metamri 0.1.23 → 0.2.0

data/Rakefile

@@ -1,22 +1,7 @@
-# 
-# To change this template, choose Tools | Templates
-# and open the template in the editor.
-
 require 'rubygems'
 require 'rake'
-# require 'echoe'
-# 
-# Echoe.new('metamri', '0.1.0') do |p|
-#   p.description    = "Extraction of MRI metadata and insertion into compatible sqlite3 databases."
-#   p.url            = "http://github.com/brainmap/metamri"
-#   p.author         = "Kristopher J. Kosmatka"
-#   p.email          = "kk4@medicine.wisc.edu"
-#   p.ignore_pattern = ["nbproject/*"]
-#   p.development_dependencies = []
-# end
-# 
-# Dir["#{File.dirname(__FILE__)}/tasks/*.rake"].sort.each { |ext| load ext }
-
+require 'rake/rdoctask'
+require 'rake/testtask'
 
 begin
   require 'jeweler'
@@ -35,4 +20,23 @@ begin
   Jeweler::GemcutterTasks.new
 rescue LoadError
   puts "Jeweler not available. Install it with: sudo gem install jeweler"
+end
+
+begin
+  require 'spec/rake/spectask'
+  Spec::Rake::SpecTask.new do |test|
+    test.warning = true
+    test.rcov = true
+    test.spec_files = FileList['spec/**/*_spec.rb']
+  end
+rescue LoadError
+  task :spec do
+    abort "RSpec is not available.  In order to run specs, you must: sudo gem install rspec"
+  end
+end
+
+Rake::TestTask.new do |t|
+  t.libs << "test"
+  t.test_files = FileList['test/*test*.rb']
+  t.verbose = true
 end

data/VERSION

@@ -1 +1 @@
-0.1.23
+0.2.0

data/bin/list_visit

@@ -42,14 +42,18 @@ def run!
   
   # Default to scanning the current directory if no argument was given,
   # otherwise go through each list.
-  unless ARGV.length == 1
-    input_directories = ARGV
+  unless ARGV.length >= 1
+    input_directories = [Dir.pwd]
   else
-    input_directories = ARGV[0]
+    input_directories = ARGV
   end
   
-  input_directories.each do |raw_directory|
-    list_visit raw_directory, options
+  unless input_directories.empty?
+    input_directories.each do |raw_directory|
+      list_visit raw_directory, options
+    end
+  else
+    raise IOError, "No input directories specified."
   end
   
 end
@@ -73,10 +77,11 @@ def list_visit(raw_directory, options = {})
   begin
     raise ScriptError, "Scaning filesystem directly..." if options[:force_scan]
     
-    visit = VisitRawDataDirectoryResource.find(:first, :params => {:search => {:path => raw_directory}})
+    lookup_path = File.dirname(raw_directory).split(File::Separator).last == "dicoms" ?  File.join(raw_directory, '..') : raw_directory
+    visit = VisitRawDataDirectoryResource.find(:first, :params => {:search => {:path => File.expand_path(lookup_path)}})
     
-    raise IOError.new("Could not lookup visit using path.") unless visit
-    raise IOError.new("Returned visit does not match path.") unless visit.path == raw_directory
+    raise IOError.new("Could not lookup visit using path #{lookup_path}.") unless visit
+    raise IOError.new("Returned visit's path #{visit.path} does not match path.") unless visit.path == File.expand_path(lookup_path)
   rescue ScriptError, IOError => e
     puts e unless options[:verbose] == false
     visit = VisitRawDataDirectory.new(raw_directory)
@@ -116,6 +121,10 @@ def parse_options
       options[:verbose] = true
     end
     
+    opts.on('-i', '--ignore REGEXP', Regexp, "A regular expression to ignore heavy directories.") do |regexp|
+      options[:ignore_patterns] = [regexp]
+    end
+    
     opts.on('-g', '--grep GREP', "Search series descriptions") do |grep|
       options[:grep] = grep
       options[:verbose] = false
@@ -132,5 +141,5 @@ def parse_options
 end
 
 if File.basename(__FILE__) == File.basename($PROGRAM_NAME)
-  run!
+    run!
 end

data/lib/metamri/core_additions.rb

@@ -1,4 +1,5 @@
 require 'tmpdir'
+
 class String
   # Does same basic string replacements to ensure valid filenames.
   def escape_filename
@@ -37,14 +38,14 @@ class Pathname
   
   def each_pfile(min_file_size = MIN_PFILE_SIZE)
     entries.each do |leaf|
-      next unless leaf.to_s =~ /^P.*\.7|^P.*\.7\.bz2/
+      next unless leaf.to_s =~ /^P.{5}\.7(\.bz2)/
       branch = self + leaf
       next if branch.symlink?
       if branch.size >= min_file_size
         lc = branch.local_copy
         begin
           yield lc
-        rescue Exception => e
+        rescue StandardError => e
           puts "#{e}"
         ensure
           lc.delete
@@ -56,7 +57,7 @@ class Pathname
   def first_dicom
     entries.each do |leaf|
       branch = self + leaf
-      if leaf.to_s =~ /^I\.|\.dcm(\.bz2)?$|\.[0-9]+(\.bz2)?$/
+      if leaf.to_s =~ /^I\.(\.bz2)?$|\.dcm(\.bz2)?$|[A-Za-z^P]\.[0-9]+(\.bz2)?$/
         lc = branch.local_copy
         begin
           yield lc
@@ -112,12 +113,11 @@ class Pathname
     self.to_s =~ /^\./ || self.symlink?
   end
   
