34 research outputs found
Retinal Vessel Segmentation to Support Foveal Avascular Zone Detection
Get PDFPurpose: This study aims to perform retinal vessel segmentation to support foveal avascular zone detection. Methodology: The proposed approach consists of a multi-stage image processing approach, including preprocessing, image quality enhancementt, and segmentation of retinal blood vessel using matched filter and length filter techniques.Findings: The proposed framework has achieved remarkable results with an average sensitivity, specificity, and accuracy of 77.99%, 86.43%, and 85.24%, respectively.Value: This achievement has the potential to significantly enhance the accuracy and efficiency of detecting and diagnosing medical conditions related to the retina, improving the quality of life for countless individuals
Adaptive Modelling of Multipath Communication Channels Based on the Least Mean Square Algorithm
Get PDFThe objective of this research is to design an adaptive model for multipaths communication channels. To evaluate the performance of the model, the authors used two indicators, namely the attainment of a minimum value of error signals in the form of interface and echo on the channel, and Filter Finite Impulse Response (FIR) has the same form of equation as the channel equation used. This adaptation process uses the Least Mean Square (LMS) algorithm. The LMS algorithm is known as a simple algorithm, which is fast computation time and the computation results are quite satisfying. In this study, the authors used Simulink software on Matlab. In general, this research is divided into three stages, namely modeling multipaths communication channels as an adaptive system, designing an adaptive system model on Simulink, and designing a simulation of 250 iterations. After obtaining the simulation results in the form of errors and weight of the FIR filter, it is observed that both of the results have met the requirements of the indicators of the success of this study. With 250 iterations used, satisfactory results were obtained in the form of the fulfillment of all indicators of the success of this study
Edge and curvilinear structures detection on medical images via unsupervised, adaptive and deep learning approaches
No full textComputer-aided-diagnosis (CAD) systems are very useful to help doctors in detecting various human diseases. To build a CAD system, several computer vision algorithms are required, particularly to handle object detection and segmentation tasks automatically. To develop object detection algorithms, edges and curvilinear structures detection tasks are typically required. However, performing these tasks manually is tedious, time-consuming and prone to human errors. In this thesis, we design computer algorithms for edge and curvilinear structures detection, particularly for the application of optical disc boundary and retinal vessel detection from fundus images. The algorithms are developed based on the mathematical function that can closely represent the edge and curvilinear structures behaviours. The algorithms can detect edge and curvilinear structures under an unsupervised framework and they also allow an implementation with a deep learning architecture. This provides meaningful insight for robust edge and curvilinear structures detection algorithms developments on other image modalities.Doctor of Philosoph
Detection of Optic Disc Centre Point in Retinal Image
Get PDFGlaucoma and diabetic retinopathy (DR) are the two most common retinal related diseases occurred in the world. Glaucoma can be diagnosed by measuring optic cup to disc ratio (CDR) defined as optic cup to optic disc vertical diameter ratio of retinal fundus image. A computer based optic disc is expected to assist the ophthalmologist to find their location which are necessary for glaucoma and DR diagnosis. However, many optic disc detection algorithms available now are commonly non-automatic and only work in healthy retinal image. Therefore, there is not information on how optic disc in retinal image of unhealthy patient can be extracted computationally. In this research work, the method for automated detection of optic disc on retinal colour fundus images has been developed to facilitate and assist ophthalmologists in the diagnosis of retinal related diseases. The results indicated that the proposed method can be implemented in computer aided diagnosis of glaucoma and diabetic retinopathy system development
Segmentation of the Electrocardiography Images as a Tool to Identify Heart Diseases
Get PDFThe heart is a very vital organ. Cardiac examination can be done periodically by using an electrocardiograph. So that the heart's condition can be known. One of the optimization in helping the detection of heart disease automatically by using computer assistance. Automatic detection can be done by image processing methods as input, especially ECG images that have been segmented. In this study, ECG image segmentation is carried out through several stages, such as grayscalling, contrast enhancement, and segmentation. The hope, the results of this study can be used as input for automatic detection of heart disease
