I know this is a bit off thread, but...
Re. the work by De Dombal (I gather he sadly died this year),
this is where I was first introduced to computers, many years ago!
My father was chief technician in the Dept. of Diagnostic Methodology in
the Southern General, Glasgow.
Professor Caird, a gastroentorologist, was with that Dept.
I gather he was appointed the world's first ever Professor of Computing as applied to
medicine, at Glasgow University.
Anyway, the Dept. had a PDP-11/45 computer (much less powerful than all your PC's
these days!). Huge, removable 80MByte disk packs!
Programmed in FORTRAN by Lucas Associates.
(FORTRAN for human-computer interaction - yikes!)
Anyway, they built up a database of patient responses to questions on abodominal
pain. Using the Baysian statistics of Dr. De Dombal, the computer was able to
come up with some pretty accurate diagnoses.
A couple of the things I remember are
1) patients were more likely to tell the 'truth' to a computer, being less
embarrased. e.g. about levels of alcohol consumption.
"I only drink to be sociable, doctor"
2) Some of the computer vocabulary had to be changed to suit local tastes.
I'm sure Alan Hyslop can tell us what the 'dry boak' is!
The unit had a room where the patient sat in front of a computer terminal
(none of your windows in those days). The keyboard was replaced by a box with
two or three touch sensitive buttons.
Unknown to the patient, researchers could watch via a one-way screen
and do research on patient-computer interaction.
I gather the PDP went to the great scrapyard in the sky a few years ago. Sad.
John Hearns
%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
