Many grant cycles ago, an alert reader asked how to numerically reorder DICOM files that come from the scanner out of order. I said that I didn't know; but, not wishing to overtly display my ignorance, I kept it relegated to the comments section, away from the eye of public scrutiny.
However, another reader recently pointed out that this problem can be rectified by a script available on the UCSD website. Apparently DICOM files are generated in an out-of-order sequence by General Electric scanners (and possibly others) after an equipment upgrade. After a thorough investigation into why this was happening - why dozens, if not hundreds, of researchers were needlessly suffering from a hardware upgrade that was supposed to make their neuroimaging lives easier, not more difficult - the CEO of General Electric saw no alternative but to take action and jack up executive bonuses. We can all now rest easier.
The link to the script (which is in Perl) is here; I've also copied and pasted the script below, both to take away their Internet traffic, and to season this post with the flavor of programming rigor.
However, another reader recently pointed out that this problem can be rectified by a script available on the UCSD website. Apparently DICOM files are generated in an out-of-order sequence by General Electric scanners (and possibly others) after an equipment upgrade. After a thorough investigation into why this was happening - why dozens, if not hundreds, of researchers were needlessly suffering from a hardware upgrade that was supposed to make their neuroimaging lives easier, not more difficult - the CEO of General Electric saw no alternative but to take action and jack up executive bonuses. We can all now rest easier.
The link to the script (which is in Perl) is here; I've also copied and pasted the script below, both to take away their Internet traffic, and to season this post with the flavor of programming rigor.
#! /usr/bin/perl
{
use Shell;
use Cwd; # module for finding the current working directory
$|=1; # turn off I/O buffering
print "\n";
if ($#ARGV == -1) { # if no arguments are entered
instructions();
print "\n";
}
else { # read in the arguments
for ($j=0; $j<$#ARGV+1; $j++) {
$tempdir = $ARGV[$j];
chomp($tempdir);
if ($tempdir eq "."){
$tempdir = &cwd
}
opendir(DIR,$tempdir) or die "$tempdir does not exist or I can't open it\n"; # check the directories inputted
closedir(DIR);
@dirlist = $tempdir;
foreach my $name (@dirlist) {
&ScanDirectory($name);
print "\n";
}
}
}
sub ScanDirectory {
my ($p) = 0;
my ($workdir) = shift;
my($startdir) = &cwd; # keep track of where we began
print "Processing Directory $workdir \n";
chdir($workdir) or die "\nUnable to enter dir $workdir:$!\n";
opendir(DIR, ".") or die "\nUnable to open $workdir:$!\n";
my @names = readdir(DIR);
closedir(DIR);
$command = "mkdir backupimg";
system ($command);
foreach my $name (@names){
next if ($name eq ".");
next if ($name eq "..");
next if ($name eq "backupimg");
if (-d $name){ # is this a directory?
&ScanDirectory($name);
next;
}
#do something with file
if (grep(/\.MRDC\./, $name)){
$p = $p + 1;
$command = "cp $name backupimg/";
system ($command);
$old_name = $name;
$name =~ s/i(.*)\.MRDC\.(.*)/i\.CFMRI\.$2/;
$num = "";
$num = sprintf("%5d", $2);
$num=~ tr/ /0/;
$name = "i$num\.CFMRI\.$2";
rename("$old_name", "$name") || die "Cannot rename $old_name: $!";
}
#done
}
print " Directory $workdir has $p files processed \n"; # print size
$command = "rm -rf backupimg";
system ($command);
chdir($startdir) or die "Unable to change to dir $startdir:$!\n";
}
sub instructions {
print "This program renames and reorders the dicom files acquired on the GE scanners at UCSD - CFMRI.\n";
print "Usage: imseq [directories to convert] \n";
print "Example: imseq directory1 directory2 directory3 \n\n";
}
}
