K. Raghesh Krishnan

Bio: K. Raghesh Krishnan is an academic researcher from Amrita Vishwa Vidyapeetham. The author has contributed to research in topics: Feature extraction & Active contour model. The author has an hindex of 3, co-authored 8 publications receiving 59 citations.

Hybrid approach to classification of focal and diffused liver disorders using ultrasound images with wavelets and texture features

Abstract: This study presents a computer-based approach to classify ten different kinds of focal and diffused liver disorders using ultrasound images. The diseased portion is isolated from the ultrasound image by applying active contour segmentation technique. The segmented region is further decomposed into horizontal, vertical and diagonal component images by applying biorthogonal wavelet transform. From the above wavelet filtered component images, grey level run-length matrix features are extracted and classified using random forests by applying ten-fold cross-validation strategy. The results are compared with spatial feature extraction techniques such as intensity histogram, invariant moment features and spatial texture features such as grey-level co-occurrence matrices, grey-level run length matrices and fractal texture features. The proposed technique, which is an application of texture feature extraction on transform domain images, gives an overall classification accuracy of 91% for a combination of ten classes of similar looking diseases which is appreciable than the spatial domain only techniques for liver disease classification from ultrasound images.

Automatic Classification of Liver Diseases from Ultrasound Images Using GLRLM Texture Features

Abstract: Ultrasound imaging is considered to be one of the most cost-effective and non-invasive techniques for conclusive diagnosis in some cases and preliminary diagnosis in others. Automatic liver tissue characterization and classification from ultrasonic scans have been for long, the concern of many researchers, and has been made possible today by the availability of the most powerful and cost effective computing facilities. Automatic diagnosis and classification systems are used both for quick and accurate diagnosis and as a second opinion tool for clarifications. This paper analyzes the effect of various linear, non linear and diffusion filters in improving the quality of the liver ultrasound images before proceeding to the subsequent phases of feature extraction and classification using Gray Level Run Length Matrix Features and Support Vector Machines respectively.

Tensor Flow Based Analysis and Classification of Liver Disorders from Ultrasonography Images

Abstract: In the field of medical imaging, Ultrasonography is a popular and most frequently used diagnostic tool owing to its hazard-free, non–invasive and the cost effective nature. Liver being the largest and vital organ in the human body, liver disorders are treated very important and initial detection of the disorder is made using ultrasound imaging by the radiologists that leads to additional biopsies for confirmation, if necessary. This work focusses on the automated classification of nine types of both focal and diffused liver disorders using ultrasound images. A deep convolutional neural network architecture codenamed Inception is used. The technique achieves a new state for classification and detection of liver disease. The disease is predicted based on the score obtained as a result of training. The classification is achieved using tensor flow and it outputs the predicted labels and the corresponding scores. The method achieves reasonable accuracy using the trained model.

Analyzing Underwater Videos for Fish Detection, Counting and Classification

Abstract: Underwater video processing is a valuable tool for analyzing the presence and behaviour of fishes in underwater. Video based analysis of fishes finds its use in aquaculture, fisheries and protection of fishes in oceans. This paper proposes a system to detect, count, track and classify the fishes in underwater videos. In the proposed approach, two systems are developed, a counting system, which uses gaussian mixture model (for foreground detection), morphological operations (for denoising), blob analysis (for counting) and kalman filtering (for tracking), and a classification system, which uses bag of features approach that is used to classify the fishes. In the bag of feature approach, surf features are extracted from the images to obtain feature descriptors. k-mean clustering is applied on the feature descriptors, to obtain visual vocabulary. The test features are input to the MSVM classifier, which uses visual vocabulary to classify the images. The proposed system achieves an average accuracy of 90% in counting and 88.9% in classification, respectively.

A study of the phases of classification of liver diseases from ultrasound images and gray level difference weights based segmentation

Abstract: This paper presents a study of the state of the art techniques applied to computer based analysis and classification of liver diseases from ultrasound images. The diseased portions from the ultrasound images are analyzed and categorized using techniques such as Despeckling, Segmentation, Feature extraction and Classification. Automatic segmentation of ultrasound images is complicated due to the fact that the image may include other organs which are close to the liver, irregular structure of disease, poor quality of image, lack of color cues, and lack of definite boundaries and presence of noise. This work makes a study of different techniques used in the different phases of biomedical liver ultrasound processing such as noise removal, segmentation, Feature Extraction and classification. This work also presents the segmentation results obtained using Gray Level Difference Weights Method on 10 types of liver diseases from ultrasound images.

