Image quality assessment in digital mammography: part I. Technical characterization of the systems.

In many European countries, image quality for digital x-ray systems used in screening mammography is currently specified using a threshold-detail detectability method. This is a two-part study that proposes an alternative method based on calculated detectability for a model observer: the first part of the work presents a characterization of the systems. Eleven digital mammography systems were included in the study; four computed radiography (CR) systems, and a group of seven digital radiography (DR) detectors, composed of three amorphous selenium-based detectors, three caesium iodide scintillator systems and a silicon wafer-based photon counting system. The technical parameters assessed included the system response curve, detector uniformity error, pre-sampling modulation transfer function (MTF), normalized noise power spectrum (NNPS) and detective quantum efficiency (DQE). Approximate quantum noise limited exposure range was examined using a separation of noise sources based upon standard deviation. Noise separation showed that electronic noise was the dominant noise at low detector air kerma for three systems; the remaining systems showed quantum noise limited behaviour between 12.5 and 380 µGy. Greater variation in detector MTF was found for the DR group compared to the CR systems; MTF at 5 mm(-1) varied from 0.08 to 0.23 for the CR detectors against a range of 0.16-0.64 for the DR units. The needle CR detector had a higher MTF, lower NNPS and higher DQE at 5 mm(-1) than the powder CR phosphors. DQE at 5 mm(-1) ranged from 0.02 to 0.20 for the CR systems, while DQE at 5 mm(-1) for the DR group ranged from 0.04 to 0.41, indicating higher DQE for the DR detectors and needle CR system than for the powder CR phosphor systems. The technical evaluation section of the study showed that the digital mammography systems were well set up and exhibiting typical performance for the detector technology employed in the respective systems.

Monitoring of illicit pill distribution networks using an image collection exploration framework

This paper proposes a novel approach for the analysis of illicit tablets based on their visual characteristics. In particular, the paper concentrates on the problem of ecstasy pill seizure profiling and monitoring. The presented method extracts the visual information from pill images and builds a representation of it, i.e. it builds a pill profile based on the pill visual appearance. Different visual features are used to build different image similarity measures, which are the basis for a pill monitoring strategy based on both discriminative and clustering models. The discriminative model permits to infer whether two pills come from the same seizure, while the clustering models groups of pills that share similar visual characteristics. The resulting clustering structure allows to perform a visual identification of the relationships between different seizures. The proposed approach was evaluated using a data set of 621 Ecstasy pill pictures. The results demonstrate that this is a feasible and cost effective method for performing pill profiling and monitoring.

Management of patient dose and image noise in routine pediatric CT abdominal examinations.

The aim was to propose a strategy for finding reasonable compromises between image noise and dose as a function of patient weight. Weighted CT dose index (CTDI(w)) was measured on a multidetector-row CT unit using CTDI test objects of 16, 24 and 32 cm in diameter at 80, 100, 120 and 140 kV. These test objects were then scanned in helical mode using a wide range of tube currents and voltages with a reconstructed slice thickness of 5 mm. For each set of acquisition parameter image noise was measured and the Rose model observer was used to test two strategies for proposing a reasonable compromise between dose and low-contrast detection performance: (1) the use of a unique noise level for all test object diameters, and (2) the use of a unique dose efficacy level defined as the noise reduction per unit dose. Published data were used to define four weight classes and an acquisition protocol was proposed for each class. The protocols have been applied in clinical routine for more than one year. CTDI(vol) values of 6.7, 9.4, 15.9 and 24.5 mGy were proposed for the following weight classes: 2.5-5, 5-15, 15-30 and 30-50 kg with image noise levels in the range of 10-15 HU. The proposed method allows patient dose and image noise to be controlled in such a way that dose reduction does not impair the detection of low-contrast lesions. The proposed values correspond to high- quality images and can be reduced if only high-contrast organs are assessed.

