A hybrid approach to breast cancer diagnosis

Sordo, M, Buxton, H and Watson, D (2001) A hybrid approach to breast cancer diagnosis. In: Jain, L and Wilde, P (eds.) Practical Applications of Computational Intelligence Techniques. Kluwer Academic Publishers, pp. 299-330. ISBN 0792373200

Full text not available from this repository.
Item Type: Book Section
Additional Information: Originality: This was a novel application of KBANN technology to the analysis of 31P MR spectroscopy of breast tumours. Significance: Multidisciplinary - combining AI/computer science with clinical MR spectroscopy/patient diagnosis and care. Rigour: It was practical approach using a real clinical dataset and significant and useful results were obtained. Impact: Showed how KBANN's could be used to analyse complex clinical data, without making use of biochemical knowledge in the interpretation of the MR spectra, thus highlighting specific metabolites in the spectra that could have a specific role in the diagnosis of breast cancer. Biochemists have since used this information about new significant metabolites to support their analysis of tumour spectra.
Schools and Departments: School of Engineering and Informatics > Informatics
Depositing User: Des Watson
Date Deposited: 06 Feb 2012 20:04
Last Modified: 30 Nov 2012 17:05
URI: http://sro.sussex.ac.uk/id/eprint/23872
📧 Request an update