scholarly journals Optimization of Network Topology in Computer-Aided Detection Schemes Using Phased Searching with NEAT in a Time-Scaled Framework

2014 ◽  
Vol 13s1 ◽  
pp. CIN.S13885
Author(s):  
Maxine Tan ◽  
Jiantao Pu ◽  
Bin Zheng
Keyword(s):  
Selection Process ◽  
Learning Task ◽  
Classification Performance ◽  
Optimal Topology ◽  
Support Vector ◽  
Mass Detection ◽  
Detection Scheme ◽  
Linear Discriminant ◽  
Computer Aided ◽  
Fixed Topology

In the field of computer-aided mammographic mass detection, many different features and classifiers have been tested. Frequently, the relevant features and optimal topology for the artificial neural network (ANN)-based approaches at the classification stage are unknown, and thus determined by trial-and-error experiments. In this study, we analyzed a classifier that evolves ANNs using genetic algorithms (GAs), which combines feature selection with the learning task. The classifier named “Phased Searching with NEAT in a Time-Scaled Framework” was analyzed using a dataset with 800 malignant and 800 normal tissue regions in a 10-fold cross-validation framework. The classification performance measured by the area under a receiver operating characteristic (ROC) curve was 0.856 ± 0.029. The result was also compared with four other well-established classifiers that include fixed-topology ANNs, support vector machines (SVMs), linear discriminant analysis (LDA), and bagged decision trees. The results show that Phased Searching outperformed the LDA and bagged decision tree classifiers, and was only significantly outperformed by SVM. Furthermore, the Phased Searching method required fewer features and discarded superfluous structure or topology, thus incurring a lower feature computational and training and validation time requirement. Analyses performed on the network complexities evolved by Phased Searching indicate that it can evolve optimal network topologies based on its complexification and simplification parameter selection process. From the results, the study also concluded that the three classifiers – SVM, fixed-topology ANN, and Phased Searching with NeuroEvolution of Augmenting Topologies (NEAT) in a Time-Scaled Framework – are performing comparably well in our mammographic mass detection scheme.

scholarly journals A New Approach to Fall Detection Based on Improved Dual Parallel Channels Convolutional Neural Network

Sensors ◽  
2019 ◽  
Vol 19 (12) ◽  
pp. 2814 ◽  
Author(s):  
Xiaoguang Liu ◽  
Huanliang Li ◽  
Cunguang Lou ◽  
Tie Liang ◽  
Xiuling Liu ◽  
...  
Keyword(s):  
Neural Network ◽  
Convolutional Neural Network ◽  
Nearest Neighbor ◽  
Fall Detection ◽  
Classification Performance ◽  
Daily Activities ◽  
Support Vector ◽  
K Nearest Neighbor ◽  
Linear Discriminant ◽  
Parallel Channels

Falls are the major cause of fatal and non-fatal injury among people aged more than 65 years. Due to the grave consequences of the occurrence of falls, it is necessary to conduct thorough research on falls. This paper presents a method for the study of fall detection using surface electromyography (sEMG) based on an improved dual parallel channels convolutional neural network (IDPC-CNN). The proposed IDPC-CNN model is designed to identify falls from daily activities using the spectral features of sEMG. Firstly, the classification accuracy of time domain features and spectrograms are compared using linear discriminant analysis (LDA), k-nearest neighbor (KNN) and support vector machine (SVM). Results show that spectrograms provide a richer way to extract pattern information and better classification performance. Therefore, the spectrogram features of sEMG are selected as the input of IDPC-CNN to distinguish between daily activities and falls. Finally, The IDPC-CNN is compared with SVM and three different structure CNNs under the same conditions. Experimental results show that the proposed IDPC-CNN achieves 92.55% accuracy, 95.71% sensitivity and 91.7% specificity. Overall, The IDPC-CNN is more effective than the comparison in accuracy, efficiency, training and generalization.