Active learning methods for remote sensing image classification

In this paper, we propose two active learning algorithms for semiautomatic definition of training samples in remote sensing image classification. Based on predefined heuristics, the classifier ranks the unlabeled pixels and automatically chooses those that are considered the most valuable for its improvement. Once the pixels have been selected, the analyst labels them manually and the process is iterated. Starting with a small and nonoptimal training set, the model itself builds the optimal set of samples which minimizes the classification error. We have applied the proposed algorithms to a variety of remote sensing data, including very high resolution and hyperspectral images, using support vector machines. Experimental results confirm the consistency of the methods. The required number of training samples can be reduced to 10% using the methods proposed, reaching the same level of accuracy as larger data sets. A comparison with a state-of-the-art active learning method, margin sampling, is provided, highlighting advantages of the methods proposed. The effect of spatial resolution and separability of the classes on the quality of the selection of pixels is also discussed.

Bayesian image reconstruction with space-variant noise suppression

In this paper we present a Bayesian image reconstruction algorithm with entropy prior (FMAPE) that uses a space-variant hyperparameter. The spatial variation of the hyperparameter allows different degrees of resolution in areas of different statistical characteristics, thus avoiding the large residuals resulting from algorithms that use a constant hyperparameter. In the first implementation of the algorithm, we begin by segmenting a Maximum Likelihood Estimator (MLE) reconstruction. The segmentation method is based on using a wavelet decomposition and a self-organizing neural network. The result is a predetermined number of extended regions plus a small region for each star or bright object. To assign a different value of the hyperparameter to each extended region and star, we use either feasibility tests or cross-validation methods. Once the set of hyperparameters is obtained, we carried out the final Bayesian reconstruction, leading to a reconstruction with decreased bias and excellent visual characteristics. The method has been applied to data from the non-refurbished Hubble Space Telescope. The method can be also applied to ground-based images.

Exporting Contours to DICOM-RT Structure Set

This paper presents an ITK implementation for exportingthe contours of the automated segmentation results toDICOM-RT Structure Set format. The âeurooeradiotherapystructure setâeuro (RTSTRUCT) object of the DICOM standard isused for the transfer of patient structures and relateddata, between the devices found within and outside theradiotherapy department. It mainly contains theinformation of regions of interest (ROIs) and points ofinterest (E.g. dose reference points). In many cases,rather than manually drawing these ROIs on the CT images,one can indeed benefit from the automated segmentationalgorithms already implemented in ITK. But at present, itis not possible to export the ROIs obtained from ITK toRTSTRUCT format. In order to bridge this gap, we havedeveloped a framework for exporting contour data toRTSTRUCT. We provide here the complete implementation ofRTSTRUCT exporter and present the details of the pipelineused. Results on a 3-D CT image of the Head and Neck(H&N) region are presented.

