Combination Of Image Features Research Articles

Quantitative Diagnosis of Tongue Cancer from Histological Images in an Animal Model.

We developed a chemically-induced oral cancer animal model and a computer aided method for tongue cancer diagnosis. The animal model allows us to monitor the progress of the lesions over time. Tongue tissue dissected from mice was sent for histological processing. Representative areas of hematoxylin and eosin (H&E) stained tissue from tongue sections were captured for classifying tumor and non-tumor tissue. The image set used in this paper consisted of 214 color images (114 tumor and 100 normal tissue samples). A total of 738 color, texture, morphometry and topology features were extracted from the histological images. The combination of image features from epithelium tissue and its constituent nuclei and cytoplasm has been demonstrated to improve the classification results. With ten iteration nested cross validation, the method achieved an average sensitivity of 96.5% and a specificity of 99% for tongue cancer detection. The next step of this research is to apply this approach to human tissue for computer aided diagnosis of tongue cancer.

No-reference Image Quality Assessment With A Gradient-induced Dictionary

Image distortions are typically characterized by degradations of structures. Dictionaries learned from natural images can capture the underlying structures in images, which are important for image quality assessment (IQA). This paper presents a general-purpose no-reference image quality metric using a GRadient-Induced Dictionary (GRID). A dictionary is first constructed based on gradients of natural images using K-means clustering. Then image features are extracted using the dictionary based on Euclidean-norm coding and max-pooling. A distortion classification model and several distortion-specific quality regression models are trained using the support vector machine (SVM) by combining image features with distortion types and subjective scores, respectively. To evaluate the quality of a test image, the distortion classification model is used to determine the probabilities that the image belongs to different kinds of distortions, while the regression models are used to predict the corresponding distortion-specific quality scores. Finally, an overall quality score is computed as the probability-weighted distortion-specific quality scores. The proposed metric can evaluate image quality accurately and efficiently using a small dictionary. The performance of the proposed method is verified on public image quality databases. Experimental results demonstrate that the proposed metric can generate quality scores highly consistent with human perception, and it outperforms the state-of-the-arts.

Small (< 4 cm) Renal Mass: Differentiation of Oncocytoma From Renal Cell Carcinoma on Biphasic Contrast-Enhanced CT

The purpose of this study was to evaluate whether small (< 4 cm) oncocytomas can be differentiated from renal cell carcinomas (RCCs) on biphasic contrast-enhanced CT. Forty-three patients with 53 oncocytomas and 123 patients with 128 RCCs (24 papillary subtype and 104 clear cell and other subtypes) who underwent biphasic contrast-enhanced CT were included in the study. Patient demographics and CT tumor characteristics were evaluated in each case. A multinomial logistic regression model was then constructed for differentiating oncocytoma from clear cell and other subtype RCCs, oncocytoma from papillary RCCs, and clear cell and other subtype RCCs from papillary RCCs. The probability of each group was calculated from the model. Diagnostic performance among three pairwise diagnoses and between oncocytoma and any RCC (clear cell and other subtypes and papillary) were assessed by AUC values. Patient age, tumor CT attenuation values and skewness (i.e., histogram analysis of CT values) in both the corticomedullary and nephrographic phases, and subjective tumor heterogeneity were statistically significant variables in the multinomial logistic regression analysis. The logistic regression model using the variables yielded AUCs of 0.82, 0.95, 0.91, and 0.84 for differentiating oncocytomas from clear cell and other subtype RCCs, oncocytomas from papillary RCCs, clear cell and other subtype RCCs from papillary RCCs, and oncocytomas from any RCC (clear cell and other subtypes and papillary), respectively. A combination of imaging features on biphasic CT, including tumor CT attenuation values and tumor texture (heterogeneity and skewness), can help differentiate oncocytoma from RCC.

A review of fusion methods of multi-spectral image

As an important branch of data fusion taken images as the research objects, image fusion perform multiple images to get a more accurate image using redundant information and complementary information. Multi-spectral image is a kind of remote sensing image, and fusion of multi-spectral image combine image features of multi-spectral image together to get a more comprehensive and clear image using the spatiotemporal correlation and information on complementary. Consequently, fusion of multi-spectral image, which is a hot issue, is an important way of information processing of remote sensing image. Fusion methods of multi-spectral image is an important issue of fusion of multi-spectral image of remote sensing image, and effective selection of an appropriate fusion method of multi-spectral image is especially significant for improving image accuracy. Along with the development of remote sensing technique, traditional fusion methods of image are difficult to meet the requirement of image accuracy. Recently, fusion methods of multi-spectral image attract increasing attention and become a new hot topic. In this paper, characteristics of different fusion methods of multi-spectral image as well as the research prospect are analyzed. This paper provides a scientific reference for the development of fusion technique of multi-spectral image.

