############################## 1. codes ##############################

1.1 Comparative experiments

a) DenseNet: Comparative1_DenseNet_NT.py

b) ResNet50: Comparative2_ResNet50_NT.py

c) MobileNet: Comparative3_MobileNet_NT.py

1.2 Our proposed deep fusion models based on VGG16: Model1 and Model2

a) VGG16 Baseline: Main1_VGG16_NT.py

b) VGG16_Fusion1: Main1_VGG16_NT_Fusion_Model1.py

c) VGG16_Fusion2: Main1_VGG16_NT_Fusion_Model2.py

1.3 Our proposed deep fusion models based on VGG19: Model1 and Model2

a) VGG19 Baseline: Main2_VGG19_NT.py

b) VGG19_Fusion1: Main2_VGG19_NT_Fusion_Model1.py

c) VGG19_Fusion2: Main2_VGG19_NT_Fusion_Model2.py

1.4 Experiment evalution

a) Optional evaluation for ROI patches:Main3_1_Evaluate_Patchwise.py

b) Evluation for ROI-wise classification: Main3_2_Evaluate_Imagewise.py

############################## 2. data ##############################

1.1 The original MIAS dataset

Please refer to the following URL for the original data provided.

http://peipa.essex.ac.uk/pix/mias/all-mias.tar.gz

https://www.repository.cam.ac.uk/handle/1810/250394?show=full

Paper for the data source:

SUCKLING J, P. (1994). The mammographic image analysis society digital mammogram database. Digital Mammo, 375-386.

1.2 The preprocessed ROIs data used in this research.

Please see the 'MIAS_ROIs' compressed file.

1.3 The MATLAB scripts for this data preprocessing.

If you want to understand the process of data processing, please refer to the MATLAB scripts below provided by us.

https://github.com/yxchspring/MIAS_Preprocess

############################## 3. Other notes ##############################

If you find this research useful, please cite our paper.

Xiangchun Yu 1, Wei Pang 2, Qing Xu 1, and Miaomiao Liang 1,, Mammographic image classification with deep fusion learning. Scientific Reports 10, 14361 (2020). https://doi.org/10.1038/s41598-020-71431-x

If you have any questions, please contact us below.

Xiangchun Yu: yuxc@jxust.edu.cn