From: frog Date: Tue, 6 Jul 2004 12:50:33 +0000 (+0000) Subject: - README some comments on images moved to TestAllEntryVerifyReference.txt. X-Git-Url: https://git.creatis.insa-lyon.fr/pubgit/?a=commitdiff_plain;h=0d2868be4e7477db2b0b2b9f39e461ec603dfb42;p=gdcmData.git - README some comments on images moved to TestAllEntryVerifyReference.txt. - TestAllEntryVerifyReference.txt: added comments on file naming convention and format of file TestAllEntryVerifyReference.txt. --- diff --git a/Changelog b/Changelog index f2fa2fd..5c9aebe 100644 --- a/Changelog +++ b/Changelog @@ -1,4 +1,7 @@ 2004-07-06 Eric Boix + - README some comments on images moved to TestAllEntryVerifyReference.txt. + - TestAllEntryVerifyReference.txt: added comments on file naming + convention and format of file TestAllEntryVerifyReference.txt. - oldACR00001.ima renamed to SIEMENS_GBS_III-16-ACR_NEMA_1.acr (directly on server so no cvs message present) - added this Changelog diff --git a/README b/README index d4e6fa2..bd545f2 100644 --- a/README +++ b/README @@ -1,140 +1,7 @@ -multiframe1.dcm -multiframe2.dcm - Cardiovasculaire multiframe - True Dicom - Explicit VR - les champs prives ne *sont pas* Explicit VR - tous les Groupes comportent un element 0000 - Pixels codes sur 10 bits - -acusson.dcm - Echographie single Frame - True Dicom - Explicit VR - 3 LUT (R, G, B) - -irmPhlipsNew1.dcm - Irm Philips LyonSud - True Dicom - Explicit VR - SQ curieuse en (0008,1140) - -pixelNonRect.dcm - Irm Philips Marseille - True Dicom - Explicit VR - les champs prives *sont* Explicit VR - les champs prives contiennent des caract non imprimables - Pixels non Rectangulaires - -esrf1.dcm - Vg Studio Max - True Dicom - Explicit VR - Manque 'Samples Per Pixel' pour etre lue par e-film - -jpeg16Bits.dcm - CT Siemens HEH - True Dicom - jpeg 16 Bits Lossy - (fait casser DcmRead 8/11/2002 :-( - repare 27/11 - -jpeglossles1.dcm - Jpeg LossLess - True Dicom - Explicit VR - SQ en (0008,2112) - contient peut etre (?!) des curiosités apres les Pixels - -jpeglossless2.dcm - CR Thoravision Philips Cardio - Jpeg LossLess - True Dicom - Explicit VR - Jpeg mal decodé par les fonctions utilisees - -jpeglossy1.dcm -jpeglossy2.dcm -jpeglossy3.dcm - Jpeg Lossy - Decodage non traite par mes fonctions - -oldACR00001.ima - ACRNEMA V1 'ancien' - Groupe 20 NE contient PAS Study et Serie Instance UID - ET Study ID et Series Number sont VIDES +This repository + :pserver:xxx@cvs.creatis.insa-lyon.fr:2402/cvs/public +contains all the images used by the test suite of gdcm (refer to +http://www.creatis.insa-lyon.fr/Public/Gdcm ) -newACR1000.nema - ACRNEMA V2 'moderne' - Groupe 20 contient Study et Serie Instance UID - ET Study ID et Series Number - -imageEcat.ecat - Image au format ECAT (Medecine Nucleaire) - On n'y echapera pas encore tres longtemps, a celles la :-( - ----------------- Ajouts 20 Nov 2002 - -mr176621 - MR Toshiba - ACRNEMA-V2 - abscence de l'element 0000 (group length) pour tous les groupes - -cr_45031 - CR Fuji - ACRNEMA V2 - 1670x2010 - -MR.6799.1 - MR Pickert - DICOM V3 like - manque le 'file preamble' de 200 octets - -cr172241 - CR Kodak - ACRNEMA V2 - (0028 3000): SQ - (fffe, e00d), (fffe,e0dd), apres (0028, 3006) !!! - -US.1.2.dcm -US.1.3.dcm - US ACUSON - DICOM 3 Multiframe - JPEG 8 bits Lossy - (0028,0002) Samples Per Pixel = 3 - (0018, 6011) : SQ - les champs à l'interieur de cette 'Sequence' - contiennent des caract non imprimables qui déconfigurent l'ecran :-( - - Ces deux examens, correspondant a 2 patients differents ont le meme Study ID - ----------------- Ajouts 27/11/2002 - -sonata.dcm - IRM Siemens Sonata (images taggees) - SQ en (0008,1140),contenant (0008,1150), (0008,1155) 3 fois - - ----------------- Ajouts 16/12/2002 -icone.dcm - CT Picker - image 512x512 - (28 200 )contient