSM) and an Artificial Neural Network (ANN) for the prediction of maximum arsenic removal. The predicted optimum condition using the desirability function of both models was 39 mg kg-1 for the arsenic concentration in soil, an elapsed time of 42 days (the sampling day) and an aeration rate of 0.22 L/min, with the predicted values of arsenic removal by RSM and ANN being 72.6% and 71.4%, respectively. The validation of the predicted optimum point showed an actual arsenic removal of 70.6%. This was achieved with the deviation between the validation value and the predicted values being within 3.49% (RSM) and 1.87% (ANN). The performance evaluation of the RSM and ANN models showed that ANN performs better than RSM with a higher R2 (0.97) close to 1.0 and very small Average Absolute Deviation (AAD) (0.02) and Root Mean Square Error (RMSE) (0.004) values close to zero. Both models were appropriate for the optimization of arsenic removal with ANN demonstrating significantly higher predictive and fitting ability than RSM.
Fulltext DyHAP: Dynamic Hybrid ANFIS-PSO Approach for Predicting Mobile Malware

Afifi F, Anuar NB, Shamshirband S, Choo KK

PLoS One, 2016;11(9):e0162627.
PMID: 27611312 DOI: 10.1371/journal.pone.0162627

To deal with the large number of malicious mobile applications (e.g. mobile malware), a number of malware detection systems have been proposed in the literature. In this paper, we propose a hybrid method to find the optimum parameters that can be used to facilitate mobile malware identification. We also present a multi agent system architecture comprising three system agents (i.e. sniffer, extraction and selection agent) to capture and manage the pcap file for data preparation phase. In our hybrid approach, we combine an adaptive neuro fuzzy inference system (ANFIS) and particle swarm optimization (PSO). Evaluations using data captured on a real-world Android device and the MalGenome dataset demonstrate the effectiveness of our approach, in comparison to two hybrid optimization methods which are differential evolution (ANFIS-DE) and ant colony optimization (ANFIS-ACO).
Fulltext Comparison between VAR, GSTAR, FFNN-VAR and FFNN-GSTAR Models for forecasting oil production

Suhartono, Prastyo, Dedy Dwi, Kuswanto, Heri, Muhammad Hisyam Lee

MATEMATIKA, 2018;34(1):103-111.
MyJurnal

Monthly data about oil production at several drilling wells is an example of
spatio-temporal data. The aim of this research is to propose nonlinear spatio-temporal
model, i.e. Feedforward Neural Network - VectorAutoregressive (FFNN-VAR) and FFNN
- Generalized Space-Time Autoregressive (FFNN-GSTAR), and compare their forecast
accuracy to linearspatio-temporal model, i.e. VAR and GSTAR. These spatio-temporal
models are proposed and applied for forecasting monthly oil production data at three
drilling wells in East Java, Indonesia. There are 60 observations that be divided to two
parts, i.e. the first 50 observations for training data and the last 10 observations for
testing data. The results show that FFNN-GSTAR(11) and FFNN-VAR(1) as nonlinear
spatio-temporal models tend to give more accurate forecast than VAR(1) and GSTAR(11)
as linear spatio-temporal models. Moreover, further research about nonlinear spatiotemporal
models based on neural networks and GSTAR is needed for developing new
hybrid models that could improve the forecast accuracy.
A Stackelberg solution to a two-level linear fractional programming problem with interval coefficients in the objective functions

Borza M, Rambely A, Saraj M

Sains Malaysiana, 2012;41:1651-1656.

In this paper, two approaches were introduced to obtain Stackelberg solutions for two-level linear fractional programming problems with interval coefficients in the objective functions. The approaches were based on the Kth best method and the method for solving linear fractional programming problems with interval coefficients in the objective function. In the first approach, linear fractional programming with interval coefficients in the objective function and linear programming were utilized to obtain Stackelberg solution, but in the second approach only linear programming is used. Since a linear fractional programming with interval coefficients can be equivalently transformed into a linear programming, therefore both of approaches have same results. Numerical examples demonstrate the feasibility and effectiveness of the methods.
Prediction of lead seven day minimum and maximum surface air temperature using neural network and genetic programming

K. Ramesh, P. Ramalakshmi, R. Anitha

Sains Malaysiana, 2015;44:1389-1396.

