Anamorphic pixels for multi-channel superresolution

Superresolution from plenoptic cameras or camera arrays is usually treated similarly to superresolution from video streams. However, the transformation between the low-resolution views can be determined precisely from camera geometry and parallax. Furthermore, as each low-resolution image originates from a unique physical camera, its sampling properties can also be unique. We exploit this option with a custom design of either the optics or the sensor pixels. This design makes sure that the sampling matrix of the complete system is always well-formed, enabling robust and high-resolution image reconstruction. We show that simply changing the pixel aspect ratio from square to anamorphic is sufficient to achieve that goal, as long as each camera has a unique aspect ratio. We support this claim with theoretical analysis and image reconstruction of real images. We derive the optimal aspect ratios for sets of 2 or 4 cameras. Finally, we verify our solution with a camera system using an anamorphic lens.

Application of a new laser Doppler imaging system in planning and monitoring of surgical flaps

There is a demand for technologies able to assess the perfusion of surgical flaps quantitatively and reliably to avoid ischemic complications. The aim of this study is to test a new high-speed high-definition laser Doppler imaging (LDI) system (FluxEXPLORER, Microvascular Imaging, Lausanne, Switzerland) in terms of preoperative mapping of the vascular supply (perforator vessels) and postoperative flow monitoring. The FluxEXPLORER performs perfusion mapping of an area 9 x 9 cm with a resolution of 256 x 256 pixels within 6 s in high-definition imaging mode. The sensitivity and predictability to localize perforators is expressed by the coincidence of preoperatively assessed LDI high flow spots with intraoperatively verified perforators in nine patients. 18 free flaps are monitored before, during, and after total ischemia. 63% of all verified perforators correspond to a high flow spot, and 38% of all high flow spots correspond to a verified perforator (positive predictive value). All perfused flaps reveal a value of above 221 perfusion units (PUs), and all values obtained in the ischemic flaps are beneath 187 PU. In summary, we conclude that the present LDI system can serve as a reliable, fast, and easy-to-handle tool to detect ischemia in free flaps, whereas perforator vessels cannot be detected appropriately.

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.

Lipid-poor adenomas on unenhanced CT: does histogram analysis increase sensitivity compared with a mean attenuation threshold?

OBJECTIVE: The purpose of our study was to evaluate the efficacy of CT histogram analysis for further characterization of lipid-poor adenomas on unenhanced CT. MATERIALS AND METHODS: One hundred thirty-two adrenal nodules were identified in 104 patients with lung cancer who underwent PET/CT. Sixty-five nodules were classified as lipid-rich adenomas if they had an unenhanced CT attenuation of less than or equal to 10 H. Thirty-one masses were classified as lipid-poor adenomas if they had an unenhanced CT attenuation greater than 10 H and stability for more than 1 year. Thirty-six masses were classified as lung cancer metastases if they showed rapid growth in 1 year (n = 27) or were biopsy-proven (n = 9). Histogram analysis was performed for all lesions to provide the mean attenuation value and percentage of negative pixels. RESULTS: All lipid-rich adenomas had more than 10% negative pixels; 51.6% of lipid-poor adenomas had more than 10% negative pixels and would have been classified as indeterminate nodules on the basis of mean attenuation alone. None of the metastases had more than 10% negative pixels. Using an unenhanced CT mean attenuation threshold of less than 10 H yielded a sensitivity of 68% and specificity of 100% for the diagnosis of an adenoma. Using an unenhanced CT threshold of more than 10% negative pixels yielded a sensitivity of 84% and specificity of 100% for the diagnosis of an adenoma. CONCLUSION: CT histogram analysis is superior to mean CT attenuation analysis for the evaluation of adrenal nodules and may help decrease referrals for additional imaging or biopsy.