Myxoid soft-tissue neoplasms: comprehensive update of the taxonomy and MRI features.

OBJECTIVE. The purpose of this article is to review the classification, clinical presentation, and histopathologic and MRI features of myxoid soft-tissue neoplasms. CONCLUSION. MRI is the modality of choice for characterization of myxoid soft-tissue tumors. A combination of imaging features (including certain characteristic signs), clinical features, and patient demographics can help the radiologist in coming to a specific diagnosis or in narrowing down the differential diagnoses.

Shelf-Life Prediction of ‘Gros Michel’ Bananas with Different Browning Levels Using Hyperspectral Reflectance Imaging

A hyperspectral imaging system was developed in the spectral region between 400 and 1,100 nm to investigate its potential for predicting the shelf life of bananas with different browning levels. Principal component analysis (PCA) was conducted for reducing data dimensionality and selecting optimal wavelengths. Five optimal wavelengths (454, 486, 559, 686, and 728 nm) were found. Among all principal component (PC) images, PC-4 was selected to segregate browning area from normal surface using a simple threshold algorithm for extracting image features of browning area. The average spectra were obtained by calculating the mean of the spectra of all browning areas in the hyperspectral images of bananas. Then, image features and average spectra in five optimal wavelengths were used to develop classification models for predicting the shelf life of banana samples by determining their browning levels using back propagation (BP), radial basis function (RBF), and self-organizing feature maps (SOM) networks. Results indicated that the classification models using both image features and average spectra were obviously superior to those models using image features or average spectra alone. Among all classifiers, BP classifier had the best performance with the best classification rates of 95.6 % for training set and 90.5 % for testing set, respectively. The results demonstrated that hyperspectral imaging has great potential in predicting banana shelf life based on combination of image features and average spectra.

2A2-R04 画像特徴と形状特徴を複合したランドマークに基づくGraph SLAM(移動ロボットの自己位置推定と地図構築)

In this study, we aim to achieve an autonomous locomotion of the mobile robot in the unknown indoor environment. In particular, we propose a novel landmark construction and detection method. The landmark is generated by combined image and shape features. In the landmark detection step, we represent the new matching with confidences defined automatically. The effectiveness of the proposed landmark is verified by experiment. Moreover, we introduce the novel landmark based graph SLAM. In our method, that landmark detection is performed on each node. Then, if the robot find the landmark it found once, local loop closure is generated and optimization is performed. The advantage of this method is that we can perform graph optimization before find loop-closure. The effectiveness of this proposal graph SLAM is verified by simulation and experiment.

Fire detection system using random forest classification for image sequences of complex background

We present a fire alarm system based on image processing that detects fire accidents in various environments. To reduce false alarms that frequently appeared in earlier systems, we combined image features including color, motion, and blinking information. We specifically define the color conditions of fires in hue, saturation and value, and RGB color space. Fire features are represented as intensity variation, color mean and variance, motion, and image differences. Moreover, blinking fire features are modeled by using crossing patches. We propose an algorithm that classifies patches into fire or nonfire areas by using random forest supervised learning. We design an embedded surveillance device made with acrylonitrile butadiene styrene housing for stable fire detection in outdoor environments. The experimental results show that our algorithm works robustly in complex environments and is able to detect fires in real time.

MR imaging features of adrenal rest tumor

Objective To investigate the imaging features of adrenal rest tumor.Methods Twelve patients of adrenal rest tumor proved by surgery or clinical diagnosis were retrospectively analyzed.Among these 12 patients,12 were examined with ultrasound,11 with MR and 1 with CT. MR and CT were performed without and with intravenous injection of contrast material.The imaging features of adrenal rest tumor were retrospectively summarized and the relevant literatures reviewed. Results The adrenal rest tumors were found in testis in 10 of the 12 patients,and in ovaries and broad ligament in the remaining two.The imaging features of the testicular adrenal rest tumor were summarized as following:all patients had bilateral testicular masses without change of the testicular contour. On ultrasonography,the lesions were hypoechoic, with some hyperechoic areas and appeared highly vascularized on Colour Doppler ultrasonography.The masses showed iso-density on plain CT,and avid enhancement on post-contrast CT images.The masses ranging in size from0.7 cm×1.0 cm×2.2 cm to 2.3 cm ×2.7 cm ×2.9 cm with uniform signal intensity,lobulated margin on MRI.They exhibited iso- or slight hyperintensity on T1 WI and hypointensity on T2WI relative to normal testicular parenchyma.The tumors showed intense enhancement on post-contrast MR images. No abnormality was detected with Colour Doppler uhrasonography and MR in 2 patients of adrenal rest tumor in ovaries and broad ligament. Conclusion Combining imaging features with the typical clinical history,the diagnosis of adrenal rest tumor could be suggested pre-operatively. Key words: Adrenal hyperplasia, congenital ; Testicular neoplasms ; Ovarian neoplasms; Comparative study

