a b/README.md 

  1
  
# HistoTNet





  
  

  2
  





  
  

  3
  
Pytorch source code for the paper:





  
  

  4
  





  
  

  5
  
    A. Genovese, M. S. Hosseini, V. Piuri, K. N. Plataniotis, and F. Scotti, 





  
  

  6
  
    "Histopathological transfer learning for Acute Lymphoblastic Leukemia detection", 





  
  

  7
  
    in Proc. of the 2021 IEEE Int. Conf. on Computational Intelligence and Virtual Environments for Measurement Systems and Applications (CIVEMSA 2021), 





  
  

  8
  
    June 18-20, 2021, pp. 1-6. 





  
  

  9
  
    ISBN: 978-1-6654-1249-0. [DOI: 10.1109/CIVEMSA52099.2021.9493677]





  
  

  10
  
    





  
  

  11
  
Paper:





  
  

  12
  





  
  

  13
  
https://ieeexplore.ieee.org/document/9493677





  
  

  14
  
    





  
  

  15
  
Project page:





  
  

  16
  





  
  

  17
  
[https://iebil.di.unimi.it/cnnALL/index.htm](https://iebil.di.unimi.it/cnnALL/index.htm)





  
  

  18
  
    





  
  

  19
  
Outline:





  
  

  20
  
![Outline](https://iebil.di.unimi.it/cnnALL/imgs/outline_civemsa21all.jpg "Outline")





  
  

  21
  





  
  

  22
  
Citation:





  
  

  23
  





  
  

  24
  
    @InProceedings {civemsa21all,





  
  

  25
  
    author = {A. Genovese and M. S. Hosseini and V. Piuri and K. N. Plataniotis and F. Scotti},





  
  

  26
  
    booktitle = {Proc. of the 2021 IEEE Int. Conf. on Computational Intelligence and Virtual Environments for Measurement Systems and Applications (CIVEMSA 2021)},





  
  

  27
  
    title = {Histopathological transfer learning for Acute Lymphoblastic Leukemia detection},





  
  

  28
  
    month = {June},





  
  

  29
  
    day = {18-20},





  
  

  30
  
    year = {2021},}





  
  

  31
  





  
  

  32
  
Main files:





  
  

  33
  





  
  

  34
  
- (1) PyTorch_HistoNet/pytorch_histonet.py: training/testing of the HistoNet;





  
  

  35
  
- (2) PyTorch_HistoTNet/pytorch_histotnet.py: training/testing of the HistoTNet.





  
  

  36
  





  
  

  37
  
Instructions:





  
  

  38
  





  
  

  39
  
0) Install the required packages (see packages.txt)





  
  

  40
  





  
  

  41
  
1) cd to "(1) PyTorch_HistoNet" and run "pytorch_histonet.py" to train the HistoNet on the ADP database, implemented according to the paper:





  
  

  42
  





  
  

  43
  
    Mahdi S. Hosseini, Lyndon Chan, Gabriel Tse, Michael Tang, Jun Deng, Sajad Norouzi, Corwyn Rowsell, Konstantinos N. Plataniotis, Savvas Damaskinos





  
  

  44
  
    "Atlas of Digital Pathology: A Generalized Hierarchical Histological Tissue Type-Annotated Database for Deep Learning"





  
  

  45
  
    Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), 2019, pp. 11747-11756





  
  

  46
  
    





  
  

  47
  
    Required files:





  
  

  48
  
    





  
  

  49
  
    - (1) PyTorch_HistoNet/db_orig/ADP/img_res_1um_bicubic/ <br/>





  
  

  50
  
    ADP database, split in patches, obtained following the instructions at: <br/>





  
  

  51
  
    https://www.dsp.utoronto.ca/projects/ADP/ <br/>





  
  

  52
  
    e.g., (1) PyTorch_HistoNet/db_orig/ADP/img_res_1um_bicubic/001.png_crop_16.png





  
  

  53
  
    





  
  

  54
  
    - (1) PyTorch_HistoNet/db_orig/ADP/ADP_EncodedLabels_Release1_Flat.csv





  
  

  55
  
    file containing the labels of the ADP database, obtained following the instructions at: <br/>





  
  

  56
  
    https://www.dsp.utoronto.ca/projects/ADP/ <br/>





  
  

  57
  
    





  
  

  58
  
2) Copy the trained models in "(2) PyTorch_HistoTNet\pretrained_nets".





  
  

  59
  
For simplicity, some trained models are already present.





  
  

  60
  
    





  
  

  61
  
3) cd to "(2) PyTorch_HistoTNet" and run "pytorch_histotnet.py" to train the HistoTNet on the ALL-IDB database for Acute Lymphoblastic Leukemia detection.





  
  

  62
  
    





  
  

  63
  
    Required files:





  
  

  64
  
    





  
  

  65
  
    - (2) PyTorch_HistoTNet/db/ALL_IDB2 <br/>





  
  

  66
  
    ALL-IDB database, obtained following the instructions at:





  
  

  67
  
    https://homes.di.unimi.it/scotti/all/





  
  

  68
  
    e.g., (2) PyTorch_HistoTNet/db/ALL_IDB2/Im001_1.tif





  
  

  69
  
    





  
  

  70
  
The databases used in the paper can be obtained at:





  
  

  71
  





  
  

  72
  
- Atlas of Digital Pathology (ADP)<br/>





  
  

  73
  
https://www.dsp.utoronto.ca/projects/ADP/





  
  

  74
  





  
  

  75
  
    Mahdi S. Hosseini, Lyndon Chan, Gabriel Tse, Michael Tang, Jun Deng, Sajad Norouzi, Corwyn Rowsell, Konstantinos N. Plataniotis, Savvas Damaskinos





  
  

  76
  
    "Atlas of Digital Pathology: A Generalized Hierarchical Histological Tissue Type-Annotated Database for Deep Learning"





  
  

  77
  
    Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), 2019, pp. 11747-11756





  
  

  78
  





  
  

  79
  
- Acute Lymphoblastic Leukemia Image Database for Image Processing (ALL-IDB) <br/>





  
  

  80
  
https://homes.di.unimi.it/scotti/all/





  
  

  81
  





  
  

  82
  
    R. Donida Labati, V. Piuri, F. Scotti





  
  

  83
  
    "ALL-IDB: the acute lymphoblastic leukemia image database for image processing"





  
  

  84
  
    in Proc. of the 2011 IEEE Int. Conf. on Image Processing (ICIP 2011), 





  
  

  85
  
    Brussels, Belgium, pp. 2045-2048, September 11-14, 2011. 





  
  

  86
  
    ISBN: 978-1-4577-1302-6. [DOI: 10.1109/ICIP.2011.6115881]





  
  

  87