Advanced Kernel Methods For Remote Sensing Image Classification

Résumé Suite aux recentes avancées technologiques, les archives d'images digitales ont connu une croissance qualitative et quantitative sans précédent. Malgré les énormes possibilités qu'elles offrent, ces avancées posent de nouvelles questions quant au traitement des masses de données saisies. Cette question est à la base de cette Thèse: les problèmes de traitement d'information digitale à très haute résolution spatiale et/ou spectrale y sont considérés en recourant à des approches d'apprentissage statistique, les méthodes à noyau. Cette Thèse étudie des problèmes de classification d'images, c'est à dire de catégorisation de pixels en un nombre réduit de classes refletant les propriétés spectrales et contextuelles des objets qu'elles représentent. L'accent est mis sur l'efficience des algorithmes, ainsi que sur leur simplicité, de manière à augmenter leur potentiel d'implementation pour les utilisateurs. De plus, le défi de cette Thèse est de rester proche des problèmes concrets des utilisateurs d'images satellite sans pour autant perdre de vue l'intéret des méthodes proposées pour le milieu du machine learning dont elles sont issues. En ce sens, ce travail joue la carte de la transdisciplinarité en maintenant un lien fort entre les deux sciences dans tous les développements proposés. Quatre modèles sont proposés: le premier répond au problème de la haute dimensionalité et de la redondance des données par un modèle optimisant les performances en classification en s'adaptant aux particularités de l'image. Ceci est rendu possible par un système de ranking des variables (les bandes) qui est optimisé en même temps que le modèle de base: ce faisant, seules les variables importantes pour résoudre le problème sont utilisées par le classifieur. Le manque d'information étiquétée et l'incertitude quant à sa pertinence pour le problème sont à la source des deux modèles suivants, basés respectivement sur l'apprentissage actif et les méthodes semi-supervisées: le premier permet d'améliorer la qualité d'un ensemble d'entraînement par interaction directe entre l'utilisateur et la machine, alors que le deuxième utilise les pixels non étiquetés pour améliorer la description des données disponibles et la robustesse du modèle. Enfin, le dernier modèle proposé considère la question plus théorique de la structure entre les outputs: l'intègration de cette source d'information, jusqu'à présent jamais considérée en télédétection, ouvre des nouveaux défis de recherche. Advanced kernel methods for remote sensing image classification Devis Tuia Institut de Géomatique et d'Analyse du Risque September 2009 Abstract The technical developments in recent years have brought the quantity and quality of digital information to an unprecedented level, as enormous archives of satellite images are available to the users. However, even if these advances open more and more possibilities in the use of digital imagery, they also rise several problems of storage and treatment. The latter is considered in this Thesis: the processing of very high spatial and spectral resolution images is treated with approaches based on data-driven algorithms relying on kernel methods. In particular, the problem of image classification, i.e. the categorization of the image's pixels into a reduced number of classes reflecting spectral and contextual properties, is studied through the different models presented. The accent is put on algorithmic efficiency and the simplicity of the approaches proposed, to avoid too complex models that would not be used by users. The major challenge of the Thesis is to remain close to concrete remote sensing problems, without losing the methodological interest from the machine learning viewpoint: in this sense, this work aims at building a bridge between the machine learning and remote sensing communities and all the models proposed have been developed keeping in mind the need for such a synergy. Four models are proposed: first, an adaptive model learning the relevant image features has been proposed to solve the problem of high dimensionality and collinearity of the image features. This model provides automatically an accurate classifier and a ranking of the relevance of the single features. The scarcity and unreliability of labeled. information were the common root of the second and third models proposed: when confronted to such problems, the user can either construct the labeled set iteratively by direct interaction with the machine or use the unlabeled data to increase robustness and quality of the description of data. Both solutions have been explored resulting into two methodological contributions, based respectively on active learning and semisupervised learning. Finally, the more theoretical issue of structured outputs has been considered in the last model, which, by integrating outputs similarity into a model, opens new challenges and opportunities for remote sensing image processing.

Image Analysis for Evaluating Air Void Parameters of Concrete, HR-396, 1998

This research project investigated the use of image analysis to measure the air void parameters of concrete specimens produced under standard laboratory conditions. The results obtained from the image analysis technique were compared to results obtained from plastic air content tests, Danish air meter tests (also referred to as Air Void Analyzer tests), high-pressure air content tests on hardened concrete, and linear traverse tests (as per ASTM C-457). Hardened concrete specimens were sent to three different laboratories for the linear traverse tests. The samples that were circulated to the three labs consisted of specimens that needed different levels of surface preparation. The first set consisted of approximately 18 specimens that had been sectioned from a 4 in. by 4 in. by 18 in. (10 cm by 10 cm by 46 cm) beam using a saw equipped with a diamond blade. These specimens were subjected to the normal sample preparation techniques that were commonly employed by the three different labs (each lab practiced slightly different specimen preparation techniques). The second set of samples consisted of eight specimens that had been ground and polished at a single laboratory. The companion labs were only supposed to retouch the sample surfaces if they exhibited major flaws. In general, the study indicated that the image analysis test results for entrained air content exhibited good to strong correlation to the average values determined via the linear traverse technique. Specimens ground and polished in a single laboratory and then circulated to the other participating laboratories for the air content determinations exhibited the strongest correlation between the image analysis and linear traverse techniques (coefficient of determination, r-squared = 0.96, for n=8). Specimens ground and polished at each of the individual laboratories exhibited considerably more scatter (coefficient of determination, r-squared = 0.78, for n=16). The image analysis technique tended to produce low estimates of the specific surface of the voids when compared to the results from the linear traverse method. This caused the image analysis spacing factor calculations to produce larger values than those obtained from the linear traverse tests. The image analysis spacing factors were still successful at distinguishing between the frost-prone test specimens and the other (more durable) test specimens that were studied in this research project.