Finite element 3D reconstruction of the pulmonary acinus imaged by synchrotron X-ray tomography

The alveolated structure of the pulmonary acinus plays a vital role in gas exchange function. Three-dimensional (3D) analysis of the parenchymal region is fundamental to understanding this structure-function relationship, but only a limited number of attempts have been conducted in the past because of technical limitations. In this study, we developed a new image processing methodology based on finite element (FE) analysis for accurate 3D structural reconstruction of the gas exchange regions of the lung. Stereologically well characterized rat lung samples (Pediatr Res 53: 72-80, 2003) were imaged using high-resolution synchrotron radiation-based X-ray tomographic microscopy. A stack of 1,024 images (each slice: 1024 x 1024 pixels) with resolution of 1.4 mum(3) per voxel were generated. For the development of FE algorithm, regions of interest (ROI), containing approximately 7.5 million voxels, were further extracted as a working subunit. 3D FEs were created overlaying the voxel map using a grid-based hexahedral algorithm. A proper threshold value for appropriate segmentation was iteratively determined to match the calculated volume density of tissue to the stereologically determined value (Pediatr Res 53: 72-80, 2003). The resulting 3D FEs are ready to be used for 3D structural analysis as well as for subsequent FE computational analyses like fluid dynamics and skeletonization.

Differential optical densities of intraretinal spaces

PURPOSE: To test the hypothesis that hyporeflective spaces in the neuroretina found on optical coherence tomography (OCT) examination have different optical reflectivities according to whether they are associated with exudation or degeneration. METHODS: Retrospective analysis of eyes with idiopathic perifoveal telangiectasia (IPT), diabetic macular edema (DME), idiopathic central serous chorioretinopathy (CSC), retinitis pigmentosa (RP), or cone dystrophy (CD) and eyes of healthy control subjects. OCT scans were performed. Raw scan data were exported and used to calculate light reflectivity profiles. Reflectivity data were acquired by projecting three rectangular boxes, each 50 pixels long and 5 pixels wide, into the intraretinal cystoid spaces, centrally onto unaffected peripheral RPE, and onto the prefoveolar vitreous. Light reflectivity in the retinal pigment epithelium (RPE), vitreous, and intraretinal spaces for the different retinal conditions and control subjects were compared. RESULTS: Reflectivities of the vitreous and the RPE were similar among the groups. Hyporeflective spaces in eyes with exudation (DME, RP, and CSC) had higher reflectivity compared with the mean reflectivity of the vitreous, whereas the cystoid spaces in the maculae of the eyes without exudation (CD and IPT) had a lower reflectivity than did the normal vitreous. CONCLUSIONS: Analysis of the light reflectivity profiles may be a tool to determine whether the density of hyporeflective spaces in the macula is greater or less than that of the vitreous, and may be a way to differentiate degenerative from exudative macular disease.

High-density electrode array for imaging in vitro electrophysiological activity

The development of a high-density active microelectrode array for in vitro electrophysiology is reported. Based on the Active Pixel Sensor (APS) concept, the array integrates 4096 gold microelectrodes (electrode separation 20 microm) on a surface of 2.5 mmx2.5 mm as well as a high-speed random addressing logic allowing the sequential selection of the measuring pixels. Following the electrical characterization in a phosphate solution, the functional evaluation has been carried out by recording the spontaneous electrical activity of neonatal rat cardiomyocytes. Signals with amplitudes from 130 microVp-p to 300 microVp-p could be recorded from different pixels. The results demonstrate the suitability of the APS concept for developing a new generation of high-resolution extracellular recording devices for in vitro electrophysiology.

IMRT credentialing for prospective trials using institutional virtual phantoms: results of a joint European Organization for the Research and Treatment of Cancer and Radiological Physics Center project.

