Skip to content

A deep learning framework for lung cancer prediction built with Lasagne/Theano

License

Notifications You must be signed in to change notification settings

EliasVansteenkiste/dsb3

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 

Repository files navigation

TODO

  • filter blobs that are too close to each other (Elias done?)
  • remove strange regions from ROI (Frederic, Elias)
  • implement test_dsb script (Andreas)
  • todo solve the DSB patient with 2 series of slices: b8bb02d229361a623a4dc57aa0e5c485
  • implement fast dsb segmentation

LUNA

  • generate blobs: test_seg_scan.py
  • probabilities for blobs (fpred): test_fpred_scan.py
  • stats over segmentation blobs: evaluate_luna_seg_scan.py
  • stats over fpred: evaluate_luna_fpred_scan.py

DSB

  • generate blobs: test_seg_scan_dsb.py
  • fpred: test_fpred_scan_dsb.py
  • plot rois as in the final data iterator: plot_dsb_roi.py
  • train classifier: train_class_dsb.py

List of patients

1.3.6.1.4.1.14519.5.2.1.6279.6001.943403138251347598519939390311 (nodule on the border, bad quality) 1.3.6.1.4.1.14519.5.2.1.6279.6001.287966244644280690737019247886 (biggest nodule 32 mm) b8bb02d229361a623a4dc57aa0e5c485 (has 2 series of data) 08528b8817429d12b7ce2bf444d264f9 (half of the lung) 6a145c28d3b722643f547dfcbdf379ae (half of the lung) 5fe048f36bd2da6bdb63d8ff3c4022cd (half of the lung)

Random

ssh -X -p 10001 ikorshun@localhost scp -P 10001 /mnt/sda3/data/kaggle-lung/stage1_sample_submission.csv ikorshun@localhost://mnt/storage/data/dsb3/

About

A deep learning framework for lung cancer prediction built with Lasagne/Theano

Resources

License

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published