-=begin
-  Creates a local, unzipped copy of a file for use in scanning.
-  Will return a pathname to the local copy if called directly, or can also be 
-  passed a block.  If it is passed a block, it will create the local copy
-  and ensure the local copy is deleted.
-=end
+
+  # Creates a local, unzipped copy of a file for use in scanning.
+  # Will return a pathname to the local copy if called directly, or can also be 
+  # passed a block.  If it is passed a block, it will create the local copy
+  # and ensure the local copy is deleted.
   def local_copy(tempdir = Dir.mktmpdir, &block)
     tfbase = self.to_s =~ /\.bz2$/ ? self.basename.to_s.chomp(".bz2") : self.basename.to_s
     tfbase.escape_filename
@@ -145,4 +145,26 @@ class Pathname
     end
   end
   
-end
+end
+
+# =begin rdoc
+# Monkey-patch Float to avoid rounding errors.
+# For more in-depth discussion, see: http://www.ruby-forum.com/topic/179361
+# Currently not in use.
+# =end
+# class Float
+#   def <=> other
+#     puts epsilon = self * 1e-14
+#     diff = self - other
+#     # return -1 if diff > epsilon
+#     # return 1 if diff < -epsilon
+#     0
+#   end
+# 
+#   def == other  #because built-in Float#== bypasses <=>
+#     (self<=>other) == 0
+#     true
+#   end
+# end
+
+

data/lib/metamri/image_dataset_quality_check_resource.rb

@@ -0,0 +1,26 @@
+# An ImageQualityCheckResource is a ruby object that represents a quality check
+# pulled down from the DataPanda database.  It contains notes on the quality of
+# specific images, with a note for each possible problem, and an overall note.
+# 
+#  Omnibus f:  NA –
+#  User: erik
+#  Motion warning: NA –
+#  Ghosting wrapping:  Pass –
+#  Nos concerns: NA –
+#  Image dataset:  12136
+#  Banding:  Pass –
+#  Fov cutoff: Pass –
+#  Registration risk:  Pass –
+#  Field inhomogeneity:  Mild – eh - it's ok.
+#  Other issues: la la la
+#  Incomplete series:  Complete –
+#  Spm mask: NA –
+#  Garbled series: Pass –
+# 
+# Check the current schema.db file for all available fields.
+class ImageDatasetQualityCheckResource < ActiveResource::Base
+  self.site = VisitRawDataDirectory::DATAPANDA_SERVER
+  self.element_name = "image_dataset_quality_check"
+  
+  
+end

data/lib/metamri/raw_image_dataset.rb

@@ -3,11 +3,10 @@ require 'sqlite3'
 require 'ftools'
 require 'metamri/nifti_builder'
 
-=begin rdoc
-A #Dataset defines a single 3D or 4D image, i.e. either a volume or a time series
-of volumes.  This encapsulation will provide easy manipulation of groups of raw
-image files including basic reconstruction.
-=end
+
+# A #RawImageDataset defines a single 3D or 4D image, i.e. either a volume or a time series
+# of volumes.  This encapsulation will provide easy manipulation of groups of raw
+# image files including basic reconstruction.
 class RawImageDataset
 
   # The directory that contains all the raw images and related files that make up
@@ -31,16 +30,26 @@ class RawImageDataset
   attr_reader :scanner_source
   # A #RawImageDatasetThumbnail object that composes the thumbnail for the dataset.
   attr_reader :thumbnail
-
-=begin rdoc
-  * dir: The directory containing the files.
-  * files: An array of #RawImageFile objects that compose the complete data set.
+  # A Description of the Study as listed in the DICOM Header
+  attr_reader :study_description
+  # A Description of the Protocol as listed in the DICOM Header
+  attr_reader :protocol_name
+  # Scan Tech Initials
+  attr_reader :operator_name
+  # Patient "Name", usually StudyID or ENUM
+  attr_reader :patient_name
+  # DICOM Series UID
+  attr_reader :dicom_series_uid
+  # DICOM Study UID
+  attr_reader :dicom_study_uid
   
-  Initialization raises errors in several cases:
-  * directory doesn't exist => IOError
-  * any of the raw image files is not actually a RawImageFile => IndexError
-  * series description, rmr number, or timestamp cannot be extracted from the first RawImageFile => IndexError
-=end
+  # * dir: The directory containing the files.
+  # * files: An array of #RawImageFile objects that compose the complete data set.
+  # 
+  # Initialization raises errors in several cases:
+  # * directory doesn't exist => IOError
+  # * any of the raw image files is not actually a RawImageFile => IndexError
+  # * series description, rmr number, or timestamp cannot be extracted from the first RawImageFile => IndexError
   def initialize(directory, raw_image_files)    
     @directory = File.expand_path(directory)
     raise(IOError, "#{@directory} not found.") if not File.directory?(@directory)
@@ -55,7 +64,7 @@ class RawImageDataset
     @raw_image_files = raw_image_files
     
