VDETECT | VisionX V4 |
vdetect − detect and report potential lung nodule locations from a whole lung CT scan.
SYNOPSISvdetect [if=] [ig=] [it=] [of=] [el=] [cf=] [cs=] [-v] [-d]
DESCRIPTIONvdetect detects and reports potential lung nodule locations from a whole lung CT scan.
A multi-stage detection algorithm using a successive nodule candidate refinement approach is used to detect pulmonary nodule candidates from whole-lung CT images. First, the detection algorithm generates a set of nodule candidates. Then, the candidates are successively refined using filters of increasing complexity. Finally, remaining candidates are classified and only potential nodule locations (x, y, z) including size are output in a tab-separated output file.
Additional output options are available. The el= parameter specifies the execution log, the cf= output provies the full features for the final candidates, and the cs= option records teh candidate counts at each feature stage. For complete intermediate results specifyu the -d option.
OPTIONS|
[if=] |
original whole lung CT scan input filename. | ||
|
[ig=] |
lung parenchyma binary mask input filename. | ||
|
[of=] |
detected nodule location text file output | ||
|
[el=] |
Specify the execution log. All commnds executed will be reported in this file | ||
|
[cf=] |
Specify a candidate feature output. All the features of the final candidates will be recorded [cs=] Specify a candidate statistics file. The cadidate number at the different filter stages will be recorded. | ||
|
-v |
verbose mode to print information about intermediate results | ||
|
-d |
Debug mode: do not delete intermediate files at the end of execution. Also, force the ceration of an execution log if el= is not specified |
vloggen(1), vlogtype(1), vlogratio(1), vattratio(1), vairdist(1), vpsdist(1), vcurvature(1), vcompact(1), vsvmclassify(1)
AUTHORAndinet Asmamaw (2005) and Sergei Fotin (September, 2008)