The document discusses a web-based interface for querying SNOMED CT, highlighting its functionalities for terminology authoring, reference set authoring, and EHR queries. It addresses the challenges of searching through over 300,000 concepts and introduces Snow Owl's meaningful query system, along with various filters and advanced search capabilities. The interface aims to balance powerful search functionalities with user-friendliness to accommodate users with varying levels of knowledge about SNOMED CT.

1. A
simple
web-­‐based
interface
for
advanced
SNOMED
CT
queries
Orsolya
Bali,
Brandon
Ulrich
–
B2i
Healthcare
IHTSDO
Showcase
2014.10.30
-­‐
Amsterdam

2. Why
search
or
query
SNOMED
CT
content?
Common
use
cases
include:
• Terminology
authoring
• Reference
set
authoring
–
intensional
sets
• Mappings
• Terminology
binding
• EHR
queries
• ValidaOon/QA
1

3. Challenges
• Over
300,000
concepts
• Lexical
search
may
be
insufficient
– Different
meanings
could
sOll
carry
the
same
descripOon
• SemanOc
search
requires
comprehensive
knowledge
of
the
concept
model
• Various
component
properOes,
extendable
by
RF2
• Average
users
have
high-­‐level
or
only
vague
understanding
of
SNOMED
CT
2

4. A
search
interface
must
be
• …
powerful
enough
to
cater
for
all
the
relevant
a[ributes
(including
lexical,
semanOc
and
metadata
searches).
• …yet
simple
and
user
friendly.
3

5. Snow
Owl
Meaningful
Query
(MQ)
• Online
browser
for
SNOMED
CT
and
extensions
– Advanced
search
– Filtered
hierarchy
• EHR
datastore
• SemanOc
query
engine
to
extract
the
semanOcs
captured
in
EHRs
• Terminology
server
integraOon
4

6. DEMO
5

7. Introduce
filters
• Concept,
DescripOon,
Property,
Refset
and
Mapping
filters
• Some
relevant
filters:
property-­‐value,
module,
descripOon
term,
reference
set,
mapping
6

8. Features
to
check
out
• Create
a
simple
filter
• Open
a
few
more
complex
filters
• Browse
view
• Filtered
browse
view
7

9. The
advanced
search
interface
8

10. Lexical
search
in
a
specific
descripMon
type
To
return
all
concepts
that
have
an
acOve
SG
English
preferred
term
containing
the
word
tetralogy.
9

11. Search
by
aNribute
(relaMonship)
To
return
all
concepts,
that
have
a
procedure
device
type
of
a[ribute,
or
any
of
its
subtypes
(eg.
direct
device),
with
the
value
of
arthroscope.
10

12. Search
by
aNribute
(concrete
domain)
To
return
all
concepts
with
the
acOve
ingredient
of
paracetamol,
whose
strength
is
between
50
and
500
mg.
11

13. Complex
query
The
below
query
combines
hierarchical,
a[ribute,
lexical
and
metadata
searches,
using
logical
and,
logical
or
operators
and
exclusion.
12

14. Search
in
reference
set
13
To
return
members
of
a
reference
set.

15. Search
by
mapping
To
search
concepts
by
their
maps.
The
below
example
returns
all
concepts,
that
are
mapped
to
the
specified
ATC
code.
14

16. Browse
hierarchy
15

17. Quick
filter
hierarchy
16

18. Apply
a
pre-­‐authored
filter
17

19. Manage
filters
18

20. Dashboard
19

21. QUESTIONS?
h[p://www.b2internaOonal.com
obali@b2internaOonal.com
20

22. THANK
YOU!
21