Retinal Blood Vessel Segmentation as a Tool to Detect Diabetic Retinopathy
Get PDFThe retina is an important part of the eye for humans. Inbesides its main function as part of the sense of sight, in the worldmedically, the retina after an image can be used to detect a numberdiseases, such as diabetic retinopathy. To detect a number of diseases,Retinal digital images taken using a digital fundus camera are used.In detecting diabetic retinopathy, digital images are neededsegmented retina. Nevertheless, automatic segmentation of digital imagesthe retina is a complex work, given the presence of artifactsas well as noise on the retinal digital image, evenly illuminated, intensitylow, low contrast, and varying lengths of retinal blood vessels.In this research, a blood vessel segmentation software system has been designed through three stagesimage processing, namely (i) preprocessing, (ii) improving image quality, (iii) andsegmentation of retinal blood vessels. With three image processing stages, the performance value is obtained, i.e. 84.62
Performance Analysis of Lung Cancer Diagnosis Algorithms on X-Ray Images
Get PDFAmong several types of cancer, lung cancer is regarded as one of the most common and serious. In this respect, early diagnosis is required and beneficial to reduce mortalities caused by this type of cancer. Such diagnosis is typically performed by doctors through manual examinations on X-Ray images. However, manual examinations are labor extensive and time consuming. In this paper, we conduct a study to analyze the performance of some computer-based lung cancer diagnosis algorithms. The algorithms are built using different feature extraction (gray-level co-occurrence matrix, pixel intensity, histogram and combination of the three) and machine learning (Multi-layer Perceptron and K-Nearest Neighbor) techniques and the performance of each algorithm is compared and analyzed. The result of the study shows that the best performance of lung cancer classification is obtained by the computer algorithm that uses the combined features to characterize lung cancer and subsequently classifies the features using Multi-layer Perceptron
Detection of Cervical Cancer Based on Learning Vector Quantization and Wavelet Transform
Get PDFCervical cancer has became the common women dsease in the world. Mostly, cervical cancer has been already known lately, because it is very dificult to detect this in early stage. In this work, a computer based software using Learning Vector Quantization (LVQ) has been designed as the early cervical cancer detection aid tool. There are six methods before the detection is performed, namely preprocessing, contrast stretching, median filtering, morphology operation, image segmentation, and Wavelet Transform based feature extraction. In tihis work, 73 cervical cell images that consist of 50 normal images and 23 cancer images are used. 35 normal images and 14 cancer images are used to train the LVQ. Then, 23 normal images and 9 cancer images are used in the testing process. Our results show 88,89 % cancer image can be detected correctly (sensitivity), 100 % normal image can be detected corerctly (specificity), and 95,83 % for overall detection (accuracy)
Implementation of GLCM for Features Extraction and Selection of Batik Images
Get PDFBatik is a craft that has high artistic value and has been a part of Indonesian culture (especially Java) for a long time. Batik cloth in Indonesia has various types of batik textures, batik cloth colors, and batik fabric patterns that reflect the regional origins of the batik cloth. Regarding the image of batik, the texture feature is an important feature because the ornaments on the batik cloth can be seen as different texture compositions. Besides batik motifs, also influenced by the shape characteristics that become parts of each batik motif. This research will add insight and knowledge to understand batik patterns based on the characteristics of batik motifs, namely texture. There are five batik motifs used, namely inland solo batik, semarang coastal batik, sidhomukti batik, parangklithik batik, and sidhodrajat batik. Initially preprocessing is done by cropping and grayscalling. Of the five image motifs, a cropping process is carried out for each motif. The next step is feature extraction. The features of GLCM were selected in this study. From the features contained in the GLCM, in this study four features were chosen, namely contrast, energy, correlation, and homogeneity. The final step is the selection or selection of features. The result of the feature selection of the four features carried out feature extraction are energy and homogeneity
