25 results
Search Results
Now showing 1 - 10 of 25
Item Introducing CNN-Based Mouse Grim Scale Analysis for Fully Automated Image-Based Assessment of Distress in Laboratory Mice(The Eurographics Association, 2018) Kopaczka, Marcin; Ernst, Lisa; Schock, Justus; Schneuing, Arne; Guth, Alexander; Tolba, Rene; Merhof, Dorit; Puig Puig, Anna and Schultz, Thomas and Vilanova, Anna and Hotz, Ingrid and Kozlikova, Barbora and Vázquez, Pere-PauInternational standards require close monitoring of distress of animals undergoing laboratory experiments in order to minimize the stress level and allow choosing minimally stressful procedures for the experiments. Currently, one of the the best established severity assessment procedures is the mouse grimace scale (MGS), a protocol in which images of the animals are taken and scored by assessing five key visual features that have been shown to be highly correlated with distress and pain. While proven to be highly reliable, MGS assessment is currently a time-consuming task requiring manual video processing for key frame extraction and subsequent expert grading. Additionally, due to the the high per-picture expert time required, MGS scoring is performed on a small number of selected frames from a video. To address these shortcomings, we introduce a method for fully automated real-time MGS scoring of orbital eye tightening, one of the five sub-scores. We define and evaluate the method which is centered around a set of convolutional neural networks (CNNs) and allows live continuous MGS assessment of a mouse in real time. We additionally describe a multithreaded client-server architecture with a graphical user interface that allows convenient use of the developed method for simultaneous real-time MGS scoring of several animals.Item Visual Analytics in Histopathology Diagnostics: a Protocol-Based Approach(The Eurographics Association, 2018) Corvò, Alberto; Westenberg, Michel A.; Driel, Marc A. van; Wijk, Jarke J.van; Puig Puig, Anna and Schultz, Thomas and Vilanova, Anna and Hotz, Ingrid and Kozlikova, Barbora and Vázquez, Pere-PauComputer-Aided-Diagnosis (CAD) systems supporting the diagnostic process are widespread in radiology. Digital Pathology is still behind in the introduction of such solutions. Several studies investigated pathologists' behavior but only a few aimed to improve the diagnostic and report process with novel applications. In this work we designed and implemented a first protocol-based CAD viewer supported by visual analytics. The system targets the optimization of the diagnostic workflow in breast cancer diagnosis by means of three image analysis features that belong to the standard grading system (Nottingham Histologic Grade). A pathologist's routine was tracked during the examination of breast cancer tissue slides and diagnostic traces were analyzed from a qualitative perspective. Accordingly, a set of generic requirements was elicited to define the design and the implementation of the CAD-Viewer. A first qualitative evaluation conducted with five pathologists shows that the interface suffices the diagnostic workflow and diminishes the manual effort. We present promising evidence of the usefulness of our CAD-viewer and opportunities for its extension and integration in clinical practice. As a conclusion, the findings demonstrate that it is feasibile to optimize the Nottingham Grading workflow and, generally, the histological diagnosis by integrating computational pathology data with visual analytics techniques.Item A Prototype Holographic Augmented Reality Interface for Image-Guided Prostate Cancer Interventions(The Eurographics Association, 2018) Morales Mojica, Cristina M.; Velazco Garcia, Jose D.; Navkar, Nikhil V.; Balakrishnan, Shidin; Abinahed, Julien; El Ansari, Walid; Al-Rumaihi, Khalid; Darweesh, Adham; Al-Ansari, Abdulla; Gharib, Mohamed; Karkoub, Mansour; Leiss, Ernst L.; Seimenis, Ioannis; Tsekos, Nikolaos V.; Puig Puig, Anna and Schultz, Thomas and Vilanova, Anna and Hotz, Ingrid and Kozlikova, Barbora and Vázquez, Pere-PauMotivated by the potential of holographic augmented reality (AR) to offer an immersive 3D appreciation of morphology and anatomy, the purpose of this work is to develop and assess an interface for image-based planning of prostate interventions with a head-mounted display (HMD). The computational system is a data and command pipeline that links a magnetic resonance imaging (MRI) scanner/data and the operator, that includes modules dedicated to image processing and segmentation, structure rendering, trajectory planning and spatial co-registration. The interface was developed with the Unity3D Engine (C#) and deployed and tested on a HoloLens HMD. For ergonomics in the surgical suite, the system was endowed with hands-free interactive manipulation of images and the holographic scene via hand gestures and voice commands. The system was tested in silico using MRI and ultrasound datasets