An optimum ANN-based breast cancer diagnosis: Bridging gaps between ANN learning and decision-making goals.

•We present a Life-Sensitive Self-Organizing Error Drive (LS-SOED) Artificial Neural network for breast cancer diagnosis.•The method matches the best performances in the literature in terms of accuracy.•The approach improves the quality of decision-making by minimizing misclassification costs.•Each patient is considered unique as their wrong diagnosis will lead to different misclassification costs.