You are in: Submissions > Select institution > City University, London > UOA 23 - Computer Science and Informatics > RA1, RA2 and RA5c

City University, London

UOA 23 - Computer Science and Informatics

RA1, RA2 and RA5c: Staff and output details and Category C staff circumstances

 

Alberdi, E - Category : A

Research Groups:

A - Software and systems dependability, B - Human-computer interaction

RA2 - Research outputs:

Number of outputs: 4

Output number: 1 of 4

Title Eliciting a Terminology for Mammographic Calcifications
Output type: Journal article
Journal title: Clinical Radiology
Month/year of publication: November 2002
Pagination: 1007-1013
Volume: 57(11)
ISSN: 0009-9260
DOI: 10.1053/crad.2002.1066 ?
Interdisciplinary output: Yes
Research group: A - Software and systems dependability
Co-authors: J. Fox, P. Taylor, R. Lee
Number of additional co-authors: 1
Other relevant details: This paper describes an innovative set of studies aimed at informing the knowledge base of a computerised diagnostic tool. The goal was to elicit discriminating descriptors for specific appearances (calcifications) in breast X-rays to be used in the tool. These studies were unique in their area in that the descriptors were obtained directly from expert practitioners (breast screening radiologists) in a relatively naturalistic decision-making environment, in contrast with other attempts based on text books or already existing standard terminologies. The studies also showed the limitations of such standard terminologies and text book descriptions. Journal ISI impact factor : 1.799 (2005)

Output number: 2 of 4

Title Elicitation and Representation of Expert Knowledge for Computer Aided Diagnosis in Mammography
Output type: Journal article
Journal title: Methods of Information in Medicine
Month/year of publication: June 2004
Pagination: 239-246
Volume: 43(3)
ISSN: 0026-1270
DOI: Not supplied ?
Interdisciplinary output: Yes
Research group: A - Software and systems dependability
Co-authors: P. Taylor, R. Lee
Other relevant details: This paper reports how the set of terms elicited in our studies in the “Clinical Radiology” journal was implemented in a knowledge-based computer tool for breast cancer diagnosis. The incorporation of explicit clinical knowledge is an advanced, unusual feature of the tool and we demonstrate the advantage of this approach over conventional leading tools in the market that lack that feature. Another important novelty of this work is the ecological (naturalistic) value of the advice provided, which is communicated at a level of description which is both effective and meaningful to clinical practitioners. Journal impact factor: 0.97 (2005).

Output number: 3 of 4

Title Expertise and the interpretation of computerised physiological data: Implications for the design of computerised physiological monitoring in neonatal intensive care
Output type: Journal article
Journal title: International Journal of Human-Computer Studies
Month/year of publication: September 2001
Pagination: 191-216
Volume: 55(3)
ISSN: 1071-5819
DOI: 10.1006/ijhc.2001.0477 ?
Interdisciplinary output: Yes
Research group: A - Software and systems dependability
Co-authors: J. Hunter, J-C. Becher, K. Gilhooly
Number of additional co-authors: 3
Other relevant details: In contrast with some recent approaches to HCI/Cognitive Engineering, which question the value of cognitive science for real world problems in software development, our paper showed the usefulness of psychological methods and cognitive theories (e.g. about expertise differences in medical reasoning) for guiding the design and evaluation of computerised decision support. Our multidisciplinary study contributed to a better understanding of clinicians’ use of monitoring information and computerised data, and thus provided practical advice for the development of computer devices managing complex information in neonatal intensive care. Journal impact factor : 1.348 (2005).

Output number: 4 of 4

Title Evaluating 'Human + Advisory computer' system: A case study
Output type: Conference contribution
Conference: Proceedings 18th British HCI Conference
Month/year of publication: 01/09/2004
Number of pages: 93-96
Media of output:
ISSN:
Research group: A - Software and systems dependability
Co-authors: A Povyakalo, L. Strigini, P. Ayton
Other relevant details: This paper introduces methods of statistical data analysis (novel for the HCI community) that highlight the “difficulty” of a decision as an important factor to predict the impact of computerised decision support tools on human decisions. We model the human and the machine as the redundant (and potentially diverse) components of a larger (human-computer) system. This analytic approach, in conjunction with our work on probabilistic modelling, leads to important counter-intuitive insights not hitherto contemplated in the relevant literature (HCI, medical informatics, or decision support in general), and provides useful criteria for steering the design of decision support.