Search

In Full-Reference Image Quality Assessment (FR-IQA) images are compared with ground truth images that are known to be of high visual quality. These metrics are utilized in order to rank algorithms under test on their image quality performance. Throughout the progress of Monte Carlo rendering processes we often wish to determine whether images being rendered are of sufficient visual quality, without the availability of a ground truth image. In such cases FR-IQA metrics are not applicable and we instead must utilise No-Reference Image Quality Assessment (NR-IQA) measures to make predictions about the perceived quality of unconverged images. In this work we propose a deep learning approach to NR-IQA, trained specifically on noise from Monte Carlo rendering processes, which significantly outperforms existing NR-IQA methods and can produce quality predictions consistent with FR-IQA measures that have access to ground truth images., Computer Graphics and Visual Computing (CGVC), Vision and Learning, 23, 31, Joss Whittle and Mark W. Jones, CCS Concepts: Computing methodologies --> Machine learning; Neural networks; Computer graphics; Image processing

CONTEXT Current endoscopes have limitations during use in polypectomies. Specifically, polyps that are flat, broad-based and sessile are more difficult to resect. Routine polypectomy procedures allow one endoscopic device to be used at a time limiting the endoscopist. More advanced scopes are not readily available at smaller community hospitals, limiting the endoscopist to using the resources available to them. METHODS The modification of the standard polypectomy method described here employs both an endoscopic forceps and an endoscopic snare to be used simultaneously during colonoscopy with a single lumen colonoscope. The forceps is introduced into the endoscope so the head is just projecting from the distal end of the scope. The snare is then placed just proximal to the head of the forceps outside of the endoscope. The endoscope is reinserted into the colon until the polyp is reached. Using the snare the polyp is elevated and then the snare secured around the base. RESULTS This resulted in easier, faster, and more complete removal of flat sessile and poorly located pedunculated polyps on the first try. This technique has been employed successfully in over 20 patients at our institution. CONCLUSIONS This new method adds another technique for endoscopists when presented with difficult polypectomies.

Background: We are increasingly using recording devices with multiple sensors operating at high frequencies to produce large volumes of data which are problematic to interpret. A particularly challenging example comes from studies on animals and humans where researchers use animal-attached accelerometers on moving subjects to attempt to quantify behaviour, energy expenditure and condition. Results: The approach taken effectively concatinated three complex lines of acceleration into one visualization that highlighted patterns that were otherwise not obvious. The summation of data points within sphere facets and presentation into histograms on the sphere surface effectively dealt with data occlusion. Further frequency binning of data within facets and representation of these bins as discs on spines radiating from the sphere allowed patterns in dynamic body accelerations (DBA) associated with different postures to become obvious. Method: We examine the extent to which novel, gravity-based spherical plots can produce revealing visualizations to incorporate the complexity of such multidimensional acceleration data using a suite of different acceleration-derived metrics with a view to highlighting patterns that are not obvious using current approaches. The basis for the visualisation involved three-dimensional plots of the smoothed acceleration values, which then occupied points on the surface of a sphere. This sphere was divided into facets and point density within each facet expressed as a histogram. Within each facet-dependent histogram, data were also grouped into frequency bins of any desirable parameters, most particularly dynamic body acceleration (DBA), which were then presented as discs on a central spine radiating from the facet. Greater radial distances from the sphere surface indicated greater DBA values while greater disc diameter indicated larger numbers of data points with that particular value. Conclusions: We indicate how this approach links behaviour and proxies for energetics and can inform our identification and understanding of movement-related processes, highlighting subtle differences in movement and its associated energetics. This approach has ramifications that should expand to areas as disparate as disease identification, lifestyle, sports practice and wild animal ecology. UCT Science Faculty Animal Ethics 2014/V10/PR (valid until 2017). Fil: Wilson, Rory P.. Swansea University; Reino Unido Fil: Holton, Mark D.. Swansea University; Reino Unido Fil: Walker, James. Swansea University; Reino Unido Fil: Shepard, Emily L. C.. Swansea University; Reino Unido Fil: Scantlebury, D. Mike. Queen’s University Belfast; Reino Unido Fil: Wilson, Vianney L.. Swansea University; Reino Unido Fil: Wilson, Gwendoline I.. Swansea University; Reino Unido Fil: Tysse, Brenda. Swansea University; Reino Unido Fil: Gravenor, Mike. Swansea University; Reino Unido Fil: Ciancio Blanc, Javier Ernesto. Consejo Nacional de Investigaciones Científicas y Técnicas. Centro Nacional Patagónico; Argentina Fil: McNarry, Melitta A.. Swansea University; Reino Unido Fil: Mackintosh, Kelly A.. Swansea University; Reino Unido Fil: Qasem, Lama. Qatar University; Qatar Fil: Rosell, Frank. Telemark University College; Noruega Fil: Graf, Patricia M.. Telemark University College; Noruega. University of Natural Resources and Life Sciences; Austria Fil: Quintana, Flavio Roberto. Consejo Nacional de Investigaciones Científicas y Técnicas. Centro Científico Tecnológico Conicet - Centro Nacional Patagónico. Instituto de Biología de Organismos Marinos; Argentina Fil: Gómez Laich, Agustina Marta. Consejo Nacional de Investigaciones Científicas y Técnicas. Centro Científico Tecnológico Conicet - Centro Nacional Patagónico. Instituto de Biología de Organismos Marinos; Argentina Fil: Sala, Juan Emilio. Consejo Nacional de Investigaciones Científicas y Técnicas. Centro Científico Tecnológico Conicet - Centro Nacional Patagónico. Instituto de Biología de Organismos Marinos; Argentina Fil: Mulvenna, Christina C.. Swansea University; Reino Unido Fil: Marks, Nicola J.. Swansea University; Reino Unido Fil: Jones, Mark W.. Swansea University; Reino Unido