BreastCancerProject

This is a school project to study Explainability and Augmentation for datasets in classification of medical images.

File Structure

report

This folder containst the LaTeX file used to generate the report. It also contains the bibiliography. Note that

this file is based on "Springer Nature Reference Style/Chemistry Reference Style" and requires the input files from Spring to compile correctly, and
the images are presumed to have been generated by the JupyterNotebooks and are not committed directly to git.

code/ProjectCode

The Jupyter Notebooks that generate the data for the report.

BreastCancer.ipynb generated the model
ExplainBreastCancerModel_LIME.ipynb generates Explainability masks using the LIME model. There is a _Generic file as well which is intended to use other models but it is a work in progress.
AnalysisExplainResults.ipynb generates charts for the report related to the Explainability.
AnalysisModelResults.ipynb generates charts for the report related to the Model itself.

code/Fiji

These are Fiji scripts to generate synthetic data

data

Stores some generic data but the primary data should be pulled from Kaggle.

References

See report. [1] Sarvamangala, D.R., Kulkarni, R.V.: Convolutional neural networks in medical image understanding: a survey. Evolutionary Intelligence (2022) https://doi.org/ 10.1007/s12065-020-00540-3
[2] Arevalo, J., Gonz´alez, F.A., Ramos-Poll´an, R., Oliveira, J.L., Guevara Lopez, M.A.: Representation learning for mammography mass lesion classification with convolutional neural networks. Computer Methods and Programs in Biomedicine 127, 248–257 (2016) https://doi.org/10.1016/j.cmpb.2015.12.014
[3] United States Cancer Statistics: Highlights from 2019 Incidence. https://www.cdc. gov/cancer/uscs/about/data-briefs/no29-USCS-highlights-2019-incidence.htm. Centers for Disease Control and Prevention, US Department of Health and Human Services (2022)
[4] W, G.A.-D.: Dataset of breast ultrasound images. Data in Brief,. https://www. kaggle.com/datasets/aryashah2k/breast-ultrasound-images-dataset (2018)
[5] Huang, C., Li, Y., Loy, C.C., Tang, X.: Learning deep representation for imbalanced classification. In: Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition (CVPR) (2016)
[6] Documentation, T.: ImageDataGenerator Class. Accessed on: November 10, 2023. https://www.tensorflow.org/api docs/python/tf/keras/preprocessing/ image/ImageDataGenerator 12
[7] Tan, M., Le, Q.V.: Efficientnet: Rethinking model scaling for convolutional neural networks. CoRR (2019)
[8] Colaboratory. https://colab.research.google.com. Google’s free, Jupyter-based notebook environment
[9] Saleem, R., Yuan, B., Kurugollu, F., Anjum, A., Liu, L.: Explaining deep neural networks: A survey on the global interpretation methods. Neurocomputing 513, 165–180 (2022) https://doi.org/10.1016/j.neucom.2022.09.129