The determination of variance of surface air temperature is very essential since it has a direct impact on vegetation, environment and human livelihood. Forecast of surface air temperature is difficult because of the complex physical phenomenon and the random-like behavior of atmospheric system which influences the temperature event on the earth surface. In this study, forecast models based on artificial neural network (ANN) and genetic programming (GP) approaches were proposed to predict lead seven days minimum and maximum surface air temperature using the weather parameters observed at the station Chennai, India. The outcome of this study stated that models formulated using ANN approach are more accurate than genetic programming for all seven days with the highest coefficient of determination (R2), least mean absolute error (MAE), root mean square error (RMSE) and mean absolute percentage error (MAPE) on deployment with independent test dataset. ANN models give statistically acceptable mean absolute error of 0.59oC for lead day one in minimum temperature forecast and 0.86oC variance for lead day one in maximum temperature forecast. The study also clarified that the level of accuracy of the proposed prediction models were found to be better for smaller lead days when compared with higher lead days with both approaches.
An offset-free MPC formulation for nonlinear systems using adaptive integral controller

Hermansson AW, Syafiie S

ISA Trans, 2019 Aug;91:66-77.
PMID: 30782432 DOI: 10.1016/j.isatra.2019.01.037

This paper investigates a novel offset-free control scheme based on a multiple model predictive controller (MMPC) and an adaptive integral action controller for nonlinear processes. Firstly, the multiple model description captures the essence of the nonlinear process, while keeping the MPC optimization linear. Multiple models also enable the controller to deal with the uncertainty associated with changing setpoint. Then, a min-max approach is utilized to counter the effect of parametric uncertainty between the linear models and the nonlinear process. Finally, to deal with other uncertainties, such as input and output disturbances, an adaptive integral action controller is run in parallel to the MMPC. Thus creating a novel offset-free approach for nonlinear systems that is more easily tuned than observer-based MPC. Simulation results for a pH-controller, which acts as an example of a nonlinear process, are presented to demonstrate the usefulness of the technique compared to using an observer-based MPC.
Comparison of Fully Automated and Semi-Automated Methods for Species Identification

Kalafi EY, Anuar MK, Sakharkar MK, Dhillon SK

Folia Biol. (Praha), 2018;64(4):137-143.
PMID: 30724159

The process of manual species identification is a daunting task, so much so that the number of taxonomists is seen to be declining. In order to assist taxonomists, many methods and algorithms have been proposed to develop semi-automated and fully automated systems for species identification. While semi-automated tools would require manual intervention by a domain expert, fully automated tools are assumed to be not as reliable as manual or semiautomated identification tools. Hence, in this study we investigate the accuracy of fully automated and semi-automated models for species identification. We have built fully automated and semi-automated species classification models using the monogenean species image dataset. With respect to monogeneans' morphology, they are differentiated based on the morphological characteristics of haptoral bars, anchors, marginal hooks and reproductive organs (male and female copulatory organs). Landmarks (in the semi-automated model) and shape morphometric features (in the fully automated model) were extracted from four monogenean species images, which were then classified using k-nearest neighbour and artificial neural network. In semi-automated models, a classification accuracy of 96.67 % was obtained using the k-nearest neighbour and 97.5 % using the artificial neural network, whereas in fully automated models, a classification accuracy of 90 % was obtained using the k-nearest neighbour and 98.8 % using the artificial neural network. As for the crossvalidation, semi-automated models performed at 91.2 %, whereas fully automated models performed slightly higher at 93.75 %.
Discrete hopfield neural network in restricted maximum k-satisfiability logic programming

Mohd Shareduwan Mohd Kasihmuddin, Saratha Sathasivam, Mohd Asyraf Mansor

Sains Malaysiana, 2018;47:1327-1335.

Maximum k-Satisfiability (MAX-kSAT) consists of the most consistent interpretation that generate the maximum number
of satisfied clauses. MAX-kSAT is an important logic representation in logic programming since not all combinatorial
problem is satisfiable in nature. This paper presents Hopfield Neural Network based on MAX-kSAT logical rule. Learning
of Hopfield Neural Network will be integrated with Wan Abdullah method and Sathasivam relaxation method to obtain
the correct final state of the neurons. The computer simulation shows that MAX-kSAT can be embedded optimally in
Hopfield Neural Network.
Load forecasting using combination model of multiple linear regression with neural network for Malaysian city

Nur Arina Bazilah Kamisan, Muhammad Hisyam Lee, Suhartono Suhartono, Abdul Ghapor Hussin, Yong Zulina Zubairi

Sains Malaysiana, 2018;47:419-426.

