fridolinvii / sentinel-lymph-node-fingerprinting

Sentinel lymph node fingerprinting

This is the code to the paper Sentinel lymph node fingerprinting (Physics in Medicine and Biology 2019).

Create Matrix Ao

createA/noWallsR3/saveA create matrix for R3 (non-symetric collimator. Rest is similar)

• Step 1) saveA.m
• Step 2) createA.m
• Step 3) createA_2.m

createA/optimal_design_withWallsN_vs_noWallsR3 creates table 1.

• Warning: Creating new sources will create new numbers of the table!
  • use auswerten.m to get the number for Table 1.
  • use vergleichen.m to get error100.mat
• Warning: using createDifferentSource.m will change source.mat and you get different numbers.

Evaluating the measurments

fingerprinting/R3 auswerten.m : Table 2

• for t={2,32} (we use here t=2 and t=32 instead of t=1 and t=16, to compare it better with the results from the with septum walls Collimator. At the point of measurement, it had two times the radioactive strength than during the measurment with the R3 collimator. Therefore, two compare the results the expose time of 1s with the first collimator is the same as if 2s with the R3 collimator

  • #1 error{t,1}.err
  • #2 error{t,5}.err
  • #3 error{t,8}.err
  • #4 error{t,4}.err
  • #5 error{t,7}.err
  • #6 error{t,6}.err
  • #7 error{t,9}.err

• copy the matrix Ao.mat created in createA/noWallsR3/saveA into the folder fingerprinting/R3

• execute lowpassfiltDict.m to create from Ao.mat the matrix DictLLN.mat

• execute fingerprinting.m to get the 3D contstruction from all the pattern in the folder image_bearbeitet which are saved in SOURCE_LL.mat

• execute auswerten.m to save the error in the matrix error.mat

• Note: This is done for noWallsN, R3, and withWallsN.

Data

Please contact Prof. Dr. Philippe C. Cattin ([email protected]) to access the data

The RAW data

• https://dbe-lakefs.dbe.unibas.ch/repositories/moonstar-slnf/main

Processed data

• https://dbe-lakefs.dbe.unibas.ch/repositories/moonstar-slnf/objects?ref=slnf&path= (Branch slnf)
• copy the folder image_bearbeitet into fingerprinting/R3
• cutter.m has the different parameters for the preprocessing of the data

Plotting the error

error_sym_vs_nonsym fingerprintL2_new.m creates Figure 14

• copy the matrix SOURCE_LL.mat from fingerprinting and executing auswerten.m will give you the error
• plotting the erros was hardcoded from the results from auswerten.m

Optimal Design

optimal_design Visualize Figure 5

createA/noWallsR3/optimal_design

• create values for the optimal design

Other

functionalC This creates Figure 8