une icone de 64x64 - --> DEUX groupes 7FE0 dans la même image - ------------------ Ajouts 13/02/2003 - -sonata.nema - MR Siemens sonata - True DICOM - (18,1310) est un champ de Value Multiplicity > 1 - vr : US mais longueur = 8 - ---------------Ajouts 21/03/2003 - -philipsMR-lossy.ima - MR Philips (Neuro) - True Dicom - indiquée comme ACR-NEMA V1 dans l'entete - JPEG Lossy (Transfer Syntax UID : [1.2.840.10008.1.2.4.51]) +For naming conventions and specific information of images, please refer to +the file TestAllEntryVerifyReference.txt. diff --git a/TestAllEntryVerifyReference.txt b/TestAllEntryVerifyReference.txt index 67582b9..6b19305 100644 --- a/TestAllEntryVerifyReference.txt +++ b/TestAllEntryVerifyReference.txt @@ -1,3 +1,74 @@ +# TestAllEntryVerifyReference.txt +# +# This file describes for (hopefully) all the images of this repository +# the Dicom field entries that we use for the test suite. +# Semantics: +# For a given image file, we store the most sensible (to the state of +# gdcm kernel) Dicom entries as a pair (key, value): +# - key is the Dicom tag (i.e. the pair [group,element] represented +# as a string (in hexadecimal with the pipe character acting as a +# separator between the group and the element). +# - value is the Dicom field that corresponds to the tag (and represented +# as a string. For binary entries [like the Pixels (the tag is 7fe0|0010 +# usually)], which don't have a string representation, we use the +# string constructed by gdcm e.g. +# "gdcm::NotLoaded (BinEntry) Address:776 Length:387200 x(5e880)". +# Format: +# - Each image shall be contained in a block delimited by [ and ] +# (square brackets). Nested blocks are an error. +# - each (key, value) pair shall stand on a single line: +# * key is of the form xxxx|yyyy (in hexadecimal) e.g. 0008|0060 +# * value is a string delimited by " (double quotes) e.g. "CR" +# - the spaces are the one of C++ [i.e. space, TAB, endline (including +# the empty line)...] +# - comments shall be preceded by # (sharp character) and be at the +# end of meaningfull information +# - All the information between blocks is discarded, but please use +# the comment syntax (line for the lines you are just reading) +# Example: +# [ filename.dcm +# # some nice comments and problems of filename.dcm +# 0008|0060 "CR" # Modality +# 0008|1090 "9000" # Manufacturer's model name +# 7fe0|0010 "gdcm::NotLoaded (BinEntry) Address:776 Length:387200 x(5e880)" +# ] +# This info shall be discarded (and has no comment which is lack of taste) +# +# [ +# otherFilename.dcm +# # some really nice comments and documentation +# 0028|0010 "440" +# 0028|0011 "440" ] # Columns +# +# +# File naming conventions: +# the general form shall be (unpresent or meaningfull field are optional) +# [Manufacturer]-[BytesPerPixel]-[PhotoInterpretation]-[SQ]-[Description].EXT +# where : +# - [Manufacturer] represents at best the aquisition device. It can +# can be a mix of the "Manufacturer" and "Manufacturer's Model Name" +# Dicom entries. +# - [BytesPerPixel] is the "Bits Stored" Dicom entry (or "Bits Allocated" +# when "Bits Stored" is absent). +# -[PhotoInterpretation] corresponds to the type of encoding of the image. +# It is a short for the "Photometric Interpretation" Dicom entry e.g. +# PAL (for "PALETTE COLOR"), RGB, YBR (for YBR_FULL), +# YBR422 (for YBR_FULL_4220, MONO2 (for MONOCHROME2)... +# - [SQ] when present indicates that the image contains a sequence. +# - [Description] shall be a consice description of the features +# specificities of the image (e.g. ACR_NEMA_1 to indicate it is an +# old ACR Nema version 1 file format). +# - .EXT is the extension and (allthough meaningless in terms of the +# dicom standard) shall either be .acr (for ACR Nema version 1 and 2) +# or .dcm for Dicom V3 images. +# Examples: +# - SIEMENS_GBS_III-16-ACR_NEMA_1.acr does it's best to express +# that this image was produced by a SIEMENS (manufacturer) "GBS III" +# (model name) imaging device. It is encoded on 16 bits per pixel, +# has no "Photometric Interpretation", no sequence, and uses the +# old ACR Nema version 1 file format. +# + [ CR-MONO1-10-chest.dcm #"Transfer Syntax UID" is absent. 