A semantic image classifier based on hierarchical fuzzy association rule mining

One of the major challenges in the content-based information retrieval and machine learning techniques is to-build-the-so-called "semantic classifier" which is able to effectively and efficiently classify semantic concepts in a large database. This paper dealt with semantic image classification based on hierarchical Fuzzy Association Rules (FARs) mining in the image database. Intuitively, an association rule is a unique and significant combination of image features and a semantic concept, which determines the degree of correlation between features and concept. The main idea behind this approach is that any image visual concept has some associated features, so that, there are strong correlations between the concepts and their corresponding features. Regardless of the semantic gap, an image concept appears when the corresponding features emerge in an image and vice versa. Specially, this paper's contribution was to propose a novel Fuzzy Association Rule for improving traditional association rules. Moreover, it was concerned with establishing a hierarchical fuzzy rule base in the training phase and setup corresponding fuzzy inference engine in order to classify images in the testing phase. The presented approach was independent from image segmentation and can be applied on multi-label images. Experimental results on a database of 6000 general-purpose images demonstrated the superiority of the proposed algorithm.

Imaging features of abdominal tuberculosis on multidetector CT and its mimics

Background: Abdominal tuberculosis is endemic in developing countries and its incidence is on an increase in the western world. The involvement of gastrointestinal tract is seen in 65-78% cases of abdominal tuberculosis with ileocaecal region being the commonest site followed by colon and jejunum. The disease can mimic various gastrointestinal disorders particularly inflammatory bowel disease, colonic malignancy or other gastro-intestinal infections. MDCT (Multidetector CT) because of multiplanar imaging plays an important role in diagnosis of abdominal tuberculosis due to excellent demonstration of changes in bowel, mesentery, lymph nodes, peritoneum, omentum and solid organs. Methods: This study was done in Dept. of Radio-diagnosis, Kasturba Medical College, Manipal between April - December 2011 where MDCT features of 23 patients with findings suspicious for Abdominal tuberculosis were evaluated. All patients underwent both plain and contrast enhanced CT (CECT). Patients with imaging features of genitourinary tuberculosis and those with HIV positive status were not included in the study. Results: Enhancing wall thickening involving ascending colon, caecum and ileo-caecal junction with adjacent lymph nodes was the most common finding seen in 8 cases. Peritoneal involvement in the form of ascites with enhancing and thickened parietal and visceral peritoneum was noted in 5 cases. Multiple strictures involving small bowel was the next most common finding in 3 cases. Lymph nodal involvement was seen in the form of mesenteric and retro-peritoneal lymphnodes with some of them showing caseation necrosis or calcific foci in 3 cases. Hepatic lesion which was proven to be tuberculosis on histopathology was seen in 1 case. There were 2 cases with thickened small bowel wall which were confirmed to be inflammatory bowel disease on histopathology and 1 case which showed ulcero-proliferative pattern which was confirmed to be malignant. Conclusion: There are variable imaging features of abdominal tuberculosis which are better demonstrated on MDCT. Although no single feature is diagnostic, a combination of imaging features when correlated with clinical and laboratory data can help in clinching the diagnosis.