     @series_description = @raw_image_files.first.series_description
-    raise(IndexError, "No series description found") if @series_description.nil?
+    validates_metainfo_for :series_description, :msg => "No series description found"
     
     @rmr_number = @raw_image_files.first.rmr_number
     raise(IndexError, "No rmr found") if @rmr_number.nil?
@@ -74,15 +83,34 @@ class RawImageDataset
     @study_id = @raw_image_files.first.study_id.nil? ? nil : @raw_image_files.first.study_id
     # raise(IndexError, "No study id / exam number found") if @study_id.nil?
     
+    @study_description = @raw_image_files.first.study_description
+    validates_metainfo_for :study_description, :msg => "No study description found" if dicom?
+    
+    @protocol_name = @raw_image_files.first.protocol_name
+    validates_metainfo_for :protocol_name, :msg => "No protocol name found" if dicom?
+    
+    @operator_name = @raw_image_files.first.operator_name
+    validates_metainfo_for :operator_name if dicom?
+    
+    @patient_name = @raw_image_files.first.patient_name
+    validates_metainfo_for :patient_name if dicom?
+    
+    @dicom_series_uid = @raw_image_files.first.dicom_series_uid
+    validates_metainfo_for :dicom_series_uid if dicom?
+    
+    @dicom_study_uid = @raw_image_files.first.dicom_study_uid
+    validates_metainfo_for :dicom_study_uid if dicom?        
+        
     $LOG ||= Logger.new(STDOUT)
   end
+  
 
-=begin rdoc
-Generates an SQL insert statement for this dataset that can be used to populate
-the Johnson Lab rails TransferScans application database backend.  The motivation
-for this is that many dataset inserts can be collected into one db transaction
-at the visit level, or even higher when doing a whole file system scan.
-=end
+
+  # Generates an SQL insert statement for this dataset that can be used to
+  # populate the Johnson Lab rails TransferScans application database backend. The
+  # motivation for this is that many dataset inserts can be collected into one db
+  # transaction at the visit level, or even higher when doing a whole file system
+  # scan.
   def db_insert(visit_id)
     "INSERT INTO image_datasets
     (rmr, series_description, path, timestamp, created_at, updated_at, visit_id, 
@@ -191,15 +219,15 @@ Returns a path to the created dataset as a string if successful.
     return nifti_conversion_command, nifti_output_file
   end
 
-=begin rdoc
-Returns a globbing wildcard that is used by to3D to gather files for
-reconstruction.  If no compatible glob is found for the data set, nil is returned.
-This is always the case for pfiles. For example if the first file in a data set is I.001, then:
-<tt>dataset.glob</tt>
-<tt>=> "I.*"</tt>
-including the quotes, which are necessary becuase some data sets (functional dicoms)
-have more component files than shell commands can handle.
-=end
+
+  # Returns a globbing wildcard that is used by to3D to gather files for
+  # reconstruction. If no compatible glob is found for the data set, nil is
+  # returned. This is always the case for pfiles. For example if the first file in
+  # a data set is I.001, then:
+  # <tt>dataset.glob</tt>
+  # <tt>=> "I.*"</tt>
+  # including the quotes, which are necessary becuase some data sets (functional dicoms)
+  # have more component files than shell commands can handle.
   def glob
     case @raw_image_files.first.filename
     when /^E.*dcm$/
@@ -273,12 +301,18 @@ have more component files than shell commands can handle.
     return relative_dataset_path
   end
   
-  # Reports series details, including description and possilby image quality
-  # check comments.  
+  # Reports series details, including description and possibly image quality
+  # check comments for #RawImageDatasetResource objects.  
   def series_details
     @series_description
   end
   
+  # Helper predicate method to check whether the dataset is a DICOM dataset or not.
+  # This just sends dicom? to the first raw file in the dataset.
+  def dicom?
+    @raw_image_files.first.dicom?
+  end
+  
 private
 
   # Gets the earliest timestamp among the raw image files in this dataset.
@@ -291,5 +325,15 @@ private
     File.basename(@directory)
   end
 
+  # Ensure that metadata is present in instance variables.
+  # validates_metainfo_for :study_description, :msg => "No study description found" 
+  def validates_metainfo_for(info_variable, options = {})
+    raise StandardError, "#{info_variable} must be a symbol" unless info_variable.kind_of? Symbol
+    if self.instance_variable_get("@" + info_variable.to_s).nil?
+      raise IndexError, options[:msg] ||= "Couldn't find #{info_variable.to_s}"
+    end
+  end
+
+
 end
 #### END OF CLASS ####

data/lib/metamri/raw_image_dataset_resource.rb

@@ -70,7 +70,7 @@ class RawImageDatasetResource < ActiveResource::Base
   # end
   
   def pfile?
-    scanned_file =~ /^P.*.7$/
+    scanned_file =~ /^P.{5}.7$/
   end
   
