Automatic mass detection in mammograms using deep convolutional neural networks.

Richa Agarwal,Robert Martí,Moi Hoon Yap,Oliver Diaz,Xavier Lladó

doi:10.1117/1.jmi.6.3.031409

Richa Agarwal, Robert Martí + Show 3 more

Open Access

https://doi.org/10.1117/1.jmi.6.3.031409

Copy DOI

Abstract

With recent advances in the field of deep learning, the use of convolutional neural networks (CNNs) in medical imaging has become very encouraging. The aim of our paper is to propose a patch-based CNN method for automated mass detection in full-field digital mammograms (FFDM). In addition to evaluating CNNs pretrained with the ImageNet dataset, we investigate the use of transfer learning for a particular domain adaptation. First, the CNN is trained using a large public database of digitized mammograms (CBIS-DDSM dataset), and then the model is transferred and tested onto the smaller database of digital mammograms (INbreast dataset). We evaluate three widely used CNNs (VGG16, ResNet50, InceptionV3) and show that the InceptionV3 obtains the best performance for classifying the mass and nonmass breast region for CBIS-DDSM. We further show the benefit of domain adaptation between the CBIS-DDSM (digitized) and INbreast (digital) datasets using the InceptionV3 CNN. Mass detection evaluation follows a fivefold cross-validation strategy using free-response operating characteristic curves. Results show that the transfer learning from CBIS-DDSM obtains a substantially higher performance with the best true positive rate (TPR) of at 1.67 false positives per image (FPI), compared with transfer learning from ImageNet with TPR of at 2.1 FPI. In addition, the proposed framework improves upon mass detection results described in the literature on the INbreast database, in terms of both TPR and FPI.

Highlights

Breast cancer is the most common form of cancer in the female population
We evaluated three popular, widely used convolutional neural networks (CNNs) architectures (VGG16, ResNet[50], and InceptionV3) that have already proven to be excellent for image classifications using the ImageNet dataset, which we use for transfer learning from natural images to digitized and digital mammograms
The results indicate that, using the transfer learning between the images of similar domains (ImageNet → CBIS-DDSM → INbreast), the testing accuracy is improved to 88.86% Æ 2.96 compared with that obtained with the database of natural images (ImageNet → INbreast)

Summary

Introduction

Breast cancer is the most common form of cancer in the female population. In the USA, it is estimated that ∼12.4% of women will be diagnosed with breast cancer at some point during their lifetime.[1]. Digital breast tomosynthesis is gradually being adopted, x-ray mammography is still the gold standard imaging modality used for breast cancer screening due to its fast acquisition and cost-effectiveness. For certain population groups (young women or women with dense breasts), it has been shown to have a reduced sensitivity, which may result in more missed cancers.[2]. Research in breast image analysis has mainly focused on the development of computer-aided detection or diagnosis (CAD) systems to assist radiologists in the diagnosis. Mammography CAD systems relied on handengineered features, which showed limited accuracy in complex scenarios. With the advent of deep learning methods, CAD systems learn automatically which image features are more relevant to be used to perform a diagnosis, boosting the performance of these systems.

Objectives

Results

Discussion

Conclusion