Hardware Design of Seizure Detection Based on Wavelet Transform and Sample Entropy

2016 ◽  
Vol 25 (09) ◽  
pp. 1650101 ◽  
Author(s):  
Yuanfa Wang ◽  
Zunchao Li ◽  
Lichen Feng ◽  
Chuang Wang ◽  
Wen Jing ◽  
...  
Keyword(s):  
Wavelet Transform ◽  
Lattice Structure ◽  
Wavelet Coefficient ◽  
Distance Matrix ◽  
Seizure Detection ◽  
Sample Entropy ◽  
Classification Performance ◽  
Support Vector ◽  
Discrete Wavelet ◽  
Detection Scheme

Detecting epileptic seizure is a very time consuming and costly task if a support vector machine (SVM) hardware processor is used. In this paper, an automated seizure detection scheme is developed by combining discrete wavelet transform (DWT), sample entropy (SampEn) and a novel classification algorithm based on each wavelet coefficient and voting strategy. In order to save circuit area, a Daubechies order 4 (db4) filter of lattice structure is introduced in DWT, only half elements of the symmetric distance matrix in the SampEn are stored and module reusing strategy is used. To speed up the detection, intermediate results are reused by reasonably organizing the SampEn calculation procedures. The seizure detection scheme is implemented in a field-programmable gate array (FPGA) and its classification performance is tested with publicly available epilepsy dataset.

scholarly journals The Impact of Pixel Resolution, Integration Scale, Preprocessing, and Feature Normalization on Texture Analysis for Mass Classification in Mammograms

2016 ◽  
Vol 2016 ◽  
pp. 1-12 ◽  
Author(s):  
Mohamed Abdel-Nasser ◽  
Jaime Melendez ◽  
Antonio Moreno ◽  
Domenec Puig
Keyword(s):  
Texture Analysis ◽  
Spatial Arrangement ◽  
Classification Performance ◽  
Support Vector ◽  
Forward Selection ◽  
Mass Detection ◽  
Pixel Resolution ◽  
Feature Normalization ◽  
Mass Classification ◽  
The Impact

Texture analysis methods are widely used to characterize breast masses in mammograms. Texture gives information about the spatial arrangement of the intensities in the region of interest. This information has been used in mammogram analysis applications such as mass detection, mass classification, and breast density estimation. In this paper, we study the effect of factors such as pixel resolution, integration scale, preprocessing, and feature normalization on the performance of those texture methods for mass classification. The classification performance was assessed considering linear and nonlinear support vector machine classifiers. To find the best combination among the studied factors, we used three approaches: greedy, sequential forward selection (SFS), and exhaustive search. On the basis of our study, we conclude that the factors studied affect the performance of texture methods, so the best combination of these factors should be determined to achieve the best performance with each texture method. SFS can be an appropriate way to approach the factor combination problem because it is less computationally intensive than the other methods.

scholarly journals Motion artifact detection in colonoscopy images

2018 ◽  
Vol 2 (3) ◽  
pp. 171-175
Author(s):  
Rukiye Nur Kaçmaz ◽  
Bülent Yılmaz ◽  
Mehmet Sait Dündar ◽  
Serkan Doğan
Keyword(s):  
Large Intestine ◽  
Evaluation Process ◽  
Motion Artifact ◽  
Histogram Equalization ◽  
Support Vector ◽  
Computer Aided Detection ◽  
Artifact Detection ◽  
Linear Discriminant ◽  
Feature Extraction Method ◽  
Computer Aided