Review on Singular-Value-Decomposition-Based Digital Watermarking Algorithm in Combination with Image Features

The combination of image features with singular value decomposition algorithm and digital watermarking algorithm based on wavelet transform respectively makes possible the advancement from pixel watermarking to content watermarking, which effectively solves the contradiction between perceptibility and robustness of conventional watermarking algorithm. Therefore, stronger robustness and lower perceptibility of watermarking are achieved. This paper first gives a review on conventional singular-value-decomposition-based digital watermarking algorithm, and then makes a thorough analysis of its respective features, advantages and defects. Improved feature extraction schemes of digital watermark which combine image features with SVD algorithm as well as the application algorithm in feature extraction based on improved SVD algorithm are put forward for the comparative analysis of corresponding principles and processing effects. The experimental results indicate that the content-based (instead of pixel-based) watermarking algorithm can better satisfy the perceptibility and robustness of digital watermarking, with huge application potentials and more development space

Fungal Damage Detection in Wheat Using Short-Wave Near-Infrared Hyperspectral and Digital Colour Imaging

Healthy and fungal-damaged wheat kernels infected by the species of storage fungi, namely Penicillium spp., Aspergillus glaucus, and A. niger, were scanned using a short-wave near-infrared hyperspectral imaging system in the 700–1100 nm wavelength range and an area scan colour camera. A multivariate image analysis was used to reduce the dimensionality of the hyperspectral data and to select the significant wavelength using principal component analysis. Wavelength 870 nm, which corresponded to the highest factor loading of first principal component, was considered to be significant. Statistical and histogram features from the 870 nm wavelength image were selected and used as input to statistical discriminant classifiers (linear, quadratic, and Mahalanobis). From the colour images, a total of 179 features (123 colour and 56 textural) were extracted and the top features selected from these features were used as input to the statistical classifiers. The linear discriminant analysis classifier correctly classified 97.3–100.0% healthy and fungal-infected wheat kernels, using the combined hyperspectral image features and the top ten features selected from 179 colour and textural features of the colour images as input.

Carotid artery recognition system: A comparison of three automated paradigms for ultrasound images

The development of completely automated techniques for arterial wall segmentation and intima-media thickness measurement requires the recognition of the artery in the image frame. Conceptually, automated techniques can be thought of as the combination of two cascaded stages: artery recognition and wall segmentation. In this paper, the authors show three carotid artery recognition systems (CARS) that are fully automated. The first technique is based on a first-order derivative Gaussian edge analysis (CARSgd). The second method is based on an integrated approach (CARSia) that combines image feature extraction, fitting, and classification. The third strategy is based on signal analysis (CARSsa). The output of all the three paradigms provide tracing of the far adventitial (AD(F)). The authors validated CARSgd, CARSia, and CARSsa on a dataset of 365 longitudinal B-Mode carotid images, acquired by different sonographers. Performance evaluation of the carotid recognition process was done in three ways: (1) visual inspection by experts; (2) by measuring the Hausdorff distance (HD) between the automatic far adventitial (AD(F)) and the manually traced AD(F), and (3) by measuring the HD between AD(F) and the lumen-intima (GT(LI)) and media-adventitia (GT(MA)) borders of the arterial walls. The average HD between AD(F) and the manual AD(F) was 1.53 ± 1.51 mm for CARSgd, 1.82 ± 3.08 mm for CARSia, and 2.56 ± 2.89 mm for CARSsa. The average HD between GT(LI) and AD(F) for CARSgd, CARSia, and CARSsa were 2.16 ± 1.16 mm, 2.71 ± 2.89 mm, and 2.66 ± 1.52 mm, respectively. The average HD between AD(F) and GT(MA) for CARSgd, CARSia, and CARSsa were 1.54 ± 1.19 mm, 1.86 ± 2.66 mm, and 1.95 ± 1.64 mm, respectively. Considering a maximum distance of 50 pixels (about 3 mm), CARSgd showed an identification accuracy of 100%, CARSia of 92%, and CARSsa of 96%. These identification accuracies were confirmed by visual inspection. All the three systems work on MATLAB, Windows OS, and on a PC based cross platform medical application written in Java called ATHEROEDGE™ with 1 s per image. CARSgd showed very accurate AD(F) profiles coupled with a low computational burden and without the need for specific tuning. It can be thought of as a reference technique for carotid localization, to be used in automated intima-media thickness measurement strategies.

Histological Image Feature Mining Reveals Emergent Diagnostic Properties for Renal Cancer.

Computer-aided histological image classification systems are important for making objective and timely cancer diagnostic decisions. These systems use combinations of image features that quantify a variety of image properties. Because researchers tend to validate their diagnostic systems on specific cancer endpoints, it is difficult to predict which image features will perform well given a new cancer endpoint. In this paper, we define a comprehensive set of common image features (consisting of 12 distinct feature subsets) that quantify a variety of image properties. We use a data-mining approach to determine which feature subsets and image properties emerge as part of an "optimal" diagnostic model when applied to specific cancer endpoints. Our goal is to assess the performance of such comprehensive image feature sets for application to a wide variety of diagnostic problems. We perform this study on 12 endpoints including 6 renal tumor subtype endpoints and 6 renal cancer grade endpoints. Keywords-histology, image mining, computer-aided diagnosis.