0008|0060 "CR" # Modality @@ -44,3 +115,145 @@ 0008|1090 "GBS III " # Manufacturer's Model Name 7fe0|0010 "gdcm::NotLoaded (BinEntry) Address:8192 Length:132096 x(20400)" ] + +############################################################################# +# multiframe1.dcm +# multiframe2.dcm +# Cardiovasculaire multiframe +# True Dicom +# Explicit VR +# les champs prives ne *sont pas* Explicit VR +# tous les Groupes comportent un element 0000 +# Pixels codes sur 10 bits +# +# acusson.dcm +# Echographie single Frame +# True Dicom +# Explicit VR +# 3 LUT (R, G, B) +# +# irmPhlipsNew1.dcm +# Irm Philips LyonSud +# True Dicom +# Explicit VR +# SQ curieuse en (0008,1140) +# +# pixelNonRect.dcm +# Irm Philips Marseille +# True Dicom +# Explicit VR +# les champs prives *sont* Explicit VR +# les champs prives contiennent des caract non imprimables +# Pixels non Rectangulaires +# +# esrf1.dcm +# Vg Studio Max +# True Dicom +# Explicit VR +# Manque 'Samples Per Pixel' pour etre lue par e-film +# +# jpeg16Bits.dcm +# CT Siemens HEH +# True Dicom +# jpeg 16 Bits Lossy +# (fait casser DcmRead 8/11/2002 :-( +# repare 27/11 +# +# jpeglossles1.dcm +# Jpeg LossLess +# True Dicom +# Explicit VR +# SQ en (0008,2112) +# contient peut etre (?!) des curiosités apres les Pixels +# +# jpeglossless2.dcm +# CR Thoravision Philips Cardio +# Jpeg LossLess +# True Dicom +# Explicit VR +# Jpeg mal decodé par les fonctions utilisees +# +# jpeglossy1.dcm +# jpeglossy2.dcm +# jpeglossy3.dcm +# Jpeg Lossy +# Decodage non traite par mes fonctions +# +# oldACR00001.ima +# ACRNEMA V1 'ancien' +# Groupe 20 NE contient PAS Study et Serie Instance UID +# ET Study ID et Series Number sont VIDES +# +# newACR1000.nema +# ACRNEMA V2 'moderne' +# Groupe 20 contient Study et Serie Instance UID +# ET Study ID et Series Number +# +# imageEcat.ecat +# Image au format ECAT (Medecine Nucleaire) +# On n'y echapera pas encore tres longtemps, a celles la :-( +# +# ---------------- Ajouts 20 Nov 2002 +# +# mr176621 +# MR Toshiba +# ACRNEMA-V2 +# abscence de l'element 0000 (group length) pour tous les groupes +# +# cr_45031 +# CR Fuji +# ACRNEMA V2 +# 1670x2010 +# +# MR.6799.1 +# MR Pickert +# DICOM V3 like +# manque le 'file preamble' de 200 octets +# +# cr172241 +# CR Kodak +# ACRNEMA V2 +# (0028 3000): SQ +# (fffe, e00d), (fffe,e0dd), apres (0028, 3006) !!! +# +# US.1.2.dcm +# US.1.3.dcm +# US ACUSON +# DICOM 3 Multiframe +# JPEG 8 bits Lossy +# (0028,0002) Samples Per Pixel = 3 +# (0018, 6011) : SQ +# les champs à l'interieur de cette 'Sequence' +# contiennent des caract non imprimables qui déconfigurent l'ecran :-( +# +# Ces deux examens, correspondant a 2 patients differents ont le meme Study ID +# +# ---------------- Ajouts 27/11/2002 +# +# sonata.dcm +# IRM Siemens Sonata (images taggees) +# SQ en (0008,1140),contenant (0008,1150), (0008,1155) 3 fois +# +# +# ---------------- Ajouts 16/12/2002 +# icone.dcm +# CT Picker +# image 512x512 +# (28 200 )contient une icone de 64x64 +# --> DEUX groupes 7FE0 dans la même image +# +# ----------------- Ajouts 13/02/2003 +# +# sonata.nema +# MR Siemens sonata +# True DICOM +# (18,1310) est un champ de Value Multiplicity > 1 +# vr : US mais longueur = 8 +# +# --------------Ajouts 21/03/2003 +# +# philipsMR-lossy.ima +# MR Philips (Neuro) +# True Dicom +# indiquée comme ACR-NEMA V1 dans l'entete +# JPEG Lossy (Transfer Syntax UID : [1.2.840.10008.1.2.4.51])