A case-based reasoning system using feature scaling for computer aided breast cancer

Elter, M.M.ElterWittenberg, T.T.WittenbergSchulz-Wendtland, R.R.Schulz-Wendtland2022-03-042022-03-042007https://publica.fraunhofer.de/handle/publica/21507810.1007/s11548-007-0106-4In this paper a CBR system for mammography CAD that uses feature scaling to improve the systems classification performance is proposed. The CBR system was evaluated on the public DDSM mammography database. We use ROC analysis and leaving-one-out sampling to show that the feature weighting approach results in improved ROC performance.en621616A case-based reasoning system using feature scaling for computer aided breast cancerjournal article