[TERM] extending data to include mismatched radiology/pathology conclusions

* Eric Prud'hommeaux <eric@w3.org> [2011-05-10 11:17-0400]
> Coding up
> <http://www.w3.org/wiki/HCLSIG/Terminology/PathRadCorrelation>, it
> looks like the radiology report includes only two of the three lumps
> reported by radiology:
> 
> sparql -8 -d ~/asdf.ttl -e "\
> PREFIX term: <term:/>\
> SELECT * {\
>   { [ a term:MammogramReport ;\
>       term:finding [\
>           term:location [\
>               term:laterality ?lat ;\
>               term:position ?clock\
>           ]\
>       ]\
>     ] } MINUS\
>   { [ a term:PathologyReport ;\
>       term:finding [\
>           term:location [\
>               term:laterality ?lat ;\
>               term:position ?clock\
>           ]\
>       ]\
>     ] }\
> }" 
> ┌──────────────┬──────────────┬──────────────┬────────────────────────┬──────────────┐
> │ _:b0x29ac5a0 │ _:b0x29ac620 │ _:b0x29beb80 │ ?clock                 │ ?lat         │
> │ _:b0x29ba430 │ _:b0x29bab40 │ _:b0x29b7500 │ "2:00"^^<units:/clock> │ <term:/left> │
> └──────────────┴──────────────┴──────────────┴────────────────────────┴──────────────┘
> 
> No biopsy for the lump at 2:00?
> <asdf.ttl attached>

Oops, forgot to attach the data. I attach two queries, for with the output is:

sparql -8 -d ~/asdf.ttl matching.rq 
┌──────────────┬────────────────────────┬──────────────────────────────────────────────┬────────────────────────────────────────────────────────────────┐
│ ?lat         │ ?clock                 │ ?radFinding                                  │ ?pathFinding                                                   │
│ <term:/left> │ "1:00"^^<units:/clock> │ "fine linear and pleomorphic calcifications" │ "DUCTAL CRIBRIFORM CARCINOMA IN SITU WITH MICROCALCIFICATIONS" │
│ <term:/left> │ "2:00"^^<units:/clock> │ "hypoechoic mass with related calcification" │                                                             -- │
│ <term:/left> │ "9:00"^^<units:/clock> │                          "spiculated margin" │                               "INVASIVE DUCTAL ADENOCARCINOMA" │
└──────────────┴────────────────────────┴──────────────────────────────────────────────┴────────────────────────────────────────────────────────────────┘
sparql -8 -d ~/asdf.ttl missing.rq 
┌──────────────┬────────────────────────┬──────────────────────────────────────────────┐
│ ?lat         │ ?clock                 │ ?radFinding                                  │
│ <term:/left> │ "2:00"^^<units:/clock> │ "hypoechoic mass with related calcification" │
└──────────────┴────────────────────────┴──────────────────────────────────────────────┘

BTW, in the turtle file, I've migrated
<http://www.w3.org/wiki/SuggestedN3PairedReport1> to use the TMO. I'd
like help extending the data to include mismatched radiology/pathology
conclusions. John, do you have some bandwidth to geek on that? I note
that the 1:00 mass showed evidence of calcification on both radiology
and pathology reports. If it did not, would that indicate a an issue
requiring further examination? Are there some scenarios which would
indicate that the excision missed the target?
-- 
-ericP