The semiotics of medical image Segmentation
•A semiotics-driven framework for understanding knowledge-driven interaction in medical image segmentation is proposed.•This framework is grounded in Peircean semiotics in order to structure and characterize how particular interactions are interpreted by both the user and the computer.•Using the not...
Saved in:
Published in | Medical image analysis Vol. 44; pp. 54 - 71 |
---|---|
Main Authors | , , , |
Format | Journal Article |
Language | English |
Published |
Netherlands
Elsevier B.V
01.02.2018
Elsevier BV |
Subjects | |
Online Access | Get full text |
Cover
Loading…
Summary: | •A semiotics-driven framework for understanding knowledge-driven interaction in medical image segmentation is proposed.•This framework is grounded in Peircean semiotics in order to structure and characterize how particular interactions are interpreted by both the user and the computer.•Using the notion of interface metaphors, this framework shows how metaphor quality metrics can be used to analyze interaction and improve ease-of-use in communicating complex anatomical knowledge.
As the interaction between clinicians and computational processes increases in complexity, more nuanced mechanisms are required to describe how their communication is mediated. Medical image segmentation in particular affords a large number of distinct loci for interaction which can act on a deep, knowledge-driven level which complicates the naive interpretation of the computer as a symbol processing machine. Using the perspective of the computer as dialogue partner, we can motivate the semiotic understanding of medical image segmentation. Taking advantage of Peircean semiotic traditions and new philosophical inquiry into the structure and quality of metaphors, we can construct a unified framework for the interpretation of medical image segmentation as a sign exchange in which each sign acts as an interface metaphor. This allows for a notion of finite semiosis, described through a schematic medium, that can rigorously describe how clinicians and computers interpret the signs mediating their interaction. Altogether, this framework provides a unified approach to the understanding and development of medical image segmentation interfaces.
[Display omitted] |
---|---|
Bibliography: | ObjectType-Article-1 SourceType-Scholarly Journals-1 ObjectType-Feature-2 content type line 23 |
ISSN: | 1361-8415 1361-8423 |
DOI: | 10.1016/j.media.2017.11.007 |