Starting to look at 5.0 repo and found this...

[Peter Abramowitsch]

Hi all,   Looking at the 5.0 repo, there's a compilation error across many
projects because what is being imported as
*org.apache.ctakes.contexttokenizer.ae.**

is actually located in package
*org.apache.ctakes.context.tokenizer.ae
*
and the maven artifact is declared as
*ctakes-context-tokenizer*

I've changed the import statements in 13 files, but if there's anyone who
feels strongly that I should leave those alone and change the package &
folder instead, let me know.

DId anyone else notice this too?

Peter

Re: Starting to look at 5.0 repo and found this... [EXTERNAL]

[Finan, Sean]

Hi Pete,

Thanks for testing!

I haven't seen those compilation errors, but I think that changing the import 
statements to fit the project structure is the best thing to do.

Just out of curiosity, are those imports in any particular module?

Thanks,

Sean

From: Peter Abramowitsch 
Sent: Friday, November 17, 2023 5:12 PM
To: dev@ctakes.apache.org 
Subject: Starting to look at 5.0 repo and found this... [EXTERNAL]

* External Email - Caution *


Hi all,   Looking at the 5.0 repo, there's a compilation error across many
projects because what is being imported as
*org.apache.ctakes.contexttokenizer.ae.**

is actually located in package
*org.apache.ctakes.context.tokenizer.ae
*
and the maven artifact is declared as
*ctakes-context-tokenizer*

I've changed the import statements in 13 files, but if there's anyone who
feels strongly that I should leave those alone and change the package &
folder instead, let me know.

DId anyone else notice this too?

Peter

Discrepancy in cTAKES Identification of 'Chemotherapy' SNOMED Codes

[Akram]

Hi AllI wonder if someone can help me answer these questions:
I am a PhD student, my research involves utilizing cTAKES. I'm using the 
desktop version, cTAKES 4.0.0.1 , for my research. I built a dictionary using 
the most recent SNOMED codes 2023 sourced from UMLS. My research involves using 
the MIMIC VI database, within which resides report #5419. This particular 
report has been diagnosed with ICD10 code Z51.11.
Upon manual inspection of the report, I observed the occurrence of the term 
"Chemotherapy" appearing 13 times. Interestingly, cTAKES identified these 
instances of "Chemotherapy" using codes 363688001 and 367336001. However, the 
specific SNOMED codes that accurately represent the diagnosis of "Chemotherapy" 
are 373832009, 373833004, 373835006, and 373836007.
This discrepancy prompts the question: Why did cTAKES opt for codes 363688001 
and 367336001 instead of employing 373832009, 373833004, 373835006, and 
373836007 to represent the concept of "Chemotherapy"? Any resources I can use 
to help me with other similar questions?
Thanks

Re: Discrepancy in cTAKES Identification of 'Chemotherapy' SNOMED Codes

[Kean Kaufmann]

Akram - How did you decide which SNOMED codes were correct?

> Interestingly