# Combining shallow and deep neural networks on pseudo-color enhanced images for digital breast tomosynthesis lesion classification

## TL;DR

This paper introduces a new CAD system for classifying breast lesions in tomosynthesis images using a dual neural network and pseudo-color enhancement.

## Contribution

The novel DBT Dual-Net architecture combines shallow and deep neural networks with pseudo-color enhancement and inter-slice majority voting for improved lesion classification.

## Key findings

- The proposed DBT Dual-Net outperforms existing classification approaches on a public dataset.
- Pseudo-color enhancement improves lesion visibility in DBT images.
- Inter-slice majority voting enhances classification accuracy by leveraging 3D spatial context.

## Abstract

The classification of lesion types in Digital Breast Tomosynthesis (DBT) images is crucial for the early diagnosis of breast cancer. However, the task remains challenging due to the complexity of breast tissue and the subtle nature of lesions. To alleviate radiologists’ workload, computer-aided diagnosis (CAD) systems have been developed. The breast lesion regions vary in size and complexity, which leads to performance degradation.

To tackle this problem, we propose a novel DBT Dual-Net architecture comprising two complementary neural network branches that extract both low-level and high-level features. By fusing different-level feature representations, the model can better capture subtle structure. Furthermore, we introduced a pseudo-color enhancement procedure to improve the visibility of lesions on DBT. Moreover, most existing DBT classification studies rely on two-dimensional (2D) slice-level analysis, neglecting the rich three-dimensional (3D) spatial context within DBT volumes. To address this limitation, we used majority voting for image-level classification from predictions across slices.

We evaluated our method on a public DBT dataset and compared its performance with several existing classification approaches. The results showed that our method outperforms baseline models.

The use of pseudo-color enhancement, extracting high and low-level features and inter-slice majority voting proposed method is effective for lesion classification in DBT. The code is available at https://github.com/xiaoerlaigeid/DBT-Dual-Net.