Matrix Designs and Methods for Secure and Efficient Compressed Sensing

The idea of balancing the resources spent in the acquisition and encoding of natural signals strictly to their intrinsic information content has interested nearly a decade of research under the name of compressed sensing. In this doctoral dissertation we develop some extensions and improvements upon this technique's foundations, by modifying the random sensing matrices on which the signals of interest are projected to achieve different objectives. Firstly, we propose two methods for the adaptation of sensing matrix ensembles to the second-order moments of natural signals. These techniques leverage the maximisation of different proxies for the quantity of information acquired by compressed sensing, and are efficiently applied in the encoding of electrocardiographic tracks with minimum-complexity digital hardware. Secondly, we focus on the possibility of using compressed sensing as a method to provide a partial, yet cryptanalysis-resistant form of encryption; in this context, we show how a random matrix generation strategy with a controlled amount of perturbations can be used to distinguish between multiple user classes with different quality of access to the encrypted information content. Finally, we explore the application of compressed sensing in the design of a multispectral imager, by implementing an optical scheme that entails a coded aperture array and Fabry-Pérot spectral filters. The signal recoveries obtained by processing real-world measurements show promising results, that leave room for an improvement of the sensing matrix calibration problem in the devised imager.

A new information fusion method for land-use classification using high resolution satellite imagery

Satellite image classification involves designing and developing efficient image classifiers. With satellite image data and image analysis methods multiplying rapidly, selecting the right mix of data sources and data analysis approaches has become critical to the generation of quality land-use maps. In this study, a new postprocessing information fusion algorithm for the extraction and representation of land-use information based on high-resolution satellite imagery is presented. This approach can produce land-use maps with sharp interregional boundaries and homogeneous regions. The proposed approach is conducted in five steps. First, a GIS layer - ATKIS data - was used to generate two coarse homogeneous regions, i.e. urban and rural areas. Second, a thematic (class) map was generated by use of a hybrid spectral classifier combining Gaussian Maximum Likelihood algorithm (GML) and ISODATA classifier. Third, a probabilistic relaxation algorithm was performed on the thematic map, resulting in a smoothed thematic map. Fourth, edge detection and edge thinning techniques were used to generate a contour map with pixel-width interclass boundaries. Fifth, the contour map was superimposed on the thematic map by use of a region-growing algorithm with the contour map and the smoothed thematic map as two constraints. For the operation of the proposed method, a software package is developed using programming language C. This software package comprises the GML algorithm, a probabilistic relaxation algorithm, TBL edge detector, an edge thresholding algorithm, a fast parallel thinning algorithm, and a region-growing information fusion algorithm. The county of Landau of the State Rheinland-Pfalz, Germany was selected as a test site. The high-resolution IRS-1C imagery was used as the principal input data.

Detection and identification of 700 drugs by multi-target screening with a 3200 Q TRAP LC-MS/MS system and library searching

The multi-target screening method described in this work allows the simultaneous detection and identification of 700 drugs and metabolites in biological fluids using a hybrid triple-quadrupole linear ion trap mass spectrometer in a single analytical run. After standardization of the method, the retention times of 700 compounds were determined and transitions for each compound were selected by a "scheduled" survey MRM scan, followed by an information-dependent acquisition using the sensitive enhanced product ion scan of a Q TRAP hybrid instrument. The identification of the compounds in the samples analyzed was accomplished by searching the tandem mass spectrometry (MS/MS) spectra against the library we developed, which contains electrospray ionization-MS/MS spectra of over 1,250 compounds. The multi-target screening method together with the library was included in a software program for routine screening and quantitation to achieve automated acquisition and library searching. With the help of this software application, the time for evaluation and interpretation of the results could be drastically reduced. This new multi-target screening method has been successfully applied for the analysis of postmortem and traffic offense samples as well as proficiency testing, and complements screening with immunoassays, gas chromatography-mass spectrometry, and liquid chromatography-diode-array detection. Other possible applications are analysis in clinical toxicology (for intoxication cases), in psychiatry (antidepressants and other psychoactive drugs), and in forensic toxicology (drugs and driving, workplace drug testing, oral fluid analysis, drug-facilitated sexual assault).

