Recovering ultrasound data -- does this look familiar?

[Colin Berry]

I've spent my afternoon working to recover several years worth of obstetric ultrasounds from magneto optical disks. I have made what for me is great progress and am now fairly certain that the data will not be lost. To summarize my copious notes (which I will post if requested) I have a directory full of tiff files with the extension .cri and a four letter index. I also have two almost identical files named printq.db and archiveq.db which contain a database in some format unknown to me which connects the four letter file index to the patient's name, id number, date of birth, and date of study.

My current plan is to migrate this to a mysql db with the tiff image as a blob and the other data in sensible columns.

My best scripting language for this job is php. I know that I can convert the data by brute force, but I think it's worth asking if it looks familiar to anyone. Seems a shame to reinvent the wheel if this is a standard db format with a readily available migration scheme.

cat -nv mnt/archiveq.db

1 00000022^M
2 0000C484^M
3 00BD^M
4 0000^M
5 ^L^@010A^M
6 SGMS0002^M
7 IM1|aaaacri^M
8 1837 4-1-86 Jane Doe ^M
9 2007:03:07 13:30:08^M
10 L2| AP|00 Vi|00 NV|01 AE|00 CE|01 PE|00 RE|00 UE|00
^M
11 FREE^M
...................................................................................
2079 FREE^M
2080 0001^M
2081 0000^M
2082 8000^M
2083 0001^M
2084 ^L

Looks like the first eleven lines describe the first image file. This repeats to the end where there are 5 lines of junk. Line 7 is the image file path ie ./mnt/im1/aaaa.cri. Line 8 is space delimited and has the patient id number assigned by my practice the date of birth and the name. (Mostly last,first but sometimes not.) Line 9 is the date and time the image was taken. The rest is greek to me.

cat -nv mnt/printq.db

This looks exactly the same except line 10 of each 11 line entry is a little different.

Does this look familiar to anybody? If not I'll just do 11 reads and throw 8 away.