Search

International audience; In information visualization, it has become mandatory to assess visualization techniques efficiency either to write a survey, optimize a technique or even design a new one. To do so, the common way is to conduct user evaluations through which human subjects are asked to solve a task on different visualization techniques while their performances are measured to assess which technique is the most efficient. These evaluations can be complex to design and setup in order not to be biased and, in the end, their results can become contestable when the evaluation methods standards evolve. To overcome these flaws, new evaluation methods are emerging, mostly making use of modern and efficient computer vision techniques such as deep learning. These new methods rely on a strong assumption that has not been studied deeply enough yet: humans and deep learning models performances can be correlated. This paper explores the performances of both a state-of-the-art deep neural network and human subjects on an outlier detection task taken from a previous experiment of the literature. The objective is to study whether the machine and humans behaviors were different or if some correlations can be observed. Our study shows that their results are significantly correlated and a machine learning model efficiently learned to predict human performances using deep neural network metrics as input. Hence, this work presents a use case where using a deep neural network to assess human subjects performances is efficient.

Purpose: Size and density measurements of the ablation zones on contrast-enhanced computed tomography (CT) after radiofrequency ablation (RFA) of hepatic metastases of primary breast or colorectal cancer were acquired over time. Materials and Methods: Twenty-five liver metastases [colorectal cancer (CRC): n = 16; mean size: 19.6 ± 8.5 mm; breast cancer (BC): n = 9; 27.9 ± 13.6 mm] in 15 patients (CRC: n = 11; age: 65.4 ± 6.5 years; BC: n = 4; 62.0 ± 13.8 years) treated by RFA were included in this retrospective study. All patients had undergone postinterventional serial follow-up using multidetector CT (MDCT) (1 day/1/4/7/10/14/18/23/>24 months) without evidence of local tumor recurrence during the follow-up. The ablation zones were evaluated using a commercial software tool (Syngo CT Oncology) in order to determine volumetric, RECIST-, WHO- and density changes over the course of time. Results were compared by applying repeated measures analysis of variance and displayed graphically. Results: The RF ablation zones demonstrated significant shrinkage (P ≤ 0.05) over the first 7 months (volume, RECIST, WHO) of the follow-up. Follow-up after 7 months did not show any significant changes in size (P > 0.05) (mean volume (ml): 55.2/34.7/26.3/16.5/12.7/10.0/8.9/8.1/7.5; RECIST (mm): 58.7/49.3/43.7/37.8/34.2/31.3/29.1/27.3/24.8; WHO (mm2): 2458.3/1769.3/1341.8/1027.1/870.1/720.2/649.0/570.4/511.3). Mean density values decreased significantly between 1 day (58.9 HU) and 1 month (47.5 HU) after the procedure. Conclusion: Typical changes in size and density values of RF-induced, recurrence-free ablation zones after RFA of hepatic metastases of colorectal and breast cancer were acquired, showing a significant decrease in density of the ablation zone within the first month and significant shrinkage within the first 7 months after RFA.

The paper introduces the need for automated evaluation and presents the experience of automating all the evaluations of a course in Computer Science in the sophomore year of an engineering degree. First, the paper describes the features needed and developed for that course and the positive results for both professors and students. The main advantage of automated evaluation is that it allows real continuous grading for all types of activities: short answers and exercises during the class, homework, short exercises evaluated every 10 days in class, medium term evaluations and the final grade for the course. A significant benefit of this practice is that it allows the professor, from the very beginning of the course, to monitor how the students perform each task. The professor can see in real time the marks of an exercise or evaluation, the global evolution of the class or the status of a specific student. The students also have immediate feedback from their exercises and the total points obtained at any given time providing greater involvement in the course.

Describing test results in a standardized format facilitates the exchange of information between different types of tools that are used to evaluate Web sites for accessibility. The Evaluation and Report Language (EARL) is a machine-readable, platform-independent, and vendor-neutral format that builds on Semantic Web technologies to provide a powerful data model for expressing test results. While the primary motivation for developing this language is to facilitate the automation of Web accessibility evaluations, it is has been designed so that it can serve generic Web quality assurance purposes as well as other use cases outside the scope of Web site evaluation.

User evaluation is generally performed early in the development process to reveal usability problems, design flaws, and errors and correct them before deploying. Due to the short iterations of agile development, implementing user evaluation as part of the development process is a challenge that is often neglected. In a previous work, we proposed an approach that would enable the integration of user evaluation throughout the development process, by managing and automating user evaluation activities from with the integrated development environment (IDE). In this work, we focus on a case study in which small-sized agile software teams, made up of students in an annual software engineering project course, applied our integrated user evaluation approach for developing their software projects. The feedbacks from these agile teams show the intuitiveness and effectiveness of our integration approach.

The purpose of this study was to investigate differences in change of size and CT value between local recurrences and tumor-free areas after CT-guided radiofrequency ablation (RFA) of hepatic metastases during follow-up by means of dedicated software for automatic evaluation of hepatic lesions. Thirty-two patients with 54 liver metastases from breast or colorectal cancer underwent triphasic contrast-enhanced multidetector-row computed tomography (MDCT) to evaluate hepatic metastatic spread and localization before CT-guided RFA and for follow-up after intervention. Sixteen of these patients (65.1 + or - 10.3 years) with 30 metastases stayed tumor-free (group 1), while the other group (n = 16 with 24 metastases; 62.0 + or - 13.8 years) suffered from local recurrent disease (group 2). Applying an automated software tool (SyngoCT Oncology; Siemens Healthcare, Forchheim, Germany), size parameters (volume, RECIST, WHO) and attenuation were measured within the lesions before, 1 day after, and 28 days after RFA treatment. The natural logarithm (ln) of the quotient of the volume 1 day versus 28 days after RFA treament was computed: lnQ1//28/0(volume). Analogously, ln ratios of RECIST, WHO, and attenuation were computed and statistically evaluated by repeated-measures ANOVA. One lesion in group 2 was excluded from further evaluation due to automated missegmentation. Statistically significant differences between the two groups were observed with respect to initial volume, RECIST, and WHO (p

The purpose of this article – is to create theoretical model in order to enable in automated manner to compare different solutions of building, its part or structure design and to select the best alternative under specific circumstances. Summarizing variant design theory and practice applications in Lithuania and abroad, evaluation model of alternative design solutions is created. By assistance of this model it will be possible to evaluate design solutions which can be characterized by quantitative and qualitative indicators which possibly have different weight, dimention and direction of optimization (maksimization or minimization). Automated alternatives’ evaluation working principle is demonstrated by selecting the best facade system to cover the building. For this purpose four building facades alternatives is under consideration. It is soundly described how all stages of alternatives comparison is performed and which multiple criteria evaluation methods can be applied by selecting the best alternative.

The Web has been growing in size and complexity and is used for the most diverse activities in our everyday life, becoming almost indispensable. Besides, Web applications are becoming more popular, and consequently used by a wide range of people. Thus, it is important to evaluate the accessibility of those applications to guarantee that everyone can access the information. Currently, there are some tools to evaluate the accessibility of classical Web pages, which use WCAG guidelines. However, Web applications impose different challenges, so it is mandatory to find a way to automatically obtain the dynamically introduced HTML code, in order to evaluate what users really experience. This paper details a new process of accessibility evaluation of Web applications, which evaluates the content by triggering possible events that partially change the Web page. It also presents an experimental study with several Web applications, demonstrating the potential of this framework in evaluating Web applications.