Post-doctoral Fellowship in Artificial Intelligence and Medical Image Interpretation
Post-doct{*filter*}Fellowship in Artificial Intelligence
Integrating Knowledge-Based Decision Support and Medical Image Interpretation
We are recruiting staff to work on a major collaborative project being led by
the Centre for Health Informatics and Multiprofessional Education at University
College London. The partners in the project include the Advanced Computation
Laboratory of the Imperial Cancer Research Fund, the Department of Medical
Physics at UCL, the Whittington Hospital Trust and INSERM's (the French
National Institute for Medical Research) Unit 494 at the La Piti-Salptrire
hospital in Paris.
This project is developing a generic model for enhanced imaging workstations
that will allow clinicians to make optimal use of imaging information. It
brings together work in computer aids for protocol management, knowledge-based
decision support and medical image processing.
The project builds on CADMIUM - Computer Assisted Decision Making for Image
Understanding in Medicine - an approach to decision support systems for
radiology. CADMIUM provides a symbolic model of decision making augmented with
a model of the tasks involved in interpreting images. The model of image
interpretation tasks is used in conjunction with a symbolic knowledge base to
control the execution of image processing operations that provide evidence as
required to support the reasoning. A pilot study has shown that the system can
significantly improve the performance of trained film-readers interpreting
mammograms. Our aim now is to carry out the research required to take CADMIUM
to the point where a demonstrator application can be given a rigorous
evaluation.
The Candidate
The project addresses number of important scientific issues. Perhaps the most
challenging aspect of the work will be the knowledge engineering required to
integrate decision support and image processing. This will mean identifying
concepts that provide a meaningful and communicable description of the
interesting properties of radiological images and defining an interface between
this set of concepts and the image processing measures used to analyse the
digital images. This part of the project will be the central responsibility of
the successful applicant.
The post-doct{*filter*}research fellow will have the opportunity to assist in
establishing the user requirements, be primarily responsible for building the
knowledge base and also have a role in designing the demonstrator be to
implemented at the end of the project.
The successful applicant will therefore be someone who can communicate easily
with the clinicians, who has experience in knowledge representation and is
interested in developing an understanding of problems in image interpretation.
He or she will be expected to contribute to the writing of research papers and
preparing presentations. Relevant experience would include work in fields such
as Cognitive Science, Artificial Intelligence, Knowledge-Based Systems,
Decision Support or Medical Informatics. Experience in programming in Prolog
would be an advantage.
The post is funded for two years, in the first instance, and will be based at
the Whittington Campus of UCL Medical School. Interested applicants should
submit a full C.V. before July 31st.
Any enquiries or requests for further information should be addressed to:
Paul Taylor
CHIME, UCL Medical School
Whittington Hospital Campus
Highgate Hill
LONDON
N19 5NF
Tel: +171 288 3548