BACKGROUND AND PURPOSE Intensity-modulated radiotherapy (IMRT) credentialing for a EORTC study was performed using an anthropomorphic head phantom from the Radiological Physics Center (RPC; RPC(PH)). Institutions were retrospectively requested to irradiate their institutional phantom (INST(PH)) using the same treatment plan in the framework of a Virtual Phantom Project (VPP) for IMRT credentialing. MATERIALS AND METHODS CT data set of the institutional phantom and measured 2D dose matrices were requested from centers and sent to a dedicated secure EORTC uploader. Data from the RPC(PH) and INST(PH) were thereafter centrally analyzed and inter-compared by the QA team using commercially available software (RIT; ver.5.2; Colorado Springs, USA). RESULTS Eighteen institutions participated to the VPP. The measurements of 6 (33%) institutions could not be analyzed centrally. All other centers passed both the VPP and the RPC ±7%/4 mm credentialing criteria. At the 5%/5 mm gamma criteria (90% of pixels passing), 11(92%) as compared to 12 (100%) centers pass the credentialing process with RPC(PH) and INST(PH) (p = 0.29), respectively. The corresponding pass rate for the 3%/3 mm gamma criteria (90% of pixels passing) was 2 (17%) and 9 (75%; p = 0.01), respectively. CONCLUSIONS IMRT dosimetry gamma evaluations in a single plane for a H&N prospective trial using the INST(PH) measurements showed agreement at the gamma index criteria of ±5%/5 mm (90% of pixels passing) for a small number of VPP measurements. Using more stringent, criteria, the RPC(PH) and INST(PH) comparison showed disagreement. More data is warranted and urgently required within the framework of prospective studies.

Coded Aperture Flow

Real cameras have a limited depth of field. The resulting defocus blur is a valuable cue for estimating the depth structure of a scene. Using coded apertures, depth can be estimated from a single frame. For optical flow estimation between frames, however, the depth dependent degradation can introduce errors. These errors are most prominent when objects move relative to the focal plane of the camera. We incorporate coded aperture defocus blur into optical flow estimation and allow for piecewise smooth 3D motion of objects. With coded aperture flow, we can establish dense correspondences between pixels in succeeding coded aperture frames. We compare several approaches to compute accurate correspondences for coded aperture images showing objects with arbitrary 3D motion.

Localization and Segmentation of 3D Intervertebral Discs in MR Images by Data Driven Estimation

This paper addresses the problem of fully-automatic localization and segmentation of 3D intervertebral discs (IVDs) from MR images. Our method contains two steps, where we first localize the center of each IVD, and then segment IVDs by classifying image pixels around each disc center as foreground (disc) or background. The disc localization is done by estimating the image displacements from a set of randomly sampled 3D image patches to the disc center. The image displacements are estimated by jointly optimizing the training and test displacement values in a data-driven way, where we take into consideration both the training data and the geometric constraint on the test image. After the disc centers are localized, we segment the discs by classifying image pixels around disc centers as background or foreground. The classification is done in a similar data-driven approach as we used for localization, but in this segmentation case we are aiming to estimate the foreground/background probability of each pixel instead of the image displacements. In addition, an extra neighborhood smooth constraint is introduced to enforce the local smoothness of the label field. Our method is validated on 3D T2-weighted turbo spin echo MR images of 35 patients from two different studies. Experiments show that compared to state of the art, our method achieves better or comparable results. Specifically, we achieve for localization a mean error of 1.6-2.0 mm, and for segmentation a mean Dice metric of 85%-88% and a mean surface distance of 1.3-1.4 mm.

Shaping the PSF to nearly top-hat profile: CHEOPS laboratory results

Spreading the PSF over a quite large amount of pixels is an increasingly used observing technique in order to reach extremely precise photometry, such as in the case of exoplanets searching and characterization via transits observations. A PSF top-hat profile helps to minimize the errors contribution due to the uncertainty on the knowledge of the detector flat field. This work has been carried out during the recent design study in the framework of the ESA small mission CHEOPS. Because of lack of perfect flat-fielding information, in the CHEOPS optics it is required to spread the light of a source into a well defined angular area, in a manner as uniform as possible. Furthermore this should be accomplished still retaining the features of a true focal plane onto the detector. In this way, for instance, the angular displacement on the focal plane is fully retained and in case of several stars in a field these look as separated as their distance is larger than the spreading size. An obvious way is to apply a defocus, while the presence of an intermediate pupil plane in the Back End Optics makes attractive to introduce here an optical device that is able to spread the light in a well defined manner, still retaining the direction of the chief ray hitting it. This can be accomplished through an holographic diffuser or through a lenslet array. Both techniques implement the concept of segmenting the pupil into several sub-zones where light is spread to a well defined angle. We present experimental results on how to deliver such PSF profile by mean of holographic diffuser and lenslet array. Both the devices are located in an intermediate pupil plane of a properly scaled laboratory setup mimicking the CHEOPS optical design configuration. © (2014) COPYRIGHT Society of Photo-Optical Instrumentation Engineers (SPIE). Downloading of the abstract is permitted for personal use only.