Abstract Computer-aided detection is an integral part of medical image evaluation process because examination of each image takes a long time and generally experts’ do not have enough time for the elimination of images with motion artifact (blurred images). Computer-aided detection is required for both increasing accuracy rate and saving experts’ time. Large intestine does not have straight structure thus camera of the colonoscopy should be moved continuously to examine inside of the large intestine and this movement causes motion artifact on colonoscopy images. In this study, images were selected from open-source colonoscopy videos and obtained at Kayseri Training and Research Hospital. Totally 100 images were analyzed half of which were clear. Firstly, a modified version of histogram equalization was applied in the pre-processing step to all images in our dataset, and then, used Laplacian, wavelet transform (WT), and discrete cosine transform-based (DCT) approaches to extract features for the discrimination of images with no artifact (clear) and images with motion artifact. The Laplacian-based feature extraction method was used for the first time in the literature on colonoscopy images. The comparison between Laplacian-based features and previously used methods such as WT and DCT has been performed. In the classification phase of our study, support vector machines (SVM), linear discriminant analysis (LDA), and k nearest neighbors (k-NN) were used as the classifiers. The results showed that Laplacian-based features were more successful in the detection of images with motion artifact when compared to popular methods used in the literature. As a result, a combination of features extracted using already existing approaches (WT and DCT) and the Laplacian-based methods reached 85% accuracy levels with SVM classification approach

scholarly journals Muscle network topology analysis for the classification of chronic neck pain based on EMG biomarkers extracted during walking

PLoS ONE ◽  
2021 ◽  
Vol 16 (6) ◽  
pp. e0252657
Author(s):  
David Jiménez-Grande ◽  
S. Farokh Atashzar ◽  
Eduardo Martinez-Valdes ◽  
Deborah Falla
Keyword(s):  
Neck Pain ◽  
Network Topology ◽  
Chronic Neck Pain ◽  
Classification Performance ◽  
Support Vector ◽  
Neck Muscle ◽  
Linear Discriminant ◽  
Upper Trapezius ◽  
Intermuscular Coherence

Neuromuscular impairments are frequently observed in patients with chronic neck pain (CNP). This study uniquely investigates whether changes in neck muscle synergies detected during gait are sensitive enough to differentiate between people with and without CNP. Surface electromyography (EMG) was recorded from the sternocleidomastoid, splenius capitis, and upper trapezius muscles bilaterally from 20 asymptomatic individuals and 20 people with CNP as they performed rectilinear and curvilinear gait. Intermuscular coherence was computed to generate the functional inter-muscle connectivity network, the topology of which is quantified based on a set of graph measures. Besides the functional network, spectrotemporal analysis of each EMG was used to form the feature set. With the use of Neighbourhood Component Analysis (NCA), we identified the most significant features and muscles for the classification/differentiation task conducted using K-Nearest Neighbourhood (K-NN), Support Vector Machine (SVM), and Linear Discriminant Analysis (LDA) algorithms. The NCA algorithm selected features from muscle network topology as one of the most relevant feature sets, which further emphasize the presence of major differences in muscle network topology between people with and without CNP. Curvilinear gait achieved the best classification performance through NCA-SVM based on only 16 features (accuracy: 85.00%, specificity: 81.81%, and sensitivity: 88.88%). Intermuscular muscle networks can be considered as a new sensitive tool for the classification of people with CNP. These findings further our understanding of how fundamental muscle networks are altered in people with CNP.

THE EFFECT OF POPULATION CONTEXTS ON CLASSIFIER PERFORMANCE

2008 ◽  
Vol 16 (04) ◽  
pp. 495-517 ◽  
Author(s):  
ASHISH CHOUDHARY ◽  
JIANPING HUA ◽  
MICHAEL L. BITTNER ◽  
EDWARD R. DOUGHERTY
Keyword(s):  
Support Vector Machine ◽  
Nearest Neighbor ◽  
Classification Performance ◽  
Support Vector ◽  
Model Parameters ◽  
Linear Discriminant ◽  
Classifier Design ◽  
Model Based ◽  
Genomics And Proteomics ◽  
Classifier Performance

