Skip to content

Latest commit

 

History

History
48 lines (31 loc) · 1.73 KB

File metadata and controls

48 lines (31 loc) · 1.73 KB

Breast Cancer segmentation using deep learning

The current research aims to design a deep learning model for nuclei segmentation using earlier research studies. This work employs a variety of methodologies, including pre-processing techniques on datasets, a multi-organ transfer learning method for segmentation. The various encoder-based models will be trained using a combination of the TNBC and MoNuSeg datasets but will be evaluated using the TNBC test dataset.

Project required packages

  • albumentations==1.1.0

  • imageio==2.12.0

  • matplotlib==3.5.0

  • numpy==1.21.4

  • opencv_python_headless==4.5.4.60

  • pandas==1.3.4

  • scikit_learn==1.0.2

  • scipy==1.7.3

  • seaborn==0.11.2

  • segmentation_models_pytorch==0.2.1

  • torch==1.10.0+cu113

  • tqdm==4.62.3

To train the model

git clone https://github.com/surya9teja/Breast-Cancer-Segmentation-using-Deep-Learning.git
cd Breast-Cancer-Segmentation-using-Deep-Learning
python3 data_pre_process.py
python3 main.py
# for Graphs and evaluation
python3 test.py 
python3 graphs.py 

Note: Before running the files it requires datasets from the drive and place them under Datasets/

Dataset link available at google drive

Results

Training Loss per epoch Validation Loss per epoch Evalaution Metrics