of prostate phantoms. The holographic AR scene rendered by the HoloLens HMD was subjectively found superior to desktop-based volume or 3D rendering with regard to structure detection and appreciation of spatial relationships, planning access paths and manual co-registration of MRI and Ultrasound. By inspecting the virtual trajectory superimposed to rendered structures and MR images, the operator observes collisions of the needle path with vital structures (e.g. urethra) and adjusts accordingly. Holographic AR interfacing with wireless HMD endowed with hands-free gesture and voice control is a promising technology. Studies need to systematically assess the clinical merit of such systems and needed functionalities.Item Frontmatter: Eurographics Workshop on Visual Computing for Biology and Medicine 2018(The Eurographics Association, 2018) Puig Puig, Anna; Schultz, Thomas; Vilanova, Anna; Hotz, Ingrid; Kozlikova, Barbora; Vázquez, Pere-Pau; Puig Puig, Anna and Schultz, Thomas and Vilanova, Anna and Hotz, Ingrid and Kozlikova, Barbora and Vázquez, Pere-PauItem Visual Analysis of Evolution of EEG Coherence Networks employing Temporal Multidimensional Scaling(The Eurographics Association, 2018) Ji, Chengtao; Maurits, Natasha M.; Roerdink, Jos B. T. M.; Puig Puig, Anna and Schultz, Thomas and Vilanova, Anna and Hotz, Ingrid and Kozlikova, Barbora and Vázquez, Pere-PauThe community structure of networks plays an important role in their analysis. It represents a high-level organization of objects within a network. However, in many application domains, the relationship between objects in a network changes over time, resulting in the change of community structure (the partition of a network), their attributes (the composition of a community and the values of relationships between communities), or both. Previous animation or timeline-based representations either visualize the change of attributes of networks or the community structure. There is no single method that can optimally show graphs that change in both structure and attributes. In this paper we propose a method for the case of dynamic EEG coherence networks to assist users in exploring the dynamic changes in both their community structure and their attributes. The method uses an initial timeline representation which was designed to provide an overview of changes in community structure. In addition, we order communities and assign colors to them based on their relationships by adapting the existing Temporal Multidimensional Scaling (TMDS) method. Users can identify evolution patterns of dynamic networks from this visualization.Item Global and Local Mesh Morphing for Complex Biological Objects from µCT Data(The Eurographics Association, 2018) Knötel, David; Becker, Carola; Scholtz, Gerhard; Baum, Daniel; Puig Puig, Anna and Schultz, Thomas and Vilanova, Anna and Hotz, Ingrid and Kozlikova, Barbora and Vázquez, Pere-PauWe show how biologically coherent mesh models of animals can be created from µCT data to generate artificial yet naturally looking intermediate objects. The whole pipeline of processing algorithms is presented, starting from generating topologically equivalent surface meshes, followed by solving the correspondence problem, and, finally, creating a surface morphing. In this pipeline, we address all the challenges that are due to dealing with complex biological, non-isometric objects. For biological objects it is often particularly important to obtain deformations that look as realistic as possible. In addition, spatially non-uniform shape morphings that only change one part of the surface and keep the rest as stable as possible are of interest for evolutionary studies, since functional modules often change independently from one another. We use Poisson interpolation for this purpose and show that it is well suited to generate both global and local shape deformations.Item Progressive and Efficient Multi-Resolution Representations for Brain Tractograms(The Eurographics Association, 2018) Mercier, Corentin; Gori, Pietro; Rohmer, Damien; Cani, Marie-Paule; Boubekeur, Tamy; Thiery, Jean-Marc; Bloch, Isabelle; Puig Puig, Anna and Schultz, Thomas and Vilanova, Anna and Hotz, Ingrid and Kozlikova, Barbora and Vázquez, Pere-PauCurrent tractography methods generate tractograms composed of millions of 3D polylines, called fibers, making visualization and interpretation extremely challenging, thus complexifying the use of this technique in a clinical environment. We propose to progressively simplify tractograms by grouping similar fibers into generalized cylinders. This produces a fine-grained multiresolution model that provides a progressive and real-time navigation through different levels of detail. This model preserves the overall structure of the tractogram and can be adapted to different measures of similarity. We also provide an efficient implementation of the method based on a Delaunay tetrahedralization. We