Real-Time Multimodal Retinal Image Registration for a Computer Assisted Laser Photocoagulation System

An algorithm for the real-time registration of a retinal video sequence captured with a scanning digital ophthalmoscope (SDO) to a retinal composite image is presented. This method is designed for a computer-assisted retinal laser photocoagulation system to compensate for retinal motion and hence enhance the accuracy, speed, and patient safety of retinal laser treatments. The procedure combines intensity and feature-based registration techniques. For the registration of an individual frame, the translational frame-to-frame motion between preceding and current frame is detected by normalized cross correlation. Next, vessel points on the current video frame are identified and an initial transformation estimate is constructed from the calculated translation vector and the quadratic registration matrix of the previous frame. The vessel points are then iteratively matched to the segmented vessel centerline of the composite image to refine the initial transformation and register the video frame to the composite image. Criteria for image quality and algorithm convergence are introduced, which assess the exclusion of single frames from the registration process and enable a loss of tracking signal if necessary. The algorithm was successfully applied to ten different video sequences recorded from patients. It revealed an average accuracy of 2.47 ± 2.0 pixels (∼23.2 ± 18.8 μm) for 2764 evaluated video frames and demonstrated that it meets the clinical requirements.

Fully Automatic Segmentation of Brain Tumor Images using Support Vector Machine Classiffication in Combination with Hierarchical Conditional Random Field Regularization

Delineating brain tumor boundaries from magnetic resonance images is an essential task for the analysis of brain cancer. We propose a fully automatic method for brain tissue segmentation, which combines Support Vector Machine classification using multispectral intensities and textures with subsequent hierarchical regularization based on Conditional Random Fields. The CRF regularization introduces spatial constraints to the powerful SVM classification, which assumes voxels to be independent from their neighbors. The approach first separates healthy and tumor tissue before both regions are subclassified into cerebrospinal fluid, white matter, gray matter and necrotic, active, edema region respectively in a novel hierarchical way. The hierarchical approach adds robustness and speed by allowing to apply different levels of regularization at different stages. The method is fast and tailored to standard clinical acquisition protocols. It was assessed on 10 multispectral patient datasets with results outperforming previous methods in terms of segmentation detail and computation times.

Geometry-Aware Multiscale Image Registration via OBBTree-Based Polyaffine Log-Demons

Non-linear image registration is an important tool in many areas of image analysis. For instance, in morphometric studies of a population of brains, free-form deformations between images are analyzed to describe the structural anatomical variability. Such a simple deformation model is justified by the absence of an easy expressible prior about the shape changes. Applying the same algorithms used in brain imaging to orthopedic images might not be optimal due to the difference in the underlying prior on the inter-subject deformations. In particular, using an un-informed deformation prior often leads to local minima far from the expected solution. To improve robustness and promote anatomically meaningful deformations, we propose a locally affine and geometry-aware registration algorithm that automatically adapts to the data. We build upon the log-domain demons algorithm and introduce a new type of OBBTree-based regularization in the registration with a natural multiscale structure. The regularization model is composed of a hierarchy of locally affine transformations via their logarithms. Experiments on mandibles show improved accuracy and robustness when used to initialize the demons, and even similar performance by direct comparison to the demons, with a significantly lower degree of freedom. This closes the gap between polyaffine and non-rigid registration and opens new ways to statistically analyze the registration results.

Delayed gadolinium-enhanced magnetic resonance imaging of hip joint cartilage: pearls and pitfalls

With the increasing advances in hip joint preservation surgery, accurate diagnosis and assessment of femoral head and acetabular cartilage status is becoming increasingly important. Magnetic resonance imaging (MRI) of the hip does present technical difficulties. The fairly thin cartilage lining necessitates high image resolution and high contrast-to-noise ratio (CNR). With MR arthrography (MRA) using intraarticular injected gadolinium, labral tears and cartilage clefts may be better identified through the contrast medium filling into the clefts. However, the ability of MRA to detect varying grades of cartilage damage is fairly limited and early histological and biochemical changes in the beginning of osteoarthritis (OA) cannot be accurately delineated. Traditional MRI thus lacks the ability to analyze the biological status of cartilage degeneration. The technique of delayed gadolinium-enhanced MRI of cartilage (dGEMRIC) is sensitive to the charge density of cartilage contributed by glycosaminoglycans (GAGs), which are lost early in the process of OA. Therefore, the dGEMRIC technique has a potential to detect early cartilage damage that is obviously critical for decision-making regarding time and extent of intervention for joint-preservation. In the last decade, cartilage imaging with dGEMRIC has been established as an accurate and reliable tool for assessment of cartilage status in the knee and hip joint.This review outlines the current status of dGEMRIC for assessment of hip joint cartilage. Practical modifications of the standard technique including three-dimensional (3D) dGEMRIC and dGEMRIC after intra-articular gadolinium instead of iv-dGEMRIC will also be addressed.

Steerable intravitreal inserts for drug delivery: in vitro and ex vivo mobility experiments

The progress of wet age-related macular degeneration can now be controlled by intravitreal drug injection. This approach requires repeated injections, which could be avoided by delivering the drug to the retina. Intraocular implants are a promising solution for drug delivery near the retina. Currently, their accurate placement is challenging, and they can only be removed after a vitrectomy. In this paper, we introduce an approach for minimally invasive retinal drug delivery using magnetic intraocular inserts. We briefly discuss the electromagnetic-control system for magnetic implants and then focus on evaluating their ability to move in the vitreous humor. The mobility of magnetic intraocular implants is estimated in vitro with synthesized vitreous humors, and ex vivo with experiments on cadaver porcine eyes. Preliminary results show that with such magnetic implants a vitrectomy can be avoided.

