Sevcan Aytac Korkmaz

Cancer detection in mammograms estimating feature weights via Kullback-Leibler measure

In this study the aim is to determine cancerous possibility of suspicious lesions in mammograms. With this aim, probabilistic values of suspicious lesions in the image are found via exponential curve fitting and texture features in order to find weight values in the objective function. Afterwards, images are classified as normal, malign, and benign by utilizing Kullback Leibler method. Here, 3×10 mammography images set selected from Digital Database for Screening Mammography (DDSM) are used, and severity of disease is probabilistically estimated. Results are indicated on a scale to eliminate the suspicious lesions. Thus, it is considered that workload of clinicians shall be reduced by easily eliminating suspicious images out of many mammography images.

RETRACTED: Diagnosis of cervical cancer cell taken from scanning electron and atomic force microscope images of the same patients using discrete wavelet entropy energy and Jensen Shannon, Hellinger, Triangle Measure classifier

The aim of this article is to provide early detection of cervical cancer by using both Atomic Force Microscope (AFM) and Scanning Electron Microscope (SEM) images of same patient. When the studies in the literature are examined, it is seen that the AFM and SEM images of the same patient are not used together for early diagnosis of cervical cancer. AFM and SEM images can be limited when using only one of them for the early detection of cervical cancer. Therefore, multi-modality solutions which give more accuracy results than single solutions have been realized in this paper. Optimum feature space has been obtained by Discrete Wavelet Entropy Energy (DWEE) applying to the 3×180 AFM and SEM images. Then, optimum features of these images are classified with Jensen Shannon, Hellinger, and Triangle Measure (JHT) Classifier for early diagnosis of cervical cancer. However, between classifiers which are Jensen Shannon, Hellinger, and triangle distance have been validated the measures via relationships. Afterwards, accuracy diagnosis of normal, benign, and malign cervical cancer cell was found by combining mean success rates of Jensen Shannon, Hellinger, and Triangle Measure which are connected with each other. Averages of accuracy diagnosis for AFM and SEM images by averaging the results obtained from these 3 classifiers are found as 98.29% and 97.10%, respectively. It has been observed that AFM images for early diagnosis of cervical cancer have higher performance than SEM images. Also in this article, surface roughness of malign AFM images in the result of the analysis made for the AFM images, according to the normal and benign AFM images is observed as larger, If the volume of particles has found as smaller.

A expert system for stomach cancer images with artificial neural network by using HOG features and linear discriminant analysis: HOG_LDA_ANN

In this study, normal (n), benign (b), and malign (m) stomach image cells have taken from faculty of Medicine the Fırat University with Light Microscope help. Total number of stomach images are 180 which be 60 n, 60 b, and 60 m. 90 of these 180 stomach images have been used for testing purposes and 90 have used for training purposes. The histograms of oriented gradient (HOG) feature vectors have been obtained for normal, benign, and malign original stomach images. The size of these HOG feature vectors is 46900×180. High-dimensional of these HOG feature vectors is reduced to lower-dimensional with Linear Discriminant Analysis (LDA). These low-dimensional data are 180×180. These low-dimensional data are classified as normal benign and malign by artificial neural network (ANN) classification. Thus, HOG_LDA_ANN method for stomach cancer images have developed. Diagnostic accuracy of classification results with this method has found as 88.9%. According to the other methods, this result has higher accuracy result. And this result has found in a shorter time.

Recognition of the stomach cancer images with probabilistic HOG feature vector histograms by using HOG features

In this study, normal (n), benign (b), and malign (m) stomach image cells have taken from faculty of Medicine the Fırat University with Light Microscope help. Total number of stomach images are 180 which be 60 n, 60 b, and 60 m. 90 of these 180 stomach images have been used for testing purposes and 90 have been used for training purposes. The histograms of oriented gradient (HOG) feature extraction method were used for these images. HOG feature vectors were obtained by plotting HOG features on normal, benign, and malign original stomach images. Using these HOG property vectors, histograms of normal, benign, and malignant stomach images were plotted. Bins and h histogram values were obtained from these drawn histograms. A bandwidth range that can be distinguished between normal, benign, and malignant stomach images was calculated by comparing the bins and h values obtained for normal (n), benign (b) and malign (m) images. This bandwidth range was found to be 0.09–0.22. According to this bandwidth range, the accuracy result of stomach cancer images is found as 100%. When the h values of the HOG feature vector between these bandwidths are examined, the h values of normal and benign stomach images are found to be higher than those of a malignant stomach image. Between this bandwidth, the h value of the normal stomach image was found to be higher than the benign stomach image.

Retraction notice to "Diagnosis of cervical cancer cell taken from scanning electron and atomic force microscope images of the same patients using discrete wavelet entropy energy and Jensen Shannon, Hellinger, Triangle Measure classifier" [SAA 160 (2016) 39-49].

This article has been retracted: please see Elsevier Policy on Article This manipulation of the peer-review process represents a clear vioWithdrawal (http://www.elsevier.com/locate/withdrawalpolicy). This article has been retracted at the request of the Editors. After a thorough investigation, the Publisher has concluded that the Editor wasmisled into accepting this article based upon the positive advice of at least one suggested reviewer report. The reportwas submitted from an email account provided by the author, that was later determined not to be the email of the supposed expert reviewer.

Recognition of the Gastric Molecular Image Based on Decision Tree and Discriminant Analysis Classifiers by using Discrete Fourier Transform and Features

ABSTRACT This article presents the development and evaluation of a computerized decision support system (DSS), aiming to Show the feasibility and potential toward maximizing the benefits of a new algorithm by combining the machine-learning techniques which are not used in the literature for automatic recognition of the gastric images. The object of this article is fivefold: first, the features Maximally Stable Extremal Regions (MSER), Speeded Up Robust Features (SRF), and Binary Robust Invariant Scalable Keypoints (BRISK) of histopathological gastric images were analyzed. Second, the Fourier Transform (FT) was applied to these properties which were calculated to equalize the dimensions of the obtained features. Third, MS and LE size reduction methods have been applied. Fourth, the decision tree (DT) and discriminant analysis (DA) classifiers are used to classify the histopathological gastric images. Fifth, these classification results have been compared. In this article, the highest accuracy result obtained by using the SRF_FT_MS_DT method is found to be 86.66%. Fast and multimodality computerized DSS can beneficial to patients for early detection of gastric diseases. It may facilitate early diagnosis of the disease.

A New Application Based on GPLVM, LMNN, and NCA for Early Detection of the Stomach Cancer

ABSTRACT In this article, speeded-up robust features (SURF) for each image have been calculated. Discrete Fourier transform (DFT) method has been applied to these SURF. High dimensions of these SURF–DFT feature vectors are reduced to low dimensions with large-margin nearest neighbor (LMNN), Gaussian process latent variable models (GPLVM), and neighborhood component analysis (NCA). When size reduction process was done, effect on the GPLVM, LMNN, and NCA of the 1, 2, 3, 4, 5, 6, 7, 8, 9, and 10 feature numbers has been examined. These features are classified by naive Bayes (NB) classifier. Thus, SURF_DFT_GPLVM_NB, SURF_DFT_NCA_NB, and SURF_DFT_LMNN_NB methods for gastric histopathological images have been developed. Classification results obtained with these methods have been compared. According to the obtained results, the highest classification result was obtained as 90.24% by using 4 features by SURF_DFT_GPLVM_NB method for second group images.