Classification is the technique of identifying and assigning individual quantities to a group or a set. In pattern recognition, K-Nearest Neighbors algorithm is a non-parametric method for classification and regression. The K-Nearest Neighbor (kNN) technique has been widely used in data mining and machine learning because it is simple yet very useful with distinguished performance. Classification is used to predict the labels of test data points after training sample data. Over the past few decades, researchers have proposed many classification methods, but still, KNN (K-Nearest Neighbor) is one of the most popular methods to classify the data set. The input consists of k closest examples in each space, the neighbors are picked up from a set of objects or objects having same properties or value, this can be considered as a training dataset. In this paper, we have used two normalization techniques to classify the IRIS Dataset and measure the accuracy of classification using Cross-Validation method using R-Programming. The two approaches considered in this paper are Data with Z-Score Normalization and Data with Min-Max Normalization.

/pdf/comparative-analysis-of-knn-algorithm-using-various-3qz9d3jx6d.pdf

Comparative Analysis of KNN Algorithm using Various Normalization Techniques

A professional estimate on the computed tomography brain tumor images using SVM-SMO for classification and MRG-GWO for segmentation

Cell is the smallest unit of tissues, whose abnormal growth causes tumor in Brain. Support Vector Machine (SVM) and Artificial Neural Network (ANN) based tumor and its stages classification in brain MRI images is presented in this research work. This work is started with the enhancement of the brain MRI images which are obtained from oncology department of University of Maryland Medical Center. The integration of Temper based K-means and modified Fuzzy C-means (TKFCM) clustering algorithm used to segment the MRI images based on gray level intensity in small portion of brain image. The values of K in Temper based K-means algorithm more than the conventional K-means again, automatically updated membership of FCM eradicates the contouring problem of detecting the tumor region. Then, from the segmented images the first order statistic and region property based features are extracted. The first kind of features is used to detect and isolate tumor from normal brain MRI images with SVM. There is second kind which is used to classify the tumors into benign and four malignant stages tumor with ANN. The accuracy of classifying normal and tumor brain this proposed method is up to 97.37% with Bit Error Rate (BER) of 0.0294 which is better than other methods.

Classification of tumors and it stages in brain MRI using support vector machine and artificial neural network

Quantifying biofilm formation on surfaces is challenging because traditional microbiological methods, such as total colony-forming units (CFUs), often rely on manual counting. These are laborious, resource intensive techniques, more susceptible to human error. Confocal laser scanning microscopy (CLSM) is a high-resolution technique that allows 3D visualisation of biofilm architecture. In combination with a live/dead stain, it can be used to quantify biofilm viability on both transparent and opaque surfaces. However, there is little consensus on the appropriate methodology to apply in confocal micrograph processing. In this study, we report the development of an image analysis approach to repeatably quantify biofilm viability and surface coverage. We also demonstrate its use for a range of bacterial species and translational applications. This protocol has been created with ease of use and accessibility in mind, to enable researchers who do not specialise in computational techniques to be confident in applying these methods to analyse biofilm micrographs. Furthermore, the simplicity of the method enables the user to adapt it for their bespoke needs. Validation experiments demonstrate the automated analysis is robust and accurate across a range of bacterial species and an improvement on traditional microbiological analysis. Furthermore, application to translational case studies show the automated method is a reliable measurement of biomass and cell viability. This approach will ensure image analysis is an accessible option for those in the microbiology and biomaterials field, improve current detection approaches and ultimately support the development of novel strategies for preventing biofilm formation by ensuring comparability across studies.

/pdf/biofilm-viability-checker-an-open-source-tool-for-automated-1yddxv5v5h.pdf

Biofilm viability checker: An open-source tool for automated biofilm viability analysis from confocal microscopy images.

Optical polarimetry has previously imaged the spatial extent of a typical radiofrequency ablated (RFA) lesion in myocardial tissue, exhibiting significantly lower total depolarization at the necrotic core compared to healthy tissue, and intermediate values at the RFA rim region. Here, total depolarization in ablated myocardium was used to segment the total depolarization image into three (core, rim and healthy) zones. A local fuzzy thresholding algorithm was used for this multi-region segmentation, and then compared with a ground truth segmentation obtained from manual demarcation of RFA core and rim regions on the histopathology image. Quantitative comparison of the algorithm segmentation results was performed with evaluation metrics such as dice similarity coefficient (DSC = 0.78 ± 0.02 and 0.80 ± 0.02), sensitivity (Sn = 0.83 ± 0.10 and 0.91 ± 0.08), specificity (Sp = 0.76 ± 0.17 and 0.72 ± 0.17) and accuracy (Acc = 0.81 ± 0.09 and 0.71 ± 0.10) for RFA core and rim regions, respectively. This automatic segmentation of parametric depolarization images suggests a novel application of optical polarimetry, namely its use in objective RFA image quantification.

/pdf/polarization-image-segmentation-of-radiofrequency-ablated-g4wq85cdqb.pdf

Polarization image segmentation of radiofrequency ablated porcine myocardial tissue

This paper explores the possibility of applying techniques for segmenting the regions of medical image. For this we need to investigate the use of different techniques which helps for detection and classification of image regions. We also discuss some segmentation methods classified by researchers. Region classification is an essential process in the visualization of brain tissues of MRI. Brain image is basically classified into three regions; WM, GM and CSF. The forth region can be called as the tumor region, if the image is not normal. In the paper; Segmentation and characterization of Brain MR image regions using SOM and neuro fuzzy techniques, we integrate Self Organizing Map(SOM) and Neuro Fuzzy scheme to automatically extract WM, GM, CSF and tumor region of brain MRI image tested on three normal and three abnormal brain MRI images. Now in this paper this scheme is further tested on axial view images to classify the regions of brain MRI and compare the results from the Keith"s database. Using some statistical tests like accuracy, precision, sensitivity, specificity, positive predictive value, negative predictive value, false positive rate, false negative rate, likelihood ratio positive, likelihood ratio negative and prevalence of disease we calculate the effectiveness of the scheme.

/pdf/study-of-techniques-used-for-medical-image-segmentation-and-38nku2v62z.pdf

Anamika Ahirwar

Papers

Study of Techniques used for Medical Image Segmentation and Computation of Statistical Test for Region Classification of Brain MRI