Population-Based Design of Mandibular Plates Based on Bone Quality and Morphology

In this paper we present a new population-based implant design methodology, which advances the state-of-the-art approaches by combining shape and bone quality information into the design strategy. The method enhances the mechanical stability of the fixation and reduces the intra-operative in-plane bending which might impede the functionality of the locking mechanism. The method is presented for the case of mandibular locking fixation plates, where the mandibular angle and the bone quality at screw locations are taken into account. Using computational anatomy techniques, the method automatically derives, from a set of computed tomography images, the mandibular angle and the bone thickness and intensity values at the path of every screw. An optimisation strategy is then used to optimise the two parameters of plate angle and screw position. Results for the new design are presented along with a comparison with a commercially available mandibular locking fixation plate. A statistically highly significant improvement was observed. Our experiments allowed us to conclude that an angle of 126° and a screw separation of 8mm is a more suitable design than the standard 120° and 9mm.

How reliable and safe is full-body low-dose radiography (LODOX Statscan) in detecting foreign bodies ingested by adults?

OBJECTIVE: Foreign body ingestion is common and potentially lethal. This study evaluates the use of low-dose Statscans (LODOX) in emergency departments. DESIGN: This comparative cross-sectional study retrospectively assessed 28 289 digital chest x-rays and 2301 LODOX scans performed between 2006 and 2010 at a tertiary emergency centre. The radiographic appearance, image quality and location of ingested foreign bodies were evaluated in standard digital chest and LODOX radiography. The mean irradiation (μSv) and cumulative mean radiation dose per patient with the ingested foreign body were calculated according to literature-based data, together with the sensitivity and specificity for each modality. RESULTS: A total of 62 foreign bodies were detected in 39 patients, of whom 19 were investigated with LODOX and 20 with conventional digital chest radiography. Thirty-three foreign bodies were located in the two upper abdominal quadrants, 21 in the lower quadrants-which are not visible on conventional digital chest radiography-seven in the oesophagus and one in the bronchial system. The sensitivity and specificity of digital chest radiography were 44.4% and 94.1%, respectively, and for the LODOX Statscan 90% and 100%, respectively. The calculated mean radiation dose for LODOX investigations was 184 μS, compared with 524 μS for digital chest radiography. CONCLUSIONS: LODOX Statscan is superior to digital chest radiography in the diagnostic work-up of ingested foreign bodies because it makes it possible to enlarge the field of view to the entire body, has higher sensitivity and specificity, and reduces the radiation dose by 65%.

Pathology hinting as the combination of automatic segmentation with a statistical shape model

With improvements in acquisition speed and quality, the amount of medical image data to be screened by clinicians is starting to become challenging in the daily clinical practice. To quickly visualize and find abnormalities in medical images, we propose a new method combining segmentation algorithms with statistical shape models. A statistical shape model built from a healthy population will have a close fit in healthy regions. The model will however not fit to morphological abnormalities often present in the areas of pathologies. Using the residual fitting error of the statistical shape model, pathologies can be visualized very quickly. This idea is applied to finding drusen in the retinal pigment epithelium (RPE) of optical coherence tomography (OCT) volumes. A segmentation technique able to accurately segment drusen in patients with age-related macular degeneration (AMD) is applied. The segmentation is then analyzed with a statistical shape model to visualize potentially pathological areas. An extensive evaluation is performed to validate the segmentation algorithm, as well as the quality and sensitivity of the hinting system. Most of the drusen with a height of 85.5 microm were detected, and all drusen at least 93.6 microm high were detected.

Image quality in low-dose coronary computed tomography angiography with a new high-definition CT scanner

A new generation of high definition computed tomography (HDCT) 64-slice devices complemented by a new iterative image reconstruction algorithm-adaptive statistical iterative reconstruction, offer substantially higher resolution compared to standard definition CT (SDCT) scanners. As high resolution confers higher noise we have compared image quality and radiation dose of coronary computed tomography angiography (CCTA) from HDCT versus SDCT. Consecutive patients (n = 93) underwent HDCT, and were compared to 93 patients who had previously undergone CCTA with SDCT matched for heart rate (HR), HR variability and body mass index (BMI). Tube voltage and current were adapted to the patient's BMI, using identical protocols in both groups. The image quality of all CCTA scans was evaluated by two independent readers in all coronary segments using a 4-point scale (1, excellent image quality; 2, blurring of the vessel wall; 3, image with artefacts but evaluative; 4, non-evaluative). Effective radiation dose was calculated from DLP multiplied by a conversion factor (0.014 mSv/mGy × cm). The mean image quality score from HDCT versus SDCT was comparable (2.02 ± 0.68 vs. 2.00 ± 0.76). Mean effective radiation dose did not significantly differ between HDCT (1.7 ± 0.6 mSv, range 1.0-3.7 mSv) and SDCT (1.9 ± 0.8 mSv, range 0.8-5.5 mSv; P = n.s.). HDCT scanners allow low-dose 64-slice CCTA scanning with higher resolution than SDCT but maintained image quality and equally low radiation dose. Whether this will translate into higher accuracy of HDCT for CAD detection remains to be evaluated.

