The framework they present is a useful tool for increasing programmer productivity and reducing the overhead of leveraging hardware specific resources. By performing an intelligent search, our framework produces a more efficient image reconstruction implementation in a shorter amount of time.
Motivated by growing concerns with regards to the x-ray dose delivered to the patient, low-dose computed tomography (CT) has gained substantial interest in recent years. However, achieving high-quality CT reconstructions from the limited projection data collected at reduced x-ray radiation is challenging, and iterative algorithms have been shown to perform much better than conventional analytical schemes in these cases. A problem with iterative methods in general is that they require users to set many parameters, and if set incorrectly high reconstruction time and/or low image quality are likely consequences. Since the interactions among parameters can be complex and thus effective settings can be difficult to identify for a given scanning scenario, these choices are often left to a highly-experienced human expert. To help alleviate this problem, we devise a computer-based assistant for this purpose, called dose, quality and speed (DQS)-advisor. It allows users to balance the three most important CT metrics--DQS--by ways of an intuitive visual interface. Using a known gold-standard, the system uses the ant-colony optimization algorithm to generate the most effective parameter settings for a comprehensive set of DQS configurations. A visual interface then presents the numerical outcome of this optimization, while a matrix display allows users to compare the corresponding images. The interface allows users to intuitively trade-off GPU-enabled reconstruction speed with quality and dose, while the system picks the associated parameter settings automatically. Further, once the knowledge has been generated, it can be used to correctly set the parameters for any new CT scan taken at similar scenarios.
Existing high-level, source-to-source compilers can accept input programs in a high-level language (e.g., C) and perform complex automatic parallelization and other mappings using various optimizations. These optimizations often require trade-offs and can benefit from the user's involvement in the process. However, because of the inherent complexity, the barrier to entry for new users of these high-level optimizing compilers can often be high. We propose visualization as an effective gateway for non-expert users to gain insight into the effects of parameter choices and so aid them in the selection of levels best suited to their specific optimization goals. A popular optimization paradigm is polyhedral mapping which achieves optimization by loop transformations. We have augmented a commercial polyhedral-model source-to-source compiler (R-Stream) with an interactive visual tool we call the Polyhedral User Mapping and Assistant Visualizer (PUMA-V). PUMA-V is tightly integrated with the R-Stream source-to-source compiler and allows users to explore the effects of difficult mappings and express their goals to optimize trade-offs. It implements advanced multivariate visualization paradigms such as parallel coordinates and correlation graphs and applies them in the novel setting of compiler optimizations. We believe that our tool allows programmers to better understand complex program transformations and deviations of mapping properties on well understood programs. This in turn will achieve experience and performance portability across programs architectures as well as expose new communities in the computational sciences to the rich features of auto-parallelizing high-level source-to-source compilers. 1
Unlike pandemics in the past, COVID-19 has hit us in the midst of the information age. We have built vast capabilities to collect and store data of any kind that can be analyzed in myriad ways to help us mitigate the impact of this catastrophic disease. Specifically for COVID-19, data analysis can help local governments to plan the allocation of testing kits, testing stations, and primary care units, and it can help them in setting guidelines for residents, such as the need for social distancing, the use of face masks, and when to open local businesses that enable human contact. Further, it can also lead to a better understanding of pandemics in general and so inform policy makers on the regional and national level. All of this can save both cost and lives. In this article, we show the results of an ongoing study we conducted using a prominent regularly updated dataset. We used a pattern mining engine we developed to find specific characteristics of US counties that appear to expose them to higher COVID-19 mortality. Furthermore, we also show that these characteristics can be used to predict future COVID-19 mortality. CCS Concepts: • Human-centered computing → Interactive systems and tools; Visualization systems and tools; Visual analytics; • Computing methodologies → Causal reasoning and diagnostics;
Graphical processing units (GPUs) have become widely adopted in the medical imaging community. The parallel SIMD nature of GPUs maps perfectly to many reconstruction algorithms. Because of this, it is relatively straightforward to parallelize common reconstruction algorithms (e.g. FDK backprojection). This means that significant performance improvements must come from careful memory optimizations, exploiting ASICs and a few other tricks to boost instruction throughput. We present optimizations that build off of previous work to optimize a GPU accelerated FDK backprojection implementation using the RabbitCT dataset.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.