Classifying a patient based on disease type, treatment prognosis, survivability, or other such criteria has become a major focus of genomics and proteomics. From the perspective of the general population of a particular kind of cell, one would like a classifier that applies to the whole population; however, it is often the case that the population is sufficiently structurally diverse that a satisfactory classifier cannot be designed from available sample data. In such a circumstance, it can be useful to identify cellular contexts within which a disease can be reliably diagnosed, which in effect means that one would like to find classifiers that apply to different sub-populations within the overall population. Using a model-based approach, this paper quantifies the effect of contexts on classification performance as a function of the classifier used and the sample size. The advantage of a model-based approach is that we can vary the contextual confusion as a function of the model parameters, thereby allowing us to compare the classification performance in terms of the degree of discriminatory confusion caused by the contexts. We consider five popular classifiers: linear discriminant analysis, three nearest neighbor, linear support vector machine, polynomial support vector machine, and Boosting. We contrast the case where classification is done with a single classifier without discriminating between the contexts to the case where there are context markers that facilitate context separation before classifier design. We observe that little can be done if there is high contextual confusion, but when the contextual confusion is low, context separation can be beneficial, the benefit depending on the classifier.

scholarly journals Discrimination between Inhibitory and Excitatory Neurons of Mouse Hippocampus Based on the Shape of Extracellular Spike Waveforms

2021 ◽  
Author(s):  
Mehrdad Oghazian ◽  
Farzad Saffari ◽  
Ali Khadem
Keyword(s):  
Classification Performance ◽  
Support Vector ◽  
Svm Classifier ◽  
Shape Features ◽  
K Nearest Neighbors ◽  
Linear Discriminant ◽  
Mouse Hippocampus ◽  
Excitatory Neurons ◽  
Spike Waveforms

Purpose: Inhibitory and excitatory neurons play an essential role in brain function, and we aim to introduce an automatic method to discriminate these two populations based on features of the shape of their spikes. Consequently, we will explain the spike extraction from raw data of a single shank electrode and determine the best features of spike waveforms for the classification of neurons. It is noteworthy that, to the best of our knowledge, classification of inhibitory and excitatory neurons using the shape features extracted from their spike waveforms has not been done before. Materials and Methods: In this paper, we use a dataset of mouse hippocampus neurons in which the neuron types (inhibitory or excitatory) have been verified optogenetically. For the classification of mouse hippocampus neurons, we extracted eight shape features of their spike waveforms in addition to their firing rates and used three types of classifiers: K-Nearest Neighbors (KNN), Linear Discriminant Analysis (LDA), and Support Vector Machine (SVM) to analyze the discriminatory power of features based on the accuracy of the classifications. Results: We showed that Spike asymmetry, Peak-to-trough ratio, Recovery slope, and Duration between peaks were four shape features of spike waveforms participated in the optimum feature subsets that resulted in maximum classification accuracy. Moreover, the SVM classifier with RBF kernel resulted in maximum accuracy of %96.91 ± %13.03 and was identified as the best classifier. Conclusion: In this study, we found that shape features of spike waveforms can accurately classify inhibitory and excitatory neurons of mouse hippocampus. Also, we found an optimum subset of shape features of spike waveforms that resulted in better classification performance than previously proposed subsets of features used for clustering of neurons. Our findings open a promising way toward a functional classification of neurons automatically.

Comparative Study of NMR Spectral Profiling for the Characterization and Authentication of Cannabis

2017 ◽  
Vol 100 (5) ◽  
pp. 1356-1364 ◽  
Author(s):  
Xinyi Wang ◽  
Peter de B Harrington ◽  
Steven F Baugh
Keyword(s):  
Discriminant Analysis ◽  
Nmr Spectroscopy ◽  
Plant Extracts ◽  
Screening Method ◽  
Fuzzy Rule ◽  
Classification Performance ◽  
Support Vector ◽  
Classification Methods ◽  
Linear Discriminant ◽  
Projected Difference Resolution