Enhanced compressed sensing recovery with level set normals.

We propose a compressive sensing algorithm that exploits geometric properties of images to recover images of high quality from few measurements. The image reconstruction is done by iterating the two following steps: 1) estimation of normal vectors of the image level curves, and 2) reconstruction of an image fitting the normal vectors, the compressed sensing measurements, and the sparsity constraint. The proposed technique can naturally extend to nonlocal operators and graphs to exploit the repetitive nature of textured images to recover fine detail structures. In both cases, the problem is reduced to a series of convex minimization problems that can be efficiently solved with a combination of variable splitting and augmented Lagrangian methods, leading to fast and easy-to-code algorithms. Extended experiments show a clear improvement over related state-of-the-art algorithms in the quality of the reconstructed images and the robustness of the proposed method to noise, different kind of images, and reduced measurements.

Linear and nonlinear terrain deformation maps from a reduced set of interferometric SAR images

In this paper, an advanced technique for the generation of deformation maps using synthetic aperture radar (SAR) data is presented. The algorithm estimates the linear and nonlinear components of the displacement, the error of the digital elevation model (DEM) used to cancel the topographic terms, and the atmospheric artifacts from a reduced set of low spatial resolution interferograms. The pixel candidates are selected from those presenting a good coherence level in the whole set of interferograms and the resulting nonuniform mesh tessellated with the Delauney triangulation to establish connections among them. The linear component of movement and DEM error are estimated adjusting a linear model to the data only on the connections. Later on, this information, once unwrapped to retrieve the absolute values, is used to calculate the nonlinear component of movement and atmospheric artifacts with alternate filtering techniques in both the temporal and spatial domains. The method presents high flexibility with respect to the required number of images and the baselines length. However, better results are obtained with large datasets of short baseline interferograms. The technique has been tested with European Remote Sensing SAR data from an area of Catalonia (Spain) and validated with on-field precise leveling measurements.

Retinal Image Analysis Using Machine Vision

The topic of this thesis is studying how lesions in retina caused by diabetic retinopathy can be detected from color fundus images by using machine vision methods. Methods for equalizing uneven illumination in fundus images, detecting regions of poor image quality due toinadequate illumination, and recognizing abnormal lesions were developed duringthe work. The developed methods exploit mainly the color information and simpleshape features to detect lesions. In addition, a graphical tool for collecting lesion data was developed. The tool was used by an ophthalmologist who marked lesions in the images to help method development and evaluation. The tool is a general purpose one, and thus it is possible to reuse the tool in similar projects.The developed methods were tested with a separate test set of 128 color fundus images. From test results it was calculated how accurately methods classify abnormal funduses as abnormal (sensitivity) and healthy funduses as normal (specificity). The sensitivity values were 92% for hemorrhages, 73% for red small dots (microaneurysms and small hemorrhages), and 77% for exudates (hard and soft exudates). The specificity values were 75% for hemorrhages, 70% for red small dots, and 50% for exudates. Thus, the developed methods detected hemorrhages accurately and microaneurysms and exudates moderately.

Evaluation and comparison of current fetal ultrasound image segmentation methods for biometric measurements: a grand challenge.