   
@@ -98,6 +98,10 @@ class RawImageDatasetResource < ActiveResource::Base
     return relative_dataset_path
   end
   
+  def image_quality_checks
+    
+  end
+  
   # Creates an Hirb Table for pretty output of dataset info.
   # It takes an array of either RawImageDatasets or RawImageDatasetResources
   def self.to_table(datasets)

data/lib/metamri/raw_image_dataset_thumbnail.rb

@@ -20,6 +20,7 @@ class RawImageDatasetThumbnail
   # The processor for creating the thumbnail (:rubydicom or :slicer)
   attr_reader :processor
 
+  # Creates a RawImageDatasetThumbnail instance by passing in a parent dataset to thumbnail.
   def initialize(dataset)
     if dataset.class == RawImageDataset
       @dataset = dataset
@@ -37,14 +38,32 @@ class RawImageDatasetThumbnail
   # Raises a StandardError if the format is incorrect (i.e. P-file instead of DICOM)
   # 
   # Be sure your filename is a valid unix filename - no spaces.
-  # Sets the @path instance variable and returns the full filename to the thumbnail.
-  #
-  # Pass in either a absolute or relative path or filename for the output image,
-  # and an options hash to manually specify the processor (Ruby Dicom or FSL Slicer).
-  #   {:processor => :rubydicom or :slicer}
+  # 
+  # Returns the full absolute filename to the new thumbnail image and sets it to @path instance variable.
+  # 
+  # === Parameters
+  # 
+  # * <tt>output</tt>: An optional string which specifies a directory or filename for the thumbnail image.
+  # * <tt>options</tt>: A hash of additional options.
+  # 
+  # === Options
+  # 
+  # * <tt>:processor</tt> -- Symbol. Specifies which thumbnail processor to use.  Defaults to :rubydicom, alternatively it could be :slicer
+  # 
+  # === Examples
+  # 
+  #  # Load a RawImageDataset 
+  #  ds = RawImageDataset('s01_assetcal', RawImageFile.new('./s01_assetcal/I0001.dcm'))
+  #  # Create a RawImageDatasetThumbnail instance
+  #  thumb = RawImageDatasetThumbnail.new(ds)
+  #  # Create a thumbnail in a temp directory without options, save it to a destination image, or force it to use FSL Slicer.
+  #  thumb.create_thumbnail
+  #  thumb.create_thumbnail('/tmp/asset_cal.png')
+  #  thumb.create_thumbnail('/tmp/asset_cal.png', :processor => :slicer)
   # 
   def create_thumbnail(output = nil, options = {:processor => :rubydicom})
     raise StandardError, "Thumbnail available only for DICOM format." unless dataset.raw_image_files.first.dicom?
+    raise ArgumentError, "Invalid :processor option #{options[:processor]}" unless VALID_PROCESSORS.include?(options[:processor])
     if output
       if File.directory?(output)
         # output is a directory.  Set the output directory but leave filepath nil.
@@ -60,7 +79,7 @@ class RawImageDatasetThumbnail
     end
     
     @processor = options[:processor]
-        
+    
     # Set a default filepath unless one was explicitly passed in.
     default_name = @dataset.series_description.escape_filename
     filepath ||= File.join(output_directory, default_name + '.png')

data/lib/metamri/raw_image_file.rb

@@ -4,21 +4,23 @@ require 'yaml';
 require 'sqlite3';
 require 'dicom'
 