RETICULAR PSEUDODRUSEN ON INFRARED IMAGING ARE TOPOGRAPHICALLY DISTINCT FROM SUBRETINAL DRUSENOID DEPOSITS ON EN FACE OPTICAL COHERENCE TOMOGRAPHY.

PURPOSE: To evaluate the quantitative and topographic relationship between reticular pseudodrusen (RPD) on infrared reflectance (IR) and subretinal drusenoid deposits (SDD) on en face volumetric spectral domain optical coherence tomography. METHODS: Reticular pseudodrusen were marked on IR images by a masked observer. Subretinal drusenoid deposits were visualized on en face sections of spectral domain optical coherence tomography below the external limiting membrane and identified by a semiautomated technique. Control RPD lesions were generated in a random distribution for each IR image. Binary maps of control and experimental RPD and SDD were merged and analyzed in terms of topographic localization and quantitative drusen load comparison. RESULTS: A total of 54 eyes of 41 patients diagnosed with RPD were included in this study. The average number of RPD lesions on IR images was 320 ± 44.62 compared with 127 ± 26.02 SDD lesions on en face (P < 0.001). The majority of RPD lesions did not overlap with SDD lesions and were located >30 μm away (92%). The percentage of total SDD lesions overlapping RPD was 2.91 ± 0.87% compared with 1.73 ± 0.68% overlapping control RPD lesions (P < 0.05). The percentage of total SDD lesions between 1 and 3 pixels of the nearest RPD lesion was 5.08 ± 1.40% compared with 3.33 ± 1.07% between 1 and 3 pixels of the nearest control RPD lesion (P < 0.05). CONCLUSION: This study identified significantly more RPD lesions on IR compared with SDD lesions on en face spectral domain optical coherence tomography and found that a large majority of SDD (>90% of lesions) were >30 μm away from the nearest RPD. Together, our findings indicate that RPD and SDD are two entities that are only occasionally topographically associated, suggesting that at some stage in their development, they may be pathologically related.

Temporally Coherent Disparity Maps Using CRFs with Fast 4D Filtering

State of the art methods for disparity estimation achieve good results for single stereo frames, but temporal coherence in stereo videos is often neglected. In this paper we present a method to compute temporally coherent disparity maps. We define an energy over whole stereo sequences and optimize their Conditional Random Field (CRF) distributions using mean-field approximation. We introduce novel terms for smoothness and consistency between the left and right views, and perform CRF optimization by fast, iterative spatio-temporal filtering with linear complexity in the total number of pixels. Our results rank among the state of the art while having significantly less flickering artifacts in stereo sequences.

Regularizing Image Reconstruction for Gradient-Domain Rendering with Feature Patches

We present a novel algorithm to reconstruct high-quality images from sampled pixels and gradients in gradient-domain rendering. Our approach extends screened Poisson reconstruction by adding additional regularization constraints. Our key idea is to exploit local patches in feature images, which contain per-pixels normals, textures, position, etc., to formulate these constraints. We describe a GPU implementation of our approach that runs on the order of seconds on megapixel images. We demonstrate a significant improvement in image quality over screened Poisson reconstruction under the L1 norm. Because we adapt the regularization constraints to the noise level in the input, our algorithm is consistent and converges to the ground truth.