This paper presents the evaluation results of the methods submitted to Challenge US: Biometric Measurements from Fetal Ultrasound Images, a segmentation challenge held at the IEEE International Symposium on Biomedical Imaging 2012. The challenge was set to compare and evaluate current fetal ultrasound image segmentation methods. It consisted of automatically segmenting fetal anatomical structures to measure standard obstetric biometric parameters, from 2D fetal ultrasound images taken on fetuses at different gestational ages (21 weeks, 28 weeks, and 33 weeks) and with varying image quality to reflect data encountered in real clinical environments. Four independent sub-challenges were proposed, according to the objects of interest measured in clinical practice: abdomen, head, femur, and whole fetus. Five teams participated in the head sub-challenge and two teams in the femur sub-challenge, including one team who tackled both. Nobody attempted the abdomen and whole fetus sub-challenges. The challenge goals were two-fold and the participants were asked to submit the segmentation results as well as the measurements derived from the segmented objects. Extensive quantitative (region-based, distance-based, and Bland-Altman measurements) and qualitative evaluation was performed to compare the results from a representative selection of current methods submitted to the challenge. Several experts (three for the head sub-challenge and two for the femur sub-challenge), with different degrees of expertise, manually delineated the objects of interest to define the ground truth used within the evaluation framework. For the head sub-challenge, several groups produced results that could be potentially used in clinical settings, with comparable performance to manual delineations. The femur sub-challenge had inferior performance to the head sub-challenge due to the fact that it is a harder segmentation problem and that the techniques presented relied more on the femur's appearance.

Unsupervised active regions for multiresolution image segmentation

An unsupervised approach to image segmentation which fuses region and boundary information is presented. The proposed approach takes advantage of the combined use of 3 different strategies: the guidance of seed placement, the control of decision criterion, and the boundary refinement. The new algorithm uses the boundary information to initialize a set of active regions which compete for the pixels in order to segment the whole image. The method is implemented on a multiresolution representation which ensures noise robustness as well as computation efficiency. The accuracy of the segmentation results has been proven through an objective comparative evaluation of the method

Image Segmentation Using Information Bottleneck Method

In image processing, segmentation algorithms constitute one of the main focuses of research. In this paper, new image segmentation algorithms based on a hard version of the information bottleneck method are presented. The objective of this method is to extract a compact representation of a variable, considered the input, with minimal loss of mutual information with respect to another variable, considered the output. First, we introduce a split-and-merge algorithm based on the definition of an information channel between a set of regions (input) of the image and the intensity histogram bins (output). From this channel, the maximization of the mutual information gain is used to optimize the image partitioning. Then, the merging process of the regions obtained in the previous phase is carried out by minimizing the loss of mutual information. From the inversion of the above channel, we also present a new histogram clustering algorithm based on the minimization of the mutual information loss, where now the input variable represents the histogram bins and the output is given by the set of regions obtained from the above split-and-merge algorithm. Finally, we introduce two new clustering algorithms which show how the information bottleneck method can be applied to the registration channel obtained when two multimodal images are correctly aligned. Different experiments on 2-D and 3-D images show the behavior of the proposed algorithms

Color ad spectral image assessment using novel quality and fidelity techniques

The ongoing development of the digital media has brought a new set of challenges with it. As images containing more than three wavelength bands, often called spectral images, are becoming a more integral part of everyday life, problems in the quality of the RGB reproduction from the spectral images have turned into an important area of research. The notion of image quality is often thought to comprise two distinctive areas – image quality itself and image fidelity, both dealing with similar questions, image quality being the degree of excellence of the image, and image fidelity the measure of the match of the image under study to the original. In this thesis, both image fidelity and image quality are considered, with an emphasis on the influence of color and spectral image features on both. There are very few works dedicated to the quality and fidelity of spectral images. Several novel image fidelity measures were developed in this study, which include kernel similarity measures and 3D-SSIM (structural similarity index). The kernel measures incorporate the polynomial, Gaussian radial basis function (RBF) and sigmoid kernels. The 3D-SSIM is an extension of a traditional gray-scale SSIM measure developed to incorporate spectral data. The novel image quality model presented in this study is based on the assumption that the statistical parameters of the spectra of an image influence the overall appearance. The spectral image quality model comprises three parameters of quality: colorfulness, vividness and naturalness. The quality prediction is done by modeling the preference function expressed in JNDs (just noticeable difference). Both image fidelity measures and the image quality model have proven to be effective in the respective experiments.