Forecasting a multiple seasonal data is differ from a usual seasonal data since it contains more than one cycle in a
data. Multiple linear regression (MLR) models have been used widely in load forecasting because of its usefulness in the
forecast a linear relationship with other factors but MLR has a disadvantage of having difficulties in modelling a nonlinear
relationship between the variables and influencing factors. Neural network (NN) model, on the other hand, is a good
model for modelling a nonlinear data. Therefore, in this study, a combination of MLR and NN models has proposed this
combination to overcome the problem. This hybrid model is then compared with MLR and NN models to see the performance
of the hybrid model. RMSE is used as a performance indicator and a proposed graphical error plot is introduce to see the
error graphically. From the result obtained this model gives a better forecast compare to the other two models.
A tree-based multiclassification of breast tumor histopathology images through deep learning

Murtaza G, Abdul Wahab AW, Raza G, Shuib L

Comput Med Imaging Graph, 2021 04;89:101870.
PMID: 33545489 DOI: 10.1016/j.compmedimag.2021.101870

Worldwide, the burden of cancer is drastically increasing over the past few years. Among all types of cancers in women, breast cancer (BrC) is the main cause of unnatural deaths. For early diagnosis, histopathology (Hp) imaging is a gold standard for positive and detailed (at tissue level) diagnosis of breast tumor (BrT) compared to mammogram images. A large number of studies used BrT Hp images to solve binary or multiclassification problems using high computational resources. However, classification models' performance may be compromised due to the high correlation among various types of BrT in Hp images, which raises the misclassification rate. Thus, this paper aims to develop a tree-based BrT multiclassification model via deep learning (DL) to extract discriminative features to solve the multiclassification problem with better performance using less computational resources. The main contributions of this work are to create an ensemble, tree-based DL model that is pre-trained on the BreakHis dataset, and implementation of a misclassification reduction algorithm. The ensemble, tree-based DL model, extracts discriminative BrT features from Hp images. The target dataset (i.e., Bioimaging challenge 2015 breast histology) is small in size; thus, to avoid overfitting of the proposed model, pretraining is performed on the BreakHis dataset. Whereas, misclassification reduction algorithm is implemented to enhance the performance of the classification model. The experimental results show that the proposed model outperformed the existing state-of-the-art baseline studies. The achieved classification accuracy is ranging from 87.50 % to 100 % for four subtypes of BrT. Thus, the proposed model can assist doctors as the second opinion in any healthcare centre.
Edge AI-Based Automated Detection and Classification of Road Anomalies in VANET Using Deep Learning

Bibi R, Saeed Y, Zeb A, Ghazal TM, Rahman T, Said RA, et al.

Comput Intell Neurosci, 2021;2021:6262194.
PMID: 34630550 DOI: 10.1155/2021/6262194

Road surface defects are crucial problems for safe and smooth traffic flow. Due to climate changes, low quality of construction material, large flow of traffic, and heavy vehicles, road surface anomalies are increasing rapidly. Detection and repairing of these defects are necessary for the safety of drivers, passengers, and vehicles from mechanical faults. In this modern era, autonomous vehicles are an active research area that controls itself with the help of in-vehicle sensors without human commands, especially after the emergence of deep learning (DNN) techniques. A combination of sensors and DNN techniques can be useful for unmanned vehicles for the perception of their surroundings for the detection of tracks and obstacles for smooth traveling based on the deployment of artificial intelligence in vehicles. One of the biggest challenges for autonomous vehicles is to avoid the critical road defects that may lead to dangerous situations. To solve the accident issues and share emergency information, the Intelligent Transportation System (ITS) introduced the concept of vehicular network termed as vehicular ad hoc network (VANET) for achieving security and safety in a traffic flow. A novel mechanism is proposed for the automatic detection of road anomalies by autonomous vehicles and providing road information to upcoming vehicles based on Edge AI and VANET. Road images captured via camera and deployment of the trained model for road anomaly detection in a vehicle could help to reduce the accident rate and risk of hazards on poor road conditions. The techniques Residual Convolutional Neural Network (ResNet-18) and Visual Geometry Group (VGG-11) are applied for the automatic detection and classification of the road with anomalies such as a pothole, bump, crack, and plain roads without anomalies using the dataset from different online sources. The results show that the applied models performed well than other techniques used for road anomalies identification.

Filters

Please provide feedback to Administrator (afdal@afpm.org.my)

External Links