I am currently investigating a weird layout for storing of what appears to be strings (medical format).
Looking at the dump for a bunch of those strings here is what I see:
$ hexdump -C out0000
00000000 df ff 79 17 01 09 00 49 53 4f 38 38 35 39 2d 31 |..y....ISO8859-1|
00000010 02 08 00 30 30 30 30 30 30 30 30 |...00000000|
0000001b
$ hexdump -C out0001
00000000 df ff 79 19 01 09 00 49 53 4f 38 38 35 39 2d 31 |..y....ISO8859-1|
00000010 02 0a 00 83 74 83 40 83 93 83 67 83 80 |[email protected]..|
0000001d
$ hexdump -C out0009
00000000 df ff 79 21 01 09 00 49 53 4f 38 38 35 39 2d 31 |..y!...ISO8859-1|
00000010 02 12 00 91 71 95 7e 90 ac 90 6c 95 61 83 5a 83 |....q.~...l.a.Z.|
00000020 93 83 5e 81 5b |..^.[|
00000025
$ hexdump -C out0002
00000000 df ff 79 16 01 09 00 49 53 4f 38 38 35 39 2d 31 |..y....ISO8859-1|
00000010 02 07 00 46 41 4e 54 4f 4d 55 |...FANTOMU|
0000001a
$ hexdump -C out0004
00000000 df ff 79 0f 01 09 00 49 53 4f 38 38 35 39 2d 31 |..y....ISO8859-1|
00000010 02 00 00 |...|
00000013
Has anyone seen this format before ? What is this ?
Steps to extract those weird strings:
- You need an instance (image) from a Toshiba MRT200SP5,
Extract the
Original Data
, using eg.:gdcmraw -t 700d,1008 toshiba_input.dcm original_data.raw
Use the work in progress
dump4.c
to process one portion of it (Type = WSTRING)
df ff 79
seem to be a marker or a magic number. The following byte is the size of the next part. And ISO8859-1 is a character encoding.