Back to
Projects List
Revisit RawImageGuess extension
Key Investigators
- Attila Nagy (University of Szeged, Hungary)
- Csaba Pintér (EBATINCA, Spain)
- Andras Lasso (Queen's, Canada)
- Steve Pieper (Isomics Inc, Cambridge)
- Jean-Christophe Fillion-Robin (Kitware)
Project Description
We created this extension not too long ago.
By using this extension you can find out the parameters of work file types and formats that Slicer doesn’t handle out of the box. This way you can create an .nhdr with the parameters of the images(series), and then load them into Slicer.
Objective
Would like to collect use cases, refinement suggestions and new ideas.
Approach and Plan
Collect use cases, refinement suggestions and new ideas.
Progress and Next Steps
- Ruined the original functionality :D
- Began to implement new ideas an experiment with them.
- Tried one, but didn’t really work out as expected. It is a row continuity heuristic (super cheap, super effective)
Natural images (medical, microscopy, industrial) have strong horizontal continuity.
For a guessed X:
Interpret raw data as 1D array
Split into rows of length X
Compute:
mean(|row[i] - row[i+1]|)
Do this for several rows.
Maybe needs some refinement.
The implementation of further ideas is coming.
Illustrations

Background and References
Project week from 2018:
https://projectweek.na-mic.org/PW28_2018_GranCanaria/Projects/RawImageGuess/
And 2019:
https://projectweek.na-mic.org/PW30_2019_GranCanaria/Projects/RawImageGuess/