-=begin rdoc
-Implements a collection of metadata associated with a raw image file.  In
-this case, by image we mean one single file.  For the case of Pfiles one file
-corresponds to a complete 4D data set.  For dicoms one file corresponds to a single
-2D slice, many of which are assembled later during reconstruction to create a
-4D data set.  The motivation for this class is to provide access to the metadata
-stored in image file headers so that they can be later reconstructed into nifti
-data sets.
-=end
+
+# Implements a collection of metadata associated with a raw image file.  In
+# this case, by image we mean one single file.  For the case of Pfiles one file
+# corresponds to a complete 4D data set.  For dicoms one file corresponds to a single
+# 2D slice, many of which are assembled later during reconstruction to create a
+# 4D data set.  The motivation for this class is to provide access to the metadata
+# stored in image file headers so that they can be later reconstructed into nifti
+# data sets.
+# 
+# Primarily used to instantiate a #RawImageDataset
 class RawImageFile
   #:stopdoc:
   MIN_HDR_LENGTH = 400
   DICOM_HDR = "dicom_hdr"
   RDGEHDR = "rdgehdr"
   RUBYDICOM_HDR = "rubydicom"
+  VALID_HEADERS = [DICOM_HDR, RDGEHDR, RUBYDICOM_HDR]
   MONTHS = {
     :jan => "01", :feb => "02", :mar => "03", :apr => "04", :may => "05", 
     :jun => "06", :jul => "07", :aug => "08", :sep => "09", :oct => "10", 
@@ -43,6 +45,10 @@ class RawImageFile
   # A short string describing the acquisition sequence. These come from the scanner.
   # code and are used to initialise SeriesDescription objects to find related attributes.
   attr_reader :series_description
+  # A short string describing the study sequence. These come from the scanner.
+  attr_reader :study_description
+  # A short string describing the study protocol. These come from the scanner.
+  attr_reader :protocol_name
   # M or F.
   attr_reader :gender
   # Number of slices in the data set that includes this file, used by AFNI for reconstruction.
@@ -65,23 +71,25 @@ class RawImageFile
   attr_reader :warnings
   # Serialized RubyDicomHeader Object (for DICOMs only)
   attr_reader :dicom_header
-  # DICOM Sequence UID
-  attr_reader :dicom_sequence_uid
+  # Hash of all DICOM Tags including their Names and Values (See #dicom_taghash for more information on the structure)
+  attr_reader :dicom_taghash
   # DICOM Series UID
   attr_reader :dicom_series_uid
   # DICOM Study UID
   attr_reader :dicom_study_uid
-
-=begin rdoc
-Creates a new instance of the class given a path to a valid image file.
-
-Throws IOError if the file given is not found or if the available header reading
-utilities cannot read the image header.  Also raises IOError if any of the
-attributes cannot be found in the header.  Be aware that the filename used to
-initialize your instance is used to set the "file" attribute.  If you need to
-unzip a file to a temporary location, be sure to keep the same filename for the
-temporary file.
-=end
+  # Scan Tech Initials
+  attr_reader :operator_name
+  # Patient "Name", usually StudyID or ENUM
+  attr_reader :patient_name
+  
+  # Creates a new instance of the class given a path to a valid image file.
+  # 
+  # Throws IOError if the file given is not found or if the available header reading
+  # utilities cannot read the image header.  Also raises IOError if any of the
+  # attributes cannot be found in the header.  Be aware that the filename used to
+  # initialize your instance is used to set the "file" attribute.  If you need to
+  # unzip a file to a temporary location, be sure to keep the same filename for the
+  # temporary file.
   def initialize(pathtofile)
     # raise an error if the file doesn't exist
     absfilepath = File.expand_path(pathtofile)
@@ -93,7 +101,7 @@ temporary file.
     begin
       @hdr_data, @hdr_reader = read_header(absfilepath)
     rescue Exception => e
-      raise(IOError, "Header not readable for file #{@filename}. #{e}")
+      raise(IOError, "Header not readable for file #{@filename} using #{@current_hdr_reader ? @current_hdr_reader : "unknown header reader."}. #{e}")
     end
     
     # file type is based on file name but only if the header was read successfully
@@ -103,10 +111,10 @@ temporary file.
     # are not found
     begin
       import_hdr
-    rescue ScriptError => e
-      raise ScriptError, "Could not find required DICOM Header Meta Element: #{e}"
-    rescue Exception => e
-      raise IOError, "Header import failed for file #{@filename}.  #{e}"
+    rescue ScriptError, NoMethodError => e
+      raise IOError, "Could not find required DICOM Header Meta Element: #{e}"
+    rescue StandardError => e
+      raise e, "Header import failed for file #{@filename}.  #{e}"
     end
     
     # deallocate the header data to save memory space.
@@ -114,37 +122,31 @@ temporary file.
   end
   
 
-=begin rdoc
-Predicate method that tells whether or not the file is actually an image.  This
-judgement is based on whether one of the available header reading utilities can
-actually read the header information.
-=end
+
+  # Predicate method that tells whether or not the file is actually an image.  This
+  # judgement is based on whether one of the available header reading utilities can
+  # actually read the header information.
   def image?
-    return ( @hdr_reader == RDGEHDR or @hdr_reader == DICOM_HDR )
+    return ( VALID_HEADERS.include? @hdr_reader )
   end
   
   
-=begin rdoc
-Predicate simply returns true if "pfile" is stored in the @img_type instance variable.
-=end
+
+  # Predicate simply returns true if "pfile" is stored in the @img_type instance variable.
   def pfile?
     return @file_type == "pfile"
   end
 
 
-=begin rdoc
-Predicate simply returns true if "dicom" is stored in the img_type instance variable.
-=end
+  # Predicate simply returns true if "dicom" is stored in the img_type instance variable.
   def dicom?
     return @file_type == "dicom"
   end
   
   
-=begin rdoc
-Returns a yaml string based on a subset of the attributes.  Specifically,
-the @hdr_data is not included.  This is used to generate .yaml files that are 
-placed in image directories for later scanning by YamlScanner.
-=end
+  # Returns a yaml string based on a subset of the attributes.  Specifically,
+  # the @hdr_data is not included.  This is used to generate .yaml files that are 
+  # placed in image directories for later scanning by YamlScanner.
   def to_yaml
     yamlhash = {}
     instance_variables.each do |var|
@@ -154,11 +156,9 @@ placed in image directories for later scanning by YamlScanner.
   end
  
