FYI--the American College of Radiology created the BI-RADS
terminology precisely for the purposes of providing a controlled
lexicon for radiology observations on mammography. Extensions on this
are needed to describe findings in ultrasound and MRI (future work in
BI-RADS). There is a possibility that these terms will be
incorporated into RadLex as well. Thus, you might not want to build
on this, focusing on adding the clinical indications or other
information that is not related to image observations.
Daniel
At 11:13 AM 2/12/2008, Matt Williams wrote:
Dear Helen,
At the moment we have some structure with a patient, who has one or
more studies, each of which has one or more images associated with
it. each image has zero or more lesions associated with it.
We're working with a dataset of ~400 mammogram reports, and using
this as the basis for the developing the ontology (to check
coverage, etc.). As a result, the patient info is reasonably sparse
(as not much/ if anything is supplied, apart from some demographic
info for most patients). The lesion info is really where there's a
bit more detail - calcification, masses, etc. and a little bit more
at the study level (reporting BIRADS category, etc.).
Does that answer your questions? What we've done is fairly sparse at
the moment, but in theory shouldn't take that long to flesh out, at
least to the level you've described. I'm guessing AGFA has done a
bit more on this - do you have any pointers/ papers/ work to suggest.
Thanks,
Matt
[EMAIL PROTECTED] wrote:
Matt
Can you elaborate on what aspects of mammography you include in
your ontology? I am interested in the diagnostic aspect, i.e.
image features, patient clinical information and the diagnosis to
breast cancer.
Kind regards.
Helen
*Matt Williams <[EMAIL PROTECTED]>*
Sent by: [EMAIL PROTECTED]
02/12/2008 01:35 PM
To
public-semweb-lifesci hcls <public-semweb-lifesci@w3.org>
cc
Subject
Mammographic ontology
Dear All,
Have just been reading the wiki. I note that there is a section on doing
a mammogram as a screening test. I have been doing some work on a
mammographic ontology, which we might be able to contribute (need to
talk to other authors).
Would this be useful?
How can I align it with existing terms (e.g. from GALEN, etc.)?
Matt
-- http://acl.icnet.uk/~mw
http://adhominem.blogsome.com/
+44 (0)7834 899570
--
http://acl.icnet.uk/~mw
http://adhominem.blogsome.com/
+44 (0)7834 899570