Hardy, Judy and Armitage, Alistair (2005) An advisory system for the treatment of kidney stones. In: Natural Computing Applications Forum, 2005 May, York. (Unpublished)

PDF Restricted to Registered users only Available under License Creative Commons Attribution Non-commercial. Download (757kB) | Request a copy

Abstract/Description

Kidney stones can be treated by a number of methods, some more invasive than others. Increasingly, Shock Wave Lithotripsy (a non-invasive technique) is used in preference to surgery. However, the kidney stones can recur. This study was done in conjunction with the Scottish Lithotripsy Centre. Approximately 1500 records detailing treatment were analysed. the intention was to see if a system could be developed to predict the outcome of treatment. Neural networks proved to be more successful than standard statistical techniques (logistic regression). The corrct outcome could be predicted 70-80% of the time. A brief comparison is made with results produced by other workers in this area. Some interesting insights into the factors affecting predicability were gained.

Actions (login required)