Browsing by Author "Wijk, Jarke J. van"

Now showing 1 - 5 of 5
  • Thumbnail Image
    Item
    ChronoCorrelator: Enriching Events with Time Series
    (The Eurographics Association and John Wiley & Sons Ltd., 2019) van Dortmont, Martijn; Elzen, Stef van den; Wijk, Jarke J. van; Gleicher, Michael and Viola, Ivan and Leitte, Heike
    Event sequences and time series are widely recorded in many application domains; examples are stock market prices, electronic health records, server operation and performance logs. Common goals for recording are monitoring, root cause analysis and predictive analytics. Current analysis methods generally focus on the exploration of either event sequences or time series. However, deeper insights are gained by combining both. We present a visual analytics approach where users can explore both time series and event data simultaneously, combining visualization, automated methods and human interaction. We enable users to iteratively refine the visualization. Correlations between event sequences and time series can be found by means of an interactive algorithm, which also computes the presence of monotonic effects. We illustrate the effectiveness of our method by applying it to real world and synthetic data sets.
  • Thumbnail Image
    Item
    Multiple Scale Visualization of Electronic Health Records to Support Finding Medical Narratives
    (The Eurographics Association, 2021) Linden, Sanne van der; Wijk, Jarke J. van; Funk, Mathias; Oeltze-Jafra, Steffen and Smit, Noeska N. and Sommer, Björn and Nieselt, Kay and Schultz, Thomas
    Electronic Health Records (EHRs) contain rich medical information about patients, possibly hundreds of notes, lab results, images and other information. Doctors can easily be overwhelmed by this wealth of information. For their daily work, they need to derive narratives from all this information to get insights into the main issues of their patients. Standard solutions show all the information in linear lists, often leading to cognitive overload; research solutions provide timelines and relations between the notes but provide too much fragmented information. We propose MEDeNAR, a system for enabling medical professionals to obtain insights from EHRs based on the different tasks in their workflow. The key aspects of our system are the introduction of an intermediate level that summarizes the information using clustering and NLP methods. The results are visualized along a timeline and provide easy access to the detailed descriptions in notes and lab results at the EHR level. We designed the system using an iterative design process in collaboration with 18 doctors, two nurses and 14 domain experts. During the final evaluation, the doctors ranked our system higher than a standard baseline solution and a variation for the used NLP methods.
  • Thumbnail Image
    Item
    PerSleep: A Visual Analytics Approach for Performance Assessment of Sleep Staging Models
    (The Eurographics Association, 2021) Garcia Caballero, Humberto S.; Corvò, Alberto; Meulen, Fokke van; Fonseca, Pedro; Overeem, Sebasitaan; Wijk, Jarke J. van; Westenberg, Michel A.; Oeltze-Jafra, Steffen and Smit, Noeska N. and Sommer, Björn and Nieselt, Kay and Schultz, Thomas
    Machine learning is becoming increasingly popular in the medical domain. In the near future, clinicians expect predictive models to support daily tasks such as diagnosis and prognostic analysis. For this reason, it is utterly important to evaluate and compare the performance of such models so that clinicians can safely rely on them. In this paper, we focus on sleep staging wherein machine learning models can be used to automate or support sleep scoring. Evaluation of these models is complex because sleep is a natural process, which varies among patients. For adoption in clinical routine, it is important to understand how the models perform for different groups of patients. Moreover, models can be trained to recognize different characteristics in the data, and model developers need to understand why and how performance of the different models varies. To address these challenges, we present a visual analytics approach to evaluate the performance of predictive models on sleep staging and to help experts better understand these models with respect to patient data (e.g., conditions, medication, etc.). We illustrate the effectiveness of our approach by comparing multiple models trained on real-world sleep staging data with experts.
  • Thumbnail Image
    Item
    V-Awake: A Visual Analytics Approach for Correcting Sleep Predictions from Deep Learning Models
    (The Eurographics Association and John Wiley & Sons Ltd., 2019) Garcia Caballero, Humberto; Westenberg, Michel; Gebre, Binyam; Wijk, Jarke J. van; Gleicher, Michael and Viola, Ivan and Leitte, Heike
    The usage of deep learning models for tagging input data has increased over the past years because of their accuracy and highperformance. A successful application is to score sleep stages. In this scenario, models are trained to predict the sleep stages of individuals. Although their predictive accuracy is high, there are still misclassifications that prevent doctors from properly diagnosing sleep-related disorders. This paper presents a system that allows users to explore the output of deep learning models in a real-life scenario to spot and analyze faulty predictions. These can be corrected by users to generate a sequence of sleep stages to be examined by doctors. Our approach addresses a real-life scenario with absence of ground truth. It differs from others in that our goal is not to improve the model itself, but to correct the predictions it provides. We demonstrate that our approach is effective in identifying faulty predictions and helping users to fix them in the proposed use case.
  • Thumbnail Image
    Item
    Visual Analytics in Digital Pathology: Challenges and Opportunities
    (The Eurographics Association, 2019) Corvò, Alberto; Westenberg, Michel A.; Wimberger-Friedl, Reinhold; Fromme, Stephan; Peeters, Michel M. R.; Driel, Marc A. van; Wijk, Jarke J. van; Kozlíková, Barbora and Linsen, Lars and Vázquez, Pere-Pau and Lawonn, Kai and Raidou, Renata Georgia
    The advances in high-throughput digitization, digital pathology systems, and quantitative image analysis opened new horizons in pathology. The diagnostic work of the pathologists and their role is likely to be augmented with computer-assistance and more quantitative information at hand. The recent success of artificial intelligence (AI) and computer vision methods demonstrated that in the coming years machines will support pathologists in typically tedious and highly subjective tasks and also in better patient stratification. In spite of clear future improvements in the diagnostic workflow, questions on how to effectively support the pathologists and how to integrate current data sources and quantitative information still persist. In this context, Visual Analytics (VA) - as the discipline that aids users to solve complex problems with an interactive and visual approach - can play a vital role to support the cognitive skills of pathologists and the large volumes of data available. To identify the main opportunities to employ VA in digital pathology systems, we conducted a survey with 20 pathologists to characterize the diagnostic practice and needs from a user perspective. From our findings, we discuss how VA can leverage quantitative image data to empower pathologists with new advanced digital pathology systems.