A 3-stage classification system for predicting breast cancer diagnosis via FNA biopsy features

Abstract

Abstract Using a 3-stage classification system for predicting breast cancer diagnosis via Fine Needle Aspiration biopsy features, researchers found that if a tumour is classified as benign by the first classifier, since this prediction has 100% accuracy, yet at the discretion of a physician, without undergoing any treatment the patient may be discharged imminently. Similarly, if a tumour is classified as malignant by the second classifier, due to 100% prediction accuracy, yet again at the discretion of a physician, necessary cancer treatments may commence without further ado. If a case is classified as malignant by the first, then benign by the second classifier, a third classifier will provide the physician with a probabilistic estimate. The outcome provided by this classification system can help physicians not only make better-informed decisions about the status of a suspected breast tumour, but also take subsequent action quicker with confidence. This study may also encourage hospitals to rely more on artificial intelligence to be utilized during the diagnosing process of breast cancer tumours.