Assessing Breast Cancer Risk with an Artificial Neural Network

Abstract

Objectives: Radiologists face uncertainty in making decisions based on their judgment of breast cancer risk.<br />Artificial intelligence and machine learning techniques have been widely applied in detection/recognition of cancer.<br />This study aimed to establish a model to aid radiologists in breast cancer risk estimation. This incorporated imaging<br />methods and fine needle aspiration biopsy (FNAB) for cyto-pathological diagnosis. Methods: An artificial neural<br />network (ANN) technique was used on a retrospectively collected dataset including mammographic results, risk<br />factors, and clinical findings to accurately predict the probability of breast cancer in individual patients. Area under<br />the receiver-operating characteristic curve (AUC), accuracy, sensitivity, specificity, and positive and negative predictive<br />values were used to evaluate discriminative performance. Result: The network incorporating the selected features<br />performed best (AUC = 0.955). Sensitivity and specificity of the ANN were respectively calculated as 0.82 and 0.90.<br />In addition, negative and positive predictive values were respectively computed as 0.90 and 0.80. Conclusion: ANN<br />has potential applications as a decision-support tool to help underperforming practitioners to improve the positive<br />predictive value of biopsy recommendations.