Greek Life and God Attachment: the Interaction of Sorority Influence and Religiosity on Women's Body Shame

The present study investigated the relationships between sorority women’s internalization of Greek thin ideals and body image, and dimensions of sorority women’s religiosity and body image. A combined relationship among sorority women’s internalization of Greek thin ideals, body image, and religiosity was also examined. Based on previous research it was expected that women’s internalization of Greek thin ideals would be associated with worse body image (in terms of body shame, body esteem, and drive for thinness) and that women’s religiosity (in terms of secure attachment to God) would be associated with better body image. Combinations of Greek thin ideal internalization and God attachment were expected to significantly predict changes in women’s body image. Women completed a series of survey measures assessing their awareness and internalization of Greek sociocultural thin ideals and their sense of community within their particular sorority. Women also completed a series of survey measures assessing their body shame, body esteem, and drive for thinness, in addition to survey measures assessing dimensions of their religiosity. The study’s findings revealed that women’s internalization of Greek thin ideals was associated with worse body image outcomes and that anxious attachment to God was associated with worse body image outcomes, particularly in relation to body shame. Moderation analyses revealed that Greek thin ideal internalization significantly interacted with anxious God attachment to predict body shame.

An introduction to low-level analysis methods of DNA microarray data

This article gives an overview over the methods used in the low--level analysis of gene expression data generated using DNA microarrays. This type of experiment allows to determine relative levels of nucleic acid abundance in a set of tissues or cell populations for thousands of transcripts or loci simultaneously. Careful statistical design and analysis are essential to improve the efficiency and reliability of microarray experiments throughout the data acquisition and analysis process. This includes the design of probes, the experimental design, the image analysis of microarray scanned images, the normalization of fluorescence intensities, the assessment of the quality of microarray data and incorporation of quality information in subsequent analyses, the combination of information across arrays and across sets of experiments, the discovery and recognition of patterns in expression at the single gene and multiple gene levels, and the assessment of significance of these findings, considering the fact that there is a lot of noise and thus random features in the data. For all of these components, access to a flexible and efficient statistical computing environment is an essential aspect.

Correlation between the area of increased autofluorescence surrounding geographic atrophy and disease progression in patients with AMD

PURPOSE: To test the hypothesis that the extension of areas with increased fundus autofluorescence (FAF) outside atrophic patches correlates with the rate of spread of geographic atrophy (GA) over time in eyes with age-related macular degeneration (AMD). METHODS: The database of the multicenter longitudinal natural history Fundus Autofluorescence in AMD (FAM) Study was reviewed for patients with GA recruited through the end of August 2003, with follow-up examinations within at least 1 year. Only eyes with sufficient image quality and with diffuse patterns of increased FAF surrounding atrophy were chosen. In standardized digital FAF images (excitation, 488 nm; emission, >500 nm), total size and spread of GA was measured. The convex hull (CH) of increased FAF as the minimum polygon encompassing the entire area of increased FAF surrounding the central atrophic patches was quantified at baseline. Statistical analysis was performed with the Spearman's rank correlation coefficient (rho). RESULTS: Thirty-nine eyes of 32 patients were included (median age, 75.0 years; interquartile range [IQR], 67.8-78.9); median follow-up, 1.87 years; IQR, 1.43-3.37). At baseline, the median total size of atrophy was 7.04 mm2 (IQR, 4.20-9.88). The median size of the CH was 21.47 mm2 (IQR, 15.19-28.26). The median rate of GA progression was 1.72 mm2 per year (IQR, 1.10-2.83). The area of increased FAF around the atrophy (difference between the CH and the total GA size at baseline) showed a positive correlation with GA enlargement over time (rho=0.60; P=0.0002). CONCLUSIONS: FAF characteristics that are not identified by fundus photography or fluorescein angiography may serve as a prognostic determinant in advanced atrophic AMD. As the FAF signal originates from lipofuscin (LF) in postmitotic RPE cells and since increased FAF indicates excessive LF accumulation, these findings would underscore the pathophysiological role of RPE-LF in AMD pathogenesis.