Breast Mass Classification in Mammograms using Ensemble Convolutional Neural Networks

Abstract

The paper presents quantitative results of a preliminary study undertaken as part of Decision Support and Information Management System for Breast Cancer (DESIREE). DESIREE is a European-funded project to improve the management of primary breast cancer through image-based, guidelinebased, experience-based, and case-based information systems. In this study we explore the use of ensemble deep learning for breast mass classification in mammograms. The proposed method is based on AlexNet with some modifications in order to adapt it to our classification problem. Subsequently, model selection is performed to select the best three results based on the highest validation accuracies during the validation phase. Finally, the prediction is based on the average probability of the models. Experimental evaluation shows that accuracy from individual models ranges between 75% and 77%, but combining the best models (ensemble networks) results in over 80% classification accuracy and aura under the curve.

Presentation of Novel Hybrid Algorithm for Detection and Classification of Breast Cancer Using Growth Region Method and Probabilistic Neural Network

“…Rampun et al [44] used a pretrained modified version of AlexNet with detailed adjustments in the database of CBIS-DDSM mammography images.…”

Multi-Tasking U-Shaped Network for Benign and Malignant Classification of Breast Masses

“…The research of BM classification of breast masses based on CNN is mainly focused on the improvement of CNN input [24], [25], CNN structure [26]- [28], and training methods.…”

“…Sarkar et al, used Leaky ReLU as the activation function and applied dropout to the fully connected layer to build a CNN model with four convolutional layers and three fully connected layers for the BM classification [27]. Rampun et al, removed the local response normalization layer on the basis of AlexNet, and added a BN layer after each convolution layer, using the PReLU activation function instead of the ReLU activation function to obtain better classification performance than the original AlexNet Effect [28].…”

Effective Breast Cancer Recognition Based on Fine-Grained Feature Selection

“…(5) Recent region of interest (ROI)-based breast cancer recognition models were used for indirect comparisons: Tsochatzidis's model [47], Rampun's model [48], AlexNet + sparse multiple instance learning [49], and Carneiro's model [50].…”