ViES: Visual-interactive Exploration for individualized Selection of relevant data regions
Final Report Abstract
The results of this project constitute advances on both: the methodical level and the application level. On the one hand, we established an effective mixture of conventional analysis approaches together with novel visualization and interaction techniques and statistical computations. On the other hand, we were able to demonstrate that visual analysis methods help to reveal previously undetected retinal changes at an early stage of ocular diseases and generally ease the in-depth analysis of retinal OCT data. The most important results of this project are: Regarding the ophthalmology: We evaluated several experimental and cross-sectional studies that advance the current state-of-ophthalmic-research: • Introducing a spatially precise evaluation methodology to assess retinal conditions in the context of different ocular diseases. Accelerating and simplifying the data evaluation pipeline for OCT-based cohort studies in general. • Expanding the ophthalmic diagnosis portfolio based on a deepened understanding of subtle and localized changes of intraretinal layers induced by type 1 and 2 diabetes mellitus. Providing a tool that aids biomarker development for staging diabetic peripheral neuropathy based on retinal changes. • Establishing a university network for performing interdisciplinary studies, involving ophthalmology, pediatrics, endocrinology etc., to investigate diabetes-induced retinal changes. Initializing inter-university cooperations, e.g., Gutenberg Health Study. Regarding the visualization: We developed several new visual analysis approaches that advance the current state-of-visualization-research: • Techniques for visual analysis of OCT data on different levels of granularity. New is the support for selecting relevant data regions via a seamless switch on the levels of data presentation, data resolution, and data analysis tasks. • Visualization of influencing acquisition parameters and visual analysis of OCT data quality. New is the support for choosing suitable parameter settings and for differentiating between artifacts and actual retinal changes. In addition, data screening and appropriate grid-based reduction of complex OCT data are enabled. • Feature-driven comparative visualization of multiple OCT datasets. New are the comparison of patient data to control data, the visualization via deviation maps and grids, and the application of statistical tests to highlight significant differences. • Visual analysis framework for retinal changes. New are a unified access to data from different sources, a graphical user interface that provides the developed visualization and interaction techniques in an intuitive way, and an associated workflow that eases the OCT data analysis of individual patients and patient groups. Regarding the knowledge transfer: We established an effective exchange between research and industry that helps to solve actual practical problems: • Broadened software portfolio for analyzing retinal OCT data of individual patients and groups of patients. Utilization of coordinated computational, visual, and interactive designs derived from the combined expertise of visualization, ophthalmic, and industry experts. • New options to provide an industry-level support for ophthalmic research and the evaluation of ophthalmic study data. Possibility to distribute dedicated research software, besides current commercial OCT software targeted at clinical practice. • New and continued cooperations and knowledge sharing between industry and research institutes, e.g., HE and UCSD. Informing and involving of different OCT device manufacturers, e.g., Carl Zeiss AG, Germany.
Publications
- “Visual Analysis of Optical Coherence Tomography Data in Ophthalmology”. In: Proceedings of the EuroVis Workshop on Visual Analytics (EuroVA). ed. by M. Sedlmair and C. Tominski. Barcelona, Spain: The Eurographics Association, 2017. isbn: 978-3-03868-042-0 [Best paper award]
M. Röhlig et al.
(See online at https://doi.org/10.2312/eurova.20171117) - “Visual Analysis of Retinal OCT Data”. In: Klinische Monatsblätter für Augenheilkunde 234.12 (2017), pp. 1463–1471.
M. Röhlig et al.
(See online at https://doi.org/10.1055/s-0043-121705) - Spectral-Domain Optical Coherence Tomography for Determination of Retinal Thickness in Pediatric Patients with Mild-To-Moderate Chronic Kidney Disease: A Cross-Sectional Study”. In: Current Eye Research 44.2 (2018), pp. 206–211
R. K. Prakasam et al.
(See online at https://doi.org/10.1080/02713683.2018.1522649) - “The corneal subbasal nerve plexus and thickness of the retinal layers in pediatric type 1 diabetes and matched controls”. In: Scientific Reports 8.1 (2018)
A. Götze et al.
(See online at https://doi.org/10.1038/s41598-017-18284-z) - “Visual Analysis of Retinal Changes with Optical Coherence Tomography”. In: The Visual Computer 34.9 (2018), pp. 1209–1224
M. Röhlig et al.
(See online at https://doi.org/10.1007/s00371-018-1486-x) - “Deviation maps for understanding thickness changes of inner retinal layers in children with type 1 diabetes mellitus”. In: Current Eye Research 44.7 (2019), pp. 746–752
R. K. Prakasam et al.
(See online at https://doi.org/10.1080/02713683.2019.1591463) - “Grid-Based Exploration of OCT Thickness Data of Intraretinal Layers”. In: Proceedings of the 14th International Joint Conference on Computer Vision, Imaging and Computer Graphics Theory and Applications (VISIGRAPP); ed. by A. Kerren, C. Hurter, and J. Braz. Vol. 3. (IVAPP). INSTICC. SciTePress, Feb. 2019, pp. 129–140 [Best student paper award]
M. Röhlig et al.
(See online at https://doi.org/10.5220/0007580001290140)