Learning to Diagnose Cirrhosis with Liver Capsule Guided Ultrasound Image Classification.

Abstract: This paper proposes a computer-aided cirrhosis diagnosis system to diagnose cirrhosis based on ultrasound images. We first propose a method to extract a liver capsule on an ultrasound image, then, based on the extracted liver capsule, we fine-tune a deep convolutional neural network (CNN) model to extract features from the image patches cropped around the liver capsules. Finally, a trained support vector machine (SVM) classifier is applied to classify the sample into normal or abnormal cases. Experimental results show that the proposed method can effectively extract the liver capsules and accurately classify the ultrasound images.

Assessment of combined textural and morphological features for diagnosis of breast masses in ultrasound

Abstract: The objective of this study is to assess the combined performance of textural and morphological features for the detection and diagnosis of breast masses in ultrasound images We have extracted a total of forty four features using textural and morphological techniques Support vector machine (SVM) classifier is used to discriminate the tumors into benign or malignant The performance of individual as well as combined features are assessed using accuracy(Ac), sensitivity(Se), specificity(Sp), Matthews correlation coefficient(MCC) and area AZ under receiver operating characteristics curve The individual features produced classification accuracy in the range of 6166% and 9083% and when features from each category are combined, the accuracy is improved in the range of 7916% and 9583% Moreover, the combination of gray level co-occurrence matrix (GLCM) and ratio of perimeters (P ratio ) presented highest performance among all feature combinations (Ac 9585%, Se 96%, Sp 9146%, MCC 09146 and AZ 09444)The results indicated that the discrimination performance of a computer aided breast cancer diagnosis system increases when textural and morphological features are combined

Magnetic Resonance-Based Texture Analysis Differentiating KRAS Mutation Status in Rectal Cancer.

Abstract: PURPOSE Mutation of the Kirsten Ras (KRAS) oncogene is present in 30%-40% of colorectal cancers and has prognostic significance in rectal cancer. In this study, we examined the ability of radiomics features extracted from T2-weighted magnetic resonance (MR) images to differentiate between tumors with mutant KRAS and wild-type KRAS. Materials and Methods Sixty patients with primary rectal cancer (25 with mutant KRAS, 35 with wild-type KRAS) were retrospectively enrolled. Texture analysis was performed in all regions of interest on MR images, which were manually segmented by two independent radiologists. We identified potentially useful imaging features using the two-tailed t test and used them to build a discriminant model with a decision tree to estimate whether KRAS mutation had occurred. RESULTS Three radiomic features were significantly associated with KRASmutational status (p < 0.05). The mean (and standard deviation) skewness with gradient filter value was significantly higher in the mutant KRAS group than in the wild-type group (2.04±0.94 vs. 1.59±0.69). Higher standard deviations for medium texture (SSF3 and SSF4) were able to differentiate mutant KRAS (139.81±44.19 and 267.12±89.75, respectively) and wild-type KRAS (114.55±29.30 and 224.78±62.20). The final decision tree comprised three decision nodes and four terminal nodes, two of which designated KRAS mutation. The sensitivity, specificity, and accuracy of the decision tree was 84%, 80%, and 81.7%, respectively. CONCLUSION Using MR-based texture analysis, we identified three imaging features that could differentiate mutant from wild-type KRAS. T2-weighted images could be used to predict KRAS mutation status preoperatively in patients with rectal cancer.

Classification of focal liver lesions on ultrasound images by extracting hybrid textural features and using an artificial neural network

Abstract: This paper focuses on the improvement of the diagnostic accuracy of focal liver lesions by quantifying the key features of cysts, hemangiomas, and malignant lesions on ultrasound images. The focal liver lesions were divided into 29 cysts, 37 hemangiomas, and 33 malignancies. A total of 42 hybrid textural features that composed of 5 first order statistics, 18 gray level co-occurrence matrices, 18 Law's, and echogenicity were extracted. A total of 29 key features that were selected by principal component analysis were used as a set of inputs for a feed-forward neural network. For each lesion, the performance of the diagnosis was evaluated by using the positive predictive value, negative predictive value, sensitivity, specificity, and accuracy. The results of the experiment indicate that the proposed method exhibits great performance, a high diagnosis accuracy of over 96% among all focal liver lesion groups (cyst vs. hemangioma, cyst vs. malignant, and hemangioma vs. malignant) on ultrasound images. The accuracy was slightly increased when echogenicity was included in the optimal feature set. These results indicate that it is possible for the proposed method to be applied clinically.