illustrate our method using the Human Connectome Project dataset.Item Uncertainty-Guided Semi-Automated Editing of CNN-based Retinal Layer Segmentations in Optical Coherence Tomography(The Eurographics Association, 2018) Zadeh, Shekoufeh Gorgi; Wintergerst, Maximilian W. M.; Schultz, Thomas; Puig Puig, Anna and Schultz, Thomas and Vilanova, Anna and Hotz, Ingrid and Kozlikova, Barbora and Vázquez, Pere-PauConvolutional neural networks (CNNs) have enabled dramatic improvements in the accuracy of automated medical image segmentation. Despite this, in many cases, results are still not reliable enough to be trusted ''blindly''. Consequently, a human rater is responsible to check correctness of the final result and needs to be able to correct any segmentation errors that he or she might notice. For a particular use case, segmentation of the retinal pigment epithelium and Bruch's membrane from Optical Coherence Tomography, we develop a system that makes this process more efficient by guiding the rater to segmentations that are most likely to require attention from a human expert, and by developing semi-automated tools for segmentation correction that exploit intermediate representations from the CNN.We demonstrate that our automated ranking of segmentation uncertainty correlates well with a manual assessment of segmentation quality, and with distance to a ground truth segmentation. We also show that, when used together, uncertainty guidance and our semi-automated editing tools decrease the time required for segmentation correction by more than a factor of three.Item ICG based Augmented-Reality-System for Sentinel Lymph Node Biopsy(The Eurographics Association, 2018) Noll, Matthias; Noa-Rudolph, Werner; Wesarg, Stefan; Kraly, Michael; Stoffels, Ingo; Klode, Joachim; Spass, Cédric; Spass, Gerrit; Puig Puig, Anna and Schultz, Thomas and Vilanova, Anna and Hotz, Ingrid and Kozlikova, Barbora and Vázquez, Pere-PauIn this paper we introduce a novel augmented-reality (AR) system for the sentinel lymph node (SLN) biopsy. The AR system consists of a cubic recording device with integrated stereo near-infrared (NIR) and stereo color cameras, an head mounted display (HMD) for visualizing the SLN information directly into the physicians view and a controlling software application. The labeling of the SLN is achieved using the fluorescent dye indocyanine green (ICG). The dye accumulates in the SLN where it is excited to fluorescence by applying infrared light. The fluorescence is recorded from two directions by the NIR stereo cameras using appropriate filters. Applying the known rigid camera geometry, an ICG depth map can be generated from the camera images, thus creating a live 3D representation of the SLN. The representation is then superimposed to the physicians field of view, by applying a series of coordinate system transformations, that are determined in four separate system calibration steps. To compensate for the head motion, the recording systems is continuously tracked by a single camera on the HMD using fiducial markers. Because the system does not require additional monitors, the physicians attention is kept solely on the operation site. This can potentially decrease the intervention time and render the procedure safer for the patient.Item Parameterization and Feature Extraction for the Visualization of Tree-like Structures(The Eurographics Association, 2018) Lichtenberg, Nils; Lawonn, Kai; Puig Puig, Anna and Schultz, Thomas and Vilanova, Anna and Hotz, Ingrid and Kozlikova, Barbora and Vázquez, Pere-PauThe study and visualization of vascular structures, using 3D models obtained from medical data, is an active field of research. Illustrative visualizations have been applied to this domain in multiple ways. Researchers have tried to make the geometric properties of vasculature more comprehensive and to augment the surface with representations of multivariate clinical data. Techniques that head beyond the application of color-maps or simple shading approaches require a sort of surface parameterization, i.e., texture coordinates, in order to overcome locality. When extracting 3D models, the computation of texture coordinates on the mesh is not always part of the data processing pipeline. We combine existing techniques to a simple, yet effective, parameterization approach that is suitable for tree-like structures. The parameterization is done w.r.t. to a pre-defined source vertex. For this, we present an automatic algorithm, that detects the root of a tree-structure. The parameterization is partly done in screen-space and recomputed per frame. However, the screen-space computation comes with positive features that are not present in object-space approaches. We show how the resulting texture coordinates can be used for varying hatching, contour parameterization, the display of decals, as an additional depth cue and feature extraction.
- «
- 1 (current)
- 2
- 3
- »