CONTENT-BASED IMAGE RETRIEVAL OF CULTURAL HERITAGE SYMBOLS BY INTERACTION OF VISUAL PERSPECTIVES

Content-based Image Retrieval (CBIR) has been an active area of research for retrieving similar images from large repositories, without the prerequisite of manual labeling. Most current CBIR algorithms can faithfully return a list of images that matches the visual perspective of their inventors, who might decide to use a certain combination of image features like edges, colors and textures of regions as well as their spatial distribution during processing. In practice, however, the retrieved images rarely correspond exactly to the results expected by the users, a problem that has come to be known as the semantic gap. In this paper, we propose a novel and extensible multidimensional approach called matrix of visual perspectives as a solution for addressing this semantic gap. Our approach exploits the dynamic cross-interaction (in other words, mix-and-match) of image features and similarity metrics to produce results that attempt to mimic the mental visual picture of the user. Experimental results on retrieving similar Japanese cultural heritage symbols called kamons by a prototype system confirm that the interaction of visual perspectives in the user can be effectively captured and reflected. The benefits of this approach are broader. They can be equally applicable to the development of CBIR systems for other types of images, whether cultural or noncultural, by adapting to different sets of application specific image features.

MCs Detection with Combined Image Features and Twin Support Vector Machines

Breast cancer is a common form of cancer diagnosed in women. Clustered microcalcifications(MCs) in mammograms is one of the important early sign. Their accurate detection is a key problem in computer aided detection (CDAe). In this paper, a novel approach based on the recently developed machine learning technique - twin support vector machines (TWSVM) to detect MCs in mammograms. The ground truth of MCs in mammograms is assumed to be known as a priori. First each MCs is preprocessed by using a simple artifact removal filter and a high-pass filter. Then the combined image feature extractors are employed to extract 164 image features. In the combined feature domain, the MCs detection procedure is formulated as a supervised learning and classification problem, and the trained TWSVM is used as a classifier to make decision for the presence of MCs or not. A large number of experiments were carried out to evaluate and compare the performance of the proposed MCs detection algorithm. Experimental results show that the proposed TWSVM classifier is more advantageous for real-time processing of MCs in mammograms.

A genetic programming framework for content-based image retrieval

The effectiveness of content-based image retrieval (CBIR) systems can be improved by combining image features or by weighting image similarities, as computed from multiple feature vectors. However, feature combination do not make sense always and the combined similarity function can be more complex than weight-based functions to better satisfy the users’ expectations. We address this problem by presenting a Genetic Programming framework to the design of combined similarity functions. Our method allows nonlinear combination of image similarities and is validated through several experiments, where the images are retrieved based on the shape of their objects. Experimental results demonstrate that the GP framework is suitable for the design of effective combinations functions.

A rare case of inflammatory pseudotumour of the submandibular lymphnode.

Inflammatory pseudotumour (IPT), also known as inflammatory myofibroblastic tumour, is a rare cause of benign cervical lymphadenopathy. In its acute clinical presentation it mimics malignant and infective causes of cervical lymphadenopathy. The combined clinical history, histopathological findings, and imaging features of an infiltrating soft-tissue mass should lead to the possibility of IPT. Radical surgery should be avoided before there is histological proof of a malignant tumor.

A feature-based robust digital image watermarking scheme

A robust digital image watermarking scheme that combines image feature extraction and image normalization is proposed. The goal is to resist both geometric distortion and signal processing attacks. We adopt a feature extraction method called Mexican hat wavelet scale interaction. The extracted feature points can survive a variety of attacks and be used as reference points for both watermark embedding and detection. The normalized image of an image (object) is nearly invariant with respect to rotations. As a result, the watermark detection task can be much simplified when it is applied to the normalized image. However, because image normalization is sensitive to image local variation, we apply image normalization to nonoverlapped image disks separately. The disks are centered at the extracted feature points. Several copies of a 16-bit watermark sequence are embedded in the original image to improve the robustness of watermarks. Simulation results show that our scheme can survive low-quality JPEG compression, color reduction, sharpening, Gaussian filtering, median filtering, row or column removal, shearing, rotation, local warping, cropping, and linear geometric transformations.