Deeply Supervised U-Net-for-Mass-Segmentation-in-Mammograms

📌 This repository contains the implementation of the paper mentioned below

Ravitha Rajalakshmi N, Vidhyapriya R, Elango N, Ramesh N. Deeply supervised U-Net for mass segmentation in digital mammograms. Int J Imaging Syst Technol. 2020; 1–13. https://doi.org/10.1002/ima.22516

📓 Implementation of the related models can also be found at 📁 Models

List of models

1. AU-Net
2. ASPP Net
3. Dense Net

🎯 Experiments were conducted using the data from CBIS-DDSM and INBreast

📌The CBIS-DDSM data that support the findings of this study are openly available in Mass-Training and Mass-Test folders at https://doi.org/10.7937/K9/TCIA.2016.7O02S9CY
📌The INBREAST data used in the study can be obtained on request to the authors of the article DOI: 10.1016/j.acra.2011.09.014, reference number

References

1. Lee RS, Gimenez F, Hoogi A, Miyake KK, Gorovoy M, Rubin DL. A curated mammography data set for use in computer-aided detection and diagnosis research. Scientific Data. 2017;4:170177
2. Moreira Inês C., Amaral Igor, Domingues Inês, Cardoso António, Cardoso Maria Jo~ao, Cardoso Jaime S. INbreast. Academic Radiology. 2012;19 (2):236–248. http://dx. doi.org/10.1016/j.acra.2011.09.014.
3. Sun H, Li C, Liu B, et al. AUNet: attention-guided denseupsampling networks for breast mass segmentation in whole mammograms. Physics in Medicine and Biology. 2020;65(5): 055005. https://doi.org/10.1088/1361-6560/ab5745.