Abstract

Breast cancer analysis implies that radiologists inspect mammograms to detect suspicious breast lesions and identify mass tumors. Artificial intelligence techniques offer automatic systems for breast mass segmentation to assist radiologists in their diagnosis. With the rapid development of deep learning and its application to medical imaging challenges, UNet and its variations is one of the state-of-the-art models for medical image segmentation that showed promising performance on mammography. In this paper, we propose an architecture, called Connected-UNets, which connects two UNets using additional modified skip connections. We integrate Atrous Spatial Pyramid Pooling (ASPP) in the two standard UNets to emphasize the contextual information within the encoder–decoder network architecture. We also apply the proposed architecture on the Attention UNet (AUNet) and the Residual UNet (ResUNet). We evaluated the proposed architectures on two publically available datasets, the Curated Breast Imaging Subset of Digital Database for Screening Mammography (CBIS-DDSM) and INbreast, and additionally on a private dataset. Experiments were also conducted using additional synthetic data using the cycle-consistent Generative Adversarial Network (CycleGAN) model between two unpaired datasets to augment and enhance the images. Qualitative and quantitative results show that the proposed architecture can achieve better automatic mass segmentation with a high Dice score of 89.52%, 95.28%, and 95.88% and Intersection over Union (IoU) score of 80.02%, 91.03%, and 92.27%, respectively, on CBIS-DDSM, INbreast, and the private dataset.

Highlights

  • Breast cancer is the most common type of cancer that is leading to death among women, where 41,170 death cases were reported in the United States in 2020 and it represents a rate of 15% of estimated deaths against the other types of cancer[1]

  • All experiments using the proposed architecture models were conducted on a PC with the following specifications: Intel(R) Core (TM) i7-8700K processor with 32 GB RAM, 3.70 GHz frequency, and and Intersection over Union (IoU) losses is used as a segmentation loss function using the Dice score and IoU score between true and predicted samples, as detailed in Eq (3)

  • We introduced an architecture, called ConnectedUNets, which fully connects two single UNets using additional skip connection paths

Read more

Summary

Introduction

Breast cancer is the most common type of cancer that is leading to death among women, where 41,170 death cases were reported in the United States in 2020 and it represents a rate of 15% of estimated deaths against the other types of cancer[1]. Studies emphasized the importance of frequent mammography screening in order to reduce the mortality rate by detecting the breast tumors early before being spread to normal tissues and other healthy organs[2]. Mammograms are inspected every day by radiology experts to search for abnormal lesions and detect the location, shape and type of any suspicious regions in the breast. This process is considered crucial and requires more precision and accuracy, it remains expensive and exposed to error, due to the increasing number of daily screening mammograms[3]. An automated system can benefit from the high numbers of mammograms and handle this process automatically

Methods
Results
Conclusion
Full Text
Published version (Free)

Talk to us

Join us for a 30 min session where you can share your feedback and ask us any queries you have

Schedule a call