Abstract For the authentication of botanical materials, itis difficult to obtain representative reference materials because botanicals vary significantly with respect to cultivation conditions. Chemical profiling of plant extracts or spectral fingerprinting can differentiate botanicals and group them by their chemical profiles. NMR spectroscopy yields a powerful and useful method for profiling plant extracts. Both 500 MHz 1H and 1H-1H correlation NMR spectroscopy coupled with pattern recognition were used to discriminate among Cannabis samples. A rapid method of analysis was achieved by extracting directly into the deuterated solvent. Spectral ranges including or excluding the downfield region were compared to evaluate the effect on classification accuracy by projected difference resolution. Six classification methods—fuzzy rule-building expert system, linear discriminant analysis (LDA), super partial least-squares discriminant analysis, support vector machine (SVM), and SVMclassification trees (SVMTrees)—all gave better classification performance for proton NMR spectrathan for proton-proton correlation NMR spectra for seven Cannabis samples. Among the classification methods for a set of 25 Cannabis samples, the 0.5–7.2 plus 7.4–13.0 ppm ranges gave higher prediction rates of greater than 96% when compared to the reduced range of 0.5–7.2 ppm that excluded the downfield range. The LDA method had the best prediction accuracy of 99.8 ± 0.4%. SVMTree methods provide a robust tool, and classification trees are amenable to interpretation. Hence, NMR spectroscopy combined withchemometrics could be used as a fast screening method for the authentication of Cannabis samples.

scholarly journals MCI Detection Using Kernel Eigen-Relative-Power Features of EEG Signals

Actuators ◽  
2021 ◽  
Vol 10 (7) ◽  
pp. 152
Author(s):  
Yu-Tsung Hsiao ◽  
Chia-Fen Tsai ◽  
Chien-Te Wu ◽  
Thanh-Tung Trinh ◽  
Chun-Ying Lee ◽  
...  
Keyword(s):  
Early Detection ◽  
Classification Accuracy ◽  
Spectral Power ◽  
Principal Component ◽  
Classification Performance ◽  
Kernel Principal Component Analysis ◽  
Support Vector ◽  
Relative Power ◽  
Linear Discriminant ◽  
Nonlinear Classifier

Classification between individuals with mild cognitive impairment (MCI) and healthy controls (HC) based on electroencephalography (EEG) has been considered a challenging task to be addressed for the purpose of its early detection. In this study, we proposed a novel EEG feature, the kernel eigen-relative-power (KERP) feature, for achieving high classification accuracy of MCI versus HC. First, we introduced the relative powers (RPs) between pairs of electrodes across 21 different subbands of 2-Hz width as the features, which have not yet been used in previous MCI-HC classification studies. Next, the Fisher’s class separability criterion was applied to determine the best electrode pairs (five electrodes) as well as the frequency subbands for extracting the most sensitive RP features. The kernel principal component analysis (kernel PCA) algorithm was further performed to extract a few more discriminating nonlinear principal components from the optimal RPs, and these components form a KERP feature vector. Results carried out on 51 participants (24 MCI and 27 HC) show that the newly introduced subband RP feature showed superior classification performance to commonly used spectral power features, including the band power, single-electrode relative power, and also the RP based on the conventional frequency bands. A high leave-one-participant-out cross-validation (LOPO-CV) classification accuracy 86.27% was achieved by the RP feature, using a simple linear discriminant analysis (LDA) classifier. Moreover, with the same classifier, the proposed KERP further improved the accuracy to 88.24%. Finally, cascading the KERP feature to a nonlinear classifier, the support vector machine (SVM), yields a high MCI-HC classification accuracy of 90.20% (sensitivity = 87.50% and specificity = 92.59%). The proposed method demonstrated a high accuracy and a high usability (only five electrodes are required), and therefore, has great potential to further develop an EEG-based computer-aided diagnosis system that can be applied for the early detection of MCI.

Sign in / Sign up

Export Citation Format

Share Document