Prediction Of Breast Cancer in Mammography Images Using Quantum-Based Swarm Intelligence Clustering Techniques and Deep Learning Approaches

Proceeding Paper of ICMLHEED 2023 Conference

Prediction Of Breast Cancer in Mammography Images Using Quantum-Based Swarm Intelligence Clustering Techniques and Deep Learning Approaches

Authors- Akshaya Kumar Dash, Nibedita Adhikari, Chittaranjan Mohapatra

Abstract- – Mammography is the most efficient approach for breast cancer (BC) screening. However, mammogram-based breast biopsy’s limited positive predictive value results in around 70% of needless biopsies with benign outcomes. Several computer-aided diagnostic (CAD) solutions have been proposed in recent years to help reduce the high rate of unneeded breast biopsies. The purpose of the research is to define a fuzzy c-means (FCM) approach based on quantum-inspired particle swarm optimisation (QPSO) for clustering multidimensional data. The specified QPSO technique is used to establish cluster centres for a dataset to ad-dress this disadvantage. Pre-processing, clustering, feature extraction, and classification are the four stages of this process. Initially, z-score normalization is used to perform pre-processing. Then, using the QPSO-based FCM approach, the cluster centres for a dataset are generated. Additionally, the principal component analysis (PCA) technique is suggested for extracting features from the given data. Finally, a deep learning based classifier is presented to accurately predict BC im-ages. The proposed scheme is compared to several well-established methods. Experiments demonstrate that the proposed quantum technique has an 87.72% F-measure and 89.59% accuracy compared to other approaches currently in use.

DOI: 10.61463/ijset.icmlheed-2023.102