CERR (pronounced 'sir') is a software platform for developing and sharing research results in radiology and radiation therapy treatment planning.
CERR is written in the widely-used Matlab language (version 7.0 or later), allowing for low-cost development of visualization and analysis tools.
CERR will import and display treatment plans from a wide variety of commercial or academic treatment planning systems (including both the RTOG format and DICOM-RT format).
CERR provides a common filetype for the creation of multi-institutional treatment plan databases for various types of research studies, including radiomics, dose-volume-outcomes analyses and IMRT treatment planning comparisons.
Use issue tracker to report issues.
Feb. 2014: We pushed a fix for an RTStructure import issue where slices were automatically filled-in for structures that were discontinuous in the axial direction (i.e. missing in-between slices). Here is more on how CERR assigns contours to a Scan slice and about the bug: CERR loops through all the scan slices and finds an appropriate RT-Structure contour to place at that location. In order to look for the RTStruct z-coordinate matching the corresponding CT slice, a radius equal to slicethickness/2 is used (which is alright). But, if a match is not found, this radius gets set to max(diff(RT_Struct_Zvals)). What this did was it automatically filled slices, that did not have RTStruct contoured, with the closest match. Thus, CERR would not allow RTStructs to be discontinuous in the transverse direction. This has been the case since very early version of CERR and looks this was done on purpose to make structures continuous in transverse direction by automatically filling the missing contours on in-between slices. But this behavior causes trouble when we have contours that are discontinuous in the transverse direction on purpose. Hence, this behavior has been removed. Note that this is a critical bug for CERR RTStruct import if structures were discontinuous in the axial direction. We working on a script to automatically flag buggy CERR plans by finding duplicate contours on consecutive slices and will let you know as soon as this is done. Until then, please use the latest code from gitHub that has the necessary fix for DICOM import.
Sep. 2013: We recently found a BUG in the PET import. This was related to using rescale slope (calibration factor) for PET within DICOM. This bug has affected previous versions of CERR. In some cases, the SUV max values of the PET image are truncated to lower numbers. The only way to fix previously imported PET data is to download the latest version of CERR and re-import the DICOM data into CERR. We apologize for any inconvenience.
Check CERR documentation on CERR Wiki
Download the latest version of MATLAB code on Github: https://github.com/adityaapte/CERR
or use command line: git clone git://github.com/adityaapte/CERR.git
Compiled CERR download: CERR-compiled.
MATLAB runtime component download: MATLAB 2015b RUNTIME Component.
Current version is 123123
Matlab version 2006A or later for non-compiled version.
1 GB RAM for a small archive. 2 GB RAM recommended.
Matlab Toolboxes are not required.
The current version of CERR has been tested under these combinations of Matlab/OS:
|Matlab 6.1||Matlab 6.5||Matlab 7.0||Matlab 7.11.0|
|Mac OS X|