gms | German Medical Science

63. Jahrestagung der Deutschen Gesellschaft für Medizinische Informatik, Biometrie und Epidemiologie e. V. (GMDS)

Deutsche Gesellschaft für Medizinische Informatik, Biometrie und Epidemiologie

02. - 06.09.2018, Osnabrück

An interactive hive plot for the clinical oncological routine

Meeting Abstract

  • Julia Dieter - Deutsches Krebsforschungszentrum, Heidelberg, Deutschland
  • Alexander Knurr - Deutsches Krebsforschungszentrum, Heidelberg, Deutschland
  • Janko Ahlbrandt - Deutsches Krebsforschungszentrum, Heidelberg, Deutschland
  • Analie Pascoe Pérez - Deutsches Krebsforschungszentrum, Heidelberg, Deutschland
  • Frank Ückert - Deutsches Krebsforschungszentrum, Heidelberg, Deutschland

Deutsche Gesellschaft für Medizinische Informatik, Biometrie und Epidemiologie. 63. Jahrestagung der Deutschen Gesellschaft für Medizinische Informatik, Biometrie und Epidemiologie e.V. (GMDS). Osnabrück, 02.-06.09.2018. Düsseldorf: German Medical Science GMS Publishing House; 2018. DocAbstr. 99

doi: 10.3205/18gmds022, urn:nbn:de:0183-18gmds0222

Published: August 27, 2018

© 2018 Dieter et al.
This is an Open Access article distributed under the terms of the Creative Commons Attribution 4.0 License. See license information at



Background: In healthcare, increasing amounts of multivariate data with complex relationships among their variables are continuously generated when performing clinical routines [1]. However, the high dimensionality of data heavily challenges human cognitive abilities. Therefore, analytical tools should be applied to leverage the data in order to improve patient treatment [2]. As a basis for decision-support in clinical oncological routine, patient-specific information often has to be manually extracted and collected from several clinical documentation systems or records, filtered for relevant information (e.g. patient characteristics, diagnosis or treatments), put into content-related context and set into temporal correspondences, often under time pressure. Such information-extraction workflows are commonly optimized through data visualization. Several plots already have been developed and are used on medical data. However, they focus on specific clinical specialties such as cardiac, pulmonary, chemistry/laboratory or medication data [1] without indicating relationships between them. Furthermore, the joint visualization of data within and across patients is especially complicated due to data properties (e.g. heterogeneity, complexity, time-dependence, etc.), often leading to cluttered views and information overflow.

Aim of the study: To address these problems, we propose a novel visualization approach consisting of a hive plot variation for the clinical oncological routine, presented via a prototype mock-up. The clinical hive plot depicts relationships between patient-specific data from different clinical specialties or between patients within and across cohorts of interest (e.g. patients with a specified diagnosis and treatment sequence [3]) under consideration of time-dependency and the application of interactive features.

Proposed methods: The proposed clinical hive plot prototype mock-up was designed based on the already available hive plot, originally used for network visualization [4]. After having identified the clinicians’ need for visualizations of multivariate, complex and time-dependent data, a prototype was conceptualized by an interdisciplinary team providing clinical, medical informatics and biological inputs.

Discussion: The proposed clinical hive plot comprises a customizable number of axes harboring data of different types (e.g. nominal, ordinal or interval) and clinical specialties (e.g. diagnostics, therapy or genetics), as well as parameters of interest (e.g. outcome parameters such as survival rate). Thereby, the relationship between data points can be visualized for single patients (represented by connecting lines between axes, i.e. “edge”) and compared to cohorts (multiple edges). Time-dependent visualizations are enabled by

adding adjacent axes consecutively, according to the temporal order of events (e.g. three therapy axes would reflect three consecutive therapies), and by
displaying the angle between axes proportionally to the length of time intervals (hence larger angles indicate larger time intervals).

Interactive features include a slider to define edge display thresholds, i.e. the number of patients represented by a single edge (if the slider is set to zero, each edge corresponds to a single patient). Furthermore, a hover over feature is included to fade in information on axes points or edges of interest. Cohorts can be selected by marking specific edges. As a next step, the prototype shall be evaluated for its use in the clinical oncological routine, as a prerequisite for implementation.

The authors declare that they have no competing interests.

The authors declare that an ethics committee vote is not required.


Wanderer JP, Nelson SE, Ehrenfeld JM, Monahan S, Park S. Clinical Data Visualization: The Current State and Future Needs. J Med Syst. 2016;40(12):275. DOI: 10.1007/s10916-016-0643-x External link
Yu P, Artz D, Warner J. Electronic health records (EHRs): supporting ASCO's vision of cancer care. Am Soc Clin Oncol Educ Book. 2014:225–31. DOI: 10.14694/EdBook_AM.2014.34.225 External link
Badgeley MA, Shameer K, Glicksberg BS, Tomlinson MS, Levin MA, McCormick PJ, et al. EHDViz: clinical dashboard development using open-source technologies. BMJ Open. 2016;6(3):e010579. DOI: 10.1136/bmjopen-2015-010579 External link
Krzywinski M, Birol I, Jones SJM, Marra MA. Hive plots-rational approach to visualizing networks. Brief Bioinformatics. 2012;13(5):627–44. DOI: 10.1093/bib/bbr069. External link