Extraction of diagnostic information from hand drawn images for the assessment of visuo-spatial neglect

Fairhurst, Michael and Guest, Richard and Potter, Jonathan and Donnelly, Nick (1999) Extraction of diagnostic information from hand drawn images for the assessment of visuo-spatial neglect. In: Image Processing And Its Applications, 1999. Seventh International Conference on (Conf. Publ. No. 465). Institute of Electrical Engineers pp. 387-391. ISBN 0-85296-717-9. (The full text of this publication is not available from this repository)

The full text of this publication is not available from this repository. (Contact us about this Publication)

Abstract

This paper describes the computer-based analysis of a series of hand-drawn images used to detect the severity of visuo-spatial neglect in stroke patients. Neglect is the inability of a brain damaged patient to respond to stimuli on one side of the visual field. Accurate assessment of the presence and extent of the condition is therefore important at the initial stages of treatment. Existing methods of assessment include simple `pencil and paper' tests where a patient is required to perform such tasks as the location and cancellation of targets and the drawing of simple geometric shapes. Scoring of the latter task is subjective as it requires the assessment of drawings without detailed reference models, resulting in the possibility of inter-rater disagreement and hence variation in clinical assessment. A computer based data capture and analysis system has been developed for the automated and computer-assisted assessment of these tasks. By capturing the data as a stream of co-ordinates using a standard graphics tablet, the `outcome' of the drawing can be measured with accuracy and consistency as well as facilitating the extraction of a range of novel data pertaining to the constructional aspects of the drawing. While the majority of the severe neglect cases are easy to detect by observing omission of components (elements of geometric shapes etc.) many of the diagnostic groupings are not obvious by assessing outcome alone. It is shown that the constructional features provide extra test sensitivity in predicting the outcome. A simple figure completion task is considered here to demonstrate its diagnostic features

Item Type: Conference or workshop item (Paper)
Subjects: Q Science
Divisions: Faculties > Science Technology and Medical Studies > School of Engineering and Digital Arts
Depositing User: M. Nasiriavanaki
Date Deposited: 21 Jun 2009 17:02
Last Modified: 19 May 2014 11:37
Resource URI: http://kar.kent.ac.uk/id/eprint/17320 (The current URI for this page, for reference purposes)
  • Depositors only (login required):