Devvi Sarwinda

Plasmodium parasite detection on Red Blood Cell image for the diagnosis of malaria using double thresholding

Malaria which is caused by Plasmodium parasite is one of the diseases that can cause death in patients. Detection of plasmodium parasites on the Red Blood Cell (RBC) image can help diagnose malaria quickly and accurately, especially in the areas that lacked medical expertise. This research proposes a detection method of plasmodium parasite at RBC using double thresholding for improving accuracy of detection. Our method is able to achieved Predictive Positive Value (PPV) and Sensitivity about 92.85% and 85.52% respectively.

Feature selection using kernel PCA for Alzheimer's disease detection with 3D MR Images of brain

This paper investigates the application of the kernel PCA to select the features that produced by an extraction feature method, i.e. complete local binary pattern from three orthogonal planes. The proposed approach is used to detect Alzheimers disease using 3D Magnetic Resonance Images (MRI) of brain. In this study, the feature extraction method is done by using the different radius and the number of different neighbors. A support vector machine classifier is adapted to discriminant normal from Alzheimers, normal from mild cognitive impairment (MCI) and MCI from Alzheimers. The experimental results show our proposed method achieves an accuracy of 100% for classification of Alzheimers and normal. This accuracy result is also achieved by MCI and normal classification, whereas the accuracy of Alzheimers and MCI classification is only 84%.

Texture and Gene Expression Analysis of the MRI Brain in Detection of Alzheimer’s Disease

Abstract Alzheimer’s disease is a type of dementia that can cause problems with human memory, thinking and behavior. This disease causes cell death and nerve tissue damage in the brain. The brain damage can be detected using brain volume, whole brain form, and genetic testing. In this research, we propose texture analysis of the brain and genomic analysis to detect Alzheimer’s disease. 3D MRI images were chosen to analyze the texture of the brain, and microarray data were chosen to analyze gene expression. We classified Alzheimer’s disease into three types: Alzheimer’s, Mild Cognitive Impairment (MCI), and Normal. In this study, texture analysis was carried out by using the Advanced Local Binary Pattern (ALBP) and the Gray Level Co-occurrence Matrix (GLCM). We also propose the bi-clustering method to analyze microarray data. The experimental results from texture analysis show that ALBP had better performance than GLCM in classification of Alzheimer’s disease. The ALBP method achieved an average value of accuracy of between 75% - 100% for binary classification of the whole brain data. Furthermore, Biclustering method with microarray data shows good performance gene expression, where this information show influence Alzheimer’s disease with total of bi-cluster is 6.

Detection of Alzheimer's disease using advanced local binary pattern from hippocampus and whole brain of MR images

Alzheimers disease as one type of dementia can cause problems to human memory, thinking and behavior. The brain damage can be detected using brain volume and whole brain form. The correlation between brain shrinkage and reduction of brain volume can affect to deformation texture. In this research, the enhancement texture approach was proposed, called advanced local binary pattern (ALBP) method. ALBP is introduced as a 2D and 3D feature extraction descriptor. In the ALBP, sign and magnitude value were introduced as an enhancement to the previous LBP method. Due to a great number of features are produced by ALBP, the principal component analysis (PCA) and factor analysis are used as feature selection method. Furthermore, SVM classifier is applied for multiclass classification including Alzheimers, mild cognitive impairment, and normal condition of whole brain and hippocampus. The experimental results from two scenarios (ALBP sign magnitude (2D) and ALBP sign magnitude using three orthogonal planes (3D) methods) show better accuracy and performance compare to previous method. Our proposed method achieved the average value of accuracy between 80% - 100% for both the whole brain and hippocampus data. In addition, uniform rotation invariant ALBP sign magnitude using three orthogonal planes as a 3D descriptor also outperforms other approaches with an average accuracy of 96.28% for multiclass classifications for whole brain image.

Perceptron rule improvement on FIMT-DD for large traffic data stream

This paper proposed a method to build knowledge from one and a half years of UK traffic data sets. The method used is the Fast Incremental Model Trees - Drift Detection (FIMT-DD) with an improvement on the perceptron rule. In order to predict a traditional data set, we first analyze the model. After we have analyzed the model, we then average it from different arrangements of the datasets. In a stream data set, the approach is different from the traditional data sets. The approach of a stream data set is to take several snapshots during the induction to analyze the accuracy progress of the predicted model. We used the tanh activation function to optimize the perceptron rule. Therefore, we are able to reduce the error of the result. The error measurements that we analyzed in this paper are MAE, RMSE, and SMAPE. From 100,000,000 instances of traffic data, the optimized FIMT-DD algorithm method proves to be successful with smaller error value than standard FIMT-DD algorithm. To be able to measure the performance of the error of our optimized FIMT-DD algorithm, we used evaluate prediction sequential to analyze the data. According to the error measurement results of the MAE, RMSE, and SMAPE, tanh(x) activation function has a good influence to decrease the error value and improve the accuracy. The smaller error values did not only happen in a few experiments instances, but it occurs in the majority of the instances evaluation.

Image processing based detection of lung cancer on CT scan images

In this paper, we implement and analyze the image processing method for detection of lung cancer. Image processing techniques are widely used in several medical problems for picture enhancement in the detection phase to support the early medical treatment. In this research we proposed a detection method of lung cancer based on image segmentation. Image segmentation is one of intermediate level in image processing. Marker control watershed and region growing approach are used to segment of CT scan image. Detection phases are followed by image enhancement using Gabor filter, image segmentation, and features extraction. From the experimental results, we found the effectiveness of our approach. The results show that the best approach for main features detection is watershed with masking method which has high accuracy and robust.

A comparative study for error approximation of some kernel functions in Smooth Support Vector Machines

Support Vector Machine (SVM) as one of the most popular machine learning methods is playing a significant role in statistical learning theory. Smooth Support Vector Machine (SSVM) is one of new formulation to improve the SVM. In SSVM, smoothing method is used to optimize the unconstrained model. Smoothing function can be used to replace plus function in SVM. In this paper we evaluate eight smoothing functions including quadratic polynomial, fourth order polynomial, piecewise polynomial, spline function, sixth order polynomial, advanced fourth order polynomial function, quadratic Bezier function, third order Bezier function, and fourth order Bezier function. Some of those functions have been studied previously in order to find better accuracy in SVM. In this research, we evaluate and analyze the performance of all those eight smoothing functions based on their infinity-norm values. We compare the error approximation between smoothing function and plus function (as the standard kernel function in SVM) where the best smoothing function has the minimum error of its infinity-norm. Based on theoretical analysis and numerical approximation, our results show that piecewise polynomial function is better than quadratic polynomial function, fourth polynomial function, and third order spline function. While the advanced fourth order polynomial function and sixth order polynomial function have error approximation value almost the same as that of plus function. However, since the piecewise polynomial function has less control parameters than those of the advanced fourth order polynomial function and the sixth order polynomial function, we could not conclude which one is the best. Furthermore, based on values of infinity-norm of all the smoothing functions, we found that the quadratic Bezier and quadratic polynomial show the same error values. Meanwhile the forth order Bezier function shows the smallest error approximation value among the other functions which have been tested. In conclusion, based on our results in this study we found that the forth order Bezier function is the best choice among the eight smoothing function for SSVM.