Visual analytics for the exploration of tumor tissue characterization

Information

• Publication Type: Journal Paper with Conference Talk
• Workgroup(s)/Project(s):
  • Visualization Group
• Date: 2015
• Journal: In Computer Graphics Forum
• Lecturer:
  • Hamish Carr
  • Kwan Liu Ma
  • Giuseppe Santucci

Abstract

Tumors are heterogeneous tissues consisting of multiple regions with distinct characteristics. Characterization ofthese intra-tumor regions can improve patient diagnosis and enable a better targeted treatment. Ideally, tissuecharacterization could be performed non-invasively, using medical imaging data, to derive per voxel a number offeatures, indicative of tissue properties. However, the high dimensionality and complexity of this imaging-derivedfeature space is prohibiting for easy exploration and analysis - especially when clinical researchers require toassociate observations from the feature space to other reference data, e.g., features derived from histopathologicaldata. Currently, the exploratory approach used in clinical research consists of juxtaposing these data, visuallycomparing them and mentally reconstructing their relationships. This is a time consuming and tedious process,from which it is difficult to obtain the required insight. We propose a visual tool for: (1) easy exploration and visualanalysis of the feature space of imaging-derived tissue characteristics and (2) knowledge discovery and hypothesisgeneration and confirmation, with respect to reference data used in clinical research. We employ, as central view,a 2D embedding of the imaging-derived features. Multiple linked interactive views provide functionality for theexploration and analysis of the local structure of the feature space, enabling linking to patient anatomy andclinical reference data. We performed an initial evaluation with ten clinical researchers. All participants agreedthat, unlike current practice, the proposed visual tool enables them to identify, explore and analyze heterogeneousintra-tumor regions and particularly, to generate and confirm hypotheses, with respect to clinical reference data.

Additional Files and Images

Weblinks

No further information available.

BibTeX

@article{raidou_EuroVis15,
  title =      "Visual analytics for the exploration of tumor tissue
               characterization",
  author =     "Renata Raidou and Uulke A van der Heide and Cuong V Dinh and
               Ghazaleh Ghobadi and Jesper Follsted Kallehauge and Marcel
               Breeuwer and Anna Vilanova i Bartroli",
  year =       "2015",
  abstract =   "Tumors are heterogeneous tissues consisting of multiple
               regions with distinct characteristics. Characterization
               ofthese intra-tumor regions can improve patient diagnosis
               and enable a better targeted treatment. Ideally,
               tissuecharacterization could be performed non-invasively,
               using medical imaging data, to derive per voxel a number
               offeatures, indicative of tissue properties. However, the
               high dimensionality and complexity of this
               imaging-derivedfeature space is prohibiting for easy
               exploration and analysis - especially when clinical
               researchers require toassociate observations from the
               feature space to other reference data, e.g., features
               derived from histopathologicaldata. Currently, the
               exploratory approach used in clinical research consists of
               juxtaposing these data, visuallycomparing them and mentally
               reconstructing their relationships. This is a time consuming
               and tedious process,from which it is difficult to obtain the
               required insight. We propose a visual tool for: (1) easy
               exploration and visualanalysis of the feature space of
               imaging-derived tissue characteristics and (2) knowledge
               discovery and hypothesisgeneration and confirmation, with
               respect to reference data used in clinical research. We
               employ, as central view,a 2D embedding of the
               imaging-derived features. Multiple linked interactive views
               provide functionality for theexploration  and  analysis  of 
               the  local  structure  of  the  feature  space,  enabling 
               linking  to  patient  anatomy  andclinical reference data.
               We performed an initial evaluation with ten clinical
               researchers. All participants agreedthat, unlike current
               practice, the proposed visual tool enables them to identify,
               explore and analyze heterogeneousintra-tumor regions and
               particularly, to generate and confirm hypotheses, with
               respect to clinical reference data.",
  journal =    "In Computer Graphics Forum",
  URL =        "https://www.cg.tuwien.ac.at/research/publications/2015/raidou_EuroVis15/",
}