    
-=begin rdoc
-Returns the internal, parsed data fields in an array. This is used when scanning
-dicom slices, to compare each dicom slice in a folder and make sure they all hold the
-same data.
-=end
+  # Returns the internal, parsed data fields in an array. This is used when scanning
+  # dicom slices, to compare each dicom slice in a folder and make sure they all hold the
+  # same data.
   def to_array
     return [@filename,
     @timestamp,
@@ -173,11 +173,9 @@ same data.
     @acquisition_matrix_y]
   end
   
-=begin rdoc
-Returns an SQL statement to insert this image into the raw_images table of a 
-compatible database (sqlite3).  This is intended for inserting into the rails
-backend database.
-=end
+  # Returns an SQL statement to insert this image into the raw_images table of a 
+  # compatible database (sqlite3).  This is intended for inserting into the rails
+  # backend database.
   def db_insert(image_dataset_id)
     "INSERT INTO raw_image_files
     (filename, header_reader, file_type, timestamp, source, rmr_number, series_description, 
@@ -189,27 +187,21 @@ backend database.
     #{@bold_reps}, '#{DateTime.now}', '#{DateTime.now}', #{image_dataset_id})"
   end
 
-=begin rdoc
-Returns an SQL statement to select this image file row from the raw_image_files table
-of a compatible database.
-=end
+  # Returns an SQL statement to select this image file row from the raw_image_files table
+  # of a compatible database.
   def db_fetch
     "SELECT *" + from_table_where + sql_match_conditions
   end
   
-=begin rdoc
-Returns and SQL statement to remove this image file from the raw_image_files table
-of a compatible database.
-=end
+  # Returns and SQL statement to remove this image file from the raw_image_files table
+  # of a compatible database.
   def db_remove
     "DELETE" + from_table_where + sql_match_conditions
   end
   
   
-=begin rdoc
-Uses the db_insert method to actually perform the database insert using the 
-specified database file.
-=end  
+  # Uses the db_insert method to actually perform the database insert using the 
+  # specified database file.
   def db_insert!( db_file )
     db = SQLite3::Database.new( db_file )
     db.transaction do |database|
@@ -221,20 +213,16 @@ specified database file.
     db.close
   end
 
-=begin rdoc
-Removes this instance from the raw_image_files table of the specified database.
-=end
+  # Removes this instance from the raw_image_files table of the specified database.
   def db_remove!( db_file )
     db = SQLite3::Database.new( db_file )
     db.execute( db_remove )
     db.close
   end
   
-=begin rdoc
-Finds the row in the raw_image_files table of the given db file that matches this object.
-ORM is based on combination of rmr_number, timestamp, and filename.  The row is returned 
-as an array of values (see 'sqlite3' gem docs).
-=end
+  # Finds the row in the raw_image_files table of the given db file that matches this object.
+  # ORM is based on combination of rmr_number, timestamp, and filename.  The row is returned 
+  # as an array of values (see 'sqlite3' gem docs).
   def db_fetch!( db_file )
     db = SQLite3::Database.new( db_file )
     db_row = db.execute( db_fetch )
@@ -257,40 +245,54 @@ private
     "rmr_number = '#{@rmr_number}' AND timestamp = '#{@timestamp.to_s}' AND filename = '#{@filename}'"
   end
 
-=begin rdoc
-Reads the file header using one of the available header reading utilities. 
-Returns both the header data as either a RubyDicom object or one big string, and the name of the utility 
-used to read it.
-
-Note: The rdgehdr is a binary file; the correct version for your architecture must be installed in the path.
-=end
+  # Reads the file header using one of the available header reading utilities. 
+  # Returns both the header data as either a RubyDicom object or one big string, and the name of the utility 
+  # used to read it.
+  # 
+  # Note: The rdgehdr is a binary file; the correct version for your architecture must be installed in the path.
   def read_header(absfilepath)
-    # header = DICOM::DObject.new(absfilepath)
-    # return [header, RUBYDICOM_HDR] if defined? header.read_success && header.read_success
-    
-    header = `#{DICOM_HDR} '#{absfilepath}' 2> /dev/null`
-    #header = `#{DICOM_HDR} #{absfilepath}`
-    if ( header.index("ERROR") == nil and 
-         header.chomp != "" and 
-         header.length > MIN_HDR_LENGTH )
-      return [ header, DICOM_HDR ]
-    end
-    header = `#{RDGEHDR} '#{absfilepath}' 2> /dev/null`
-    #header = `#{RDGEHDR} #{absfilepath}`
-    if ( header.chomp != "" and
-         header.length > MIN_HDR_LENGTH )
-      return [ header, RDGEHDR ]
+
+    case File.basename(absfilepath)
+    when /^P.{5}\.7$|^I\..{3}/
+      # Try reading Pfiles or Genesis I-Files with GE's rdgehdr
+      @current_hdr_reader = RDGEHDR
+      header = `#{RDGEHDR} '#{absfilepath}' 2> /dev/null`
+      #header = `#{RDGEHDR} #{absfilepath}`
+      if ( header.chomp != "" and
+           header.length > MIN_HDR_LENGTH )
+        @current_hdr_reader = nil
+        return [ header, RDGEHDR ]
+      end
+    else
+      # Try reading with RubyDICOM
+      @current_hdr_reader = RUBYDICOM_HDR
+      header = DICOM::DObject.new(absfilepath)
+      if defined? header.read_success && header.read_success
+        @current_hdr_reader = nil
+        return [header, RUBYDICOM_HDR] 
+      end
+      
+      # Try reading with AFNI's dicom_hdr
+      @current_hdr_reader = DICOM_HDR
+      header = `#{DICOM_HDR} '#{absfilepath}' 2> /dev/null`
+      #header = `#{DICOM_HDR} #{absfilepath}`
+      if ( header.index("ERROR") == nil and 
+           header.chomp != "" and 
+           header.length > MIN_HDR_LENGTH )
+        @current_hdr_reader = nil
+        return [ header, DICOM_HDR ]
+      end
     end
+
+    @current_hdr_reader = nil
     return [ nil, nil ]
   end
 
 
-=begin rdoc
-Returns a string that indicates the file type.  This is difficult because dicom
-files have no consistent naming conventions/suffixes.  Here we chose to call a
-file a "pfile" if it is an image and the file name is of the form P*.7
-All other images are called "dicom".
-=end
+  # Returns a string that indicates the file type.  This is difficult because dicom
+  # files have no consistent naming conventions/suffixes.  Here we chose to call a
+  # file a "pfile" if it is an image and the file name is of the form P*.7
+  # All other images are called "dicom".
   def determine_file_type
     return "pfile" if image? and (@filename =~ /^P.....\.7/) != nil
     return "dicom" if image? and (@filename =~ /^P.....\.7/) == nil
@@ -298,10 +300,8 @@ All other images are called "dicom".
   end
 
 
-=begin rdoc
-Parses the header data and extracts a collection of instance variables.  If 
-@hdr_data and @hdr_reader are not already availables, this function does nothing.
-=end
+  # Parses the header data and extracts a collection of instance variables.  If 
+  # @hdr_data and @hdr_reader are not already available, this function does nothing.
   def import_hdr
     raise(IndexError, "No Header Data Available.") if @hdr_data == nil
     case @hdr_reader
@@ -312,17 +312,142 @@ Parses the header data and extracts a collection of instance variables.  If
   end
 
 
-=begin rdoc
-Extract a collection of metadata from @hdr_data retrieved using RubyDicom
-=end
-def rubydicom_hdr_import
+  # Extract a collection of metadata from @hdr_data retrieved using RubyDicom
+  # 
+  #   Here are some example DICOM Tags and Values
+  #   0008,0022 Acquisition Date                     DA      8 20101103
+  #   0008,0030 Study Time                           TM      6 101538
+  #   0008,0080 Institution Name                     LO      4 Institution
+  #   0008,1010 Station Name                         SH      8 Station
+  #   0008,1030 Study Description                    LO     12 PILOT Study
+  #   0008,103E Series Description                   LO     12 3pl loc FGRE
+  #   0008,1070 Operators' Name                      PN      2 SP
+  #   0008,1090 Manufacturer's Model Name            LO     16 DISCOVERY MR750
+  #   0010,0010 Patient's Name                       PN     12 mosPilot
+  #   0010,0020 Patient ID                           LO     12 RMREKKPilot
+  #   0010,0040 Patient's Sex                        CS      2 F
+  #   0010,1010 Patient's Age                        AS      4 027Y
+  #   0010,1030 Patient's Weight                     DS      4 49.9
+  #   0018,0023 MR Acquisition Type                  CS      2 2D
+  #   0018,0050 Slice Thickness                      DS      2 10
+  #   0018,0080 Repetition Time                      DS      6 5.032
+  #   0018,0081 Echo Time                            DS      6 1.396
+  #   0018,0082 Inversion Time                       DS      2 0
+  #   0018,0083 Number of Averages                   DS      2 1
+  #   0018,0087 Magnetic Field Strength              DS      2 3
+  #   0018,0088 Spacing Between Slices               DS      4 12.5
+  #   0018,0091 Echo Train Length                    IS      2 1
+  #   0018,0093 Percent Sampling                     DS      4 100
+  #   0018,0094 Percent Phase Field of View          DS      4 100
+  #   0018,0095 Pixel Bandwidth                      DS      8 244.141
+  #   0018,1000 Device Serial Number                 LO     16 0000006080000
+  #   0018,1020 Software Version(s)                  LO     42 21\LX\MR Software release:20..
+  #   0018,1030 Protocol Name                        LO     22 MOSAIC Pilot 02Nov2010
+  #   0018,1100 Reconstruction Diameter              DS      4 240
+  #   0018,1250 Receive Coil Name                    SH      8 8HRBRAIN
+  #   0018,1310 Acquisition Matrix                   US      8 0\256\128\0
+  #   0018,1312 In-plane Phase Encoding Direction    CS      4 ROW
+  #   0018,1314 Flip Angle                           DS      2 30
+  #   0018,1315 Variable Flip Angle Flag             CS      2 N
+  #   0018,1316 SAR                                  DS      8 0.498088
+  #   0020,000D Study Instance UID                   UI     52 1.2.840.113619.6.260.4.88937..
+  #   0020,000E Series Instance UID                  UI     54 1.2.840.113619.2.260.6945.23..
+  #   0020,0010 Study ID                             SH      4 1260
+  #   0020,0011 Series Number                        IS      2 1
+  #   0020,0012 Acquisition Number                   IS      2 1
+  #   0020,0013 Instance Number                      IS      2 1
+  #   0020,0032 Image Position (Patient)             DS     22 -119.531\-159.531\-25
+  #   0020,1002 Images in Acquisition                IS      2 15
+  #   0028,0010 Rows                                 US      2 256
+  #   0028,0011 Columns                              US      2 256
+  #   0028,0030 Pixel Spacing                        DS     14 0.9375\0.9375
+  def rubydicom_hdr_import
+    dicom_tag_attributes = {
+      :source => "0008,0080",
+      :series_description => "0008,103E",
+      :study_description => "0008,1030",
+      :operator_name => "0008,1070",
+      :patient_name => "0010,0010",
+      :rmr_number => "0010,0020",
+      :gender => "0010,0040",
+      :slice_thickness => "0018,0050",
+      :reconstruction_diameter => "0018,1100",
+      :rep_time => "0018,0080",
+      :pixel_spacing => "0028,0030",
+      :flip_angle => "0018,1314",
+      :field_strength => "0018,0087",
+      :slice_spacing => "0018,0088",
+      :software_version => "0018,1020",
+      :protocol_name => "0018,1030",
+      :bold_reps => "0020,0105",
+      :dicom_series_uid => "0020,000E",
+      :dicom_study_uid => "0020,000D",
+      :study_id => "0020,0010",
+      :num_slices => "0020,1002",
+      :acquisition_matrix_x => "0028,0010",
+      :acquisition_matrix_y => "0028,0011"
+    }
+
+    
+    dicom_tag_attributes.each_pair do |name, tag|
+      begin
+        # next if tag_hash[:type] == :datetime
+        value = @hdr_data[tag].value if @hdr_data[tag]
+        raise ScriptError, "No match found for #{name}" unless value
+        instance_variable_set("@#{name.to_s}", value)
+      rescue ScriptError => e
+        @warnings << "Tag #{name} could not be found."
+      end
+    end
+    
+    @timestamp = DateTime.parse(@hdr_data["0008,0022"].value + @hdr_data["0008,0030"].value)
+    @dicom_taghash = create_dicom_taghash(@hdr_data)
+    # @dicom_header = remove_long_dicom_elements(@hdr_data)
+
+  end
   
-end
+  # # Remove long data elements from a rubydicom header.  This essentially strips 
+  # # lengthy image data.
+  # def remove_long_dicom_elements(header)
+  #   raise ScriptError, "A DICOM::DObject instance is required" unless header.kind_of? DICOM::DObject
+  #   h = header.dup
+  #   h.children.select { |element| element.length > 100 }.each do |e|
+  #     h.remove(e.tag)
+  #     # puts "Removing #{e.tag}..."
+  #   end
+  #   return h
+  # end
+  
+  # Create a super-lightweight representation of the DICOM header as a hash, where
+  # the tags are they keys and name and value are stored as an attribute hash in value.
+  # 
+  # Creates a hash like:
+  #  {"0018,0095"=>{:value=>"244.141", :name=>"Pixel Bandwidth"},
+  #   "0008,1030"=>{:value=>"MOSAIC PILOT", :name=>"Study Description"} }
+  # 
+  # When serialized with yaml, this looks like:
+  # 
+  #  0018,0095: 
+  #    :value: "244.141"
+  #    :name: Pixel Bandwidth
+  #   
+  #  0008,1030: 
+  #    :value: MOSAIC PILOT
+  #    :name: Study Description
+  # 
+  # To filter and search, you can do something like:
+  # tag_hash.each_pair {|tag, attributes| puts tag, attributes[:value] if attributes[:name] =~ /Description/i }
+  def create_dicom_taghash(header)
+    raise ScriptError, "A DICOM::DObject instance is required" unless header.kind_of? DICOM::DObject
+    h = Hash.new
+    header.children.each do |element|
+      h[element.tag] = {:value => element.instance_variable_get(:@value), :name => element.name}
+    end
+    return h
+  end
 
-=begin rdoc
-Extracts a collection of metadata from @hdr_data retrieved using the dicom_hdr
-utility.  
-=end
+  # Extracts a collection of metadata from @hdr_data retrieved using the dicom_hdr
+  # utility.  
   def dicom_hdr_import
     dicom_tag_templates = {}
     dicom_tag_templates[:rmr_number] = { 
@@ -431,10 +556,8 @@ utility.
   end
   
 
-=begin rdoc
-Extracts a collection of metadata from @hdr_data retrieved using the rdgehdr
-utility. 
-=end
+  # Extracts a collection of metadata from @hdr_data retrieved using the rdgehdr
+  # utility. 
   def rdgehdr_import
     source_pat =               /hospital [Nn]ame: ([[:graph:]\t ]+)/i
     num_slices_pat =           /Number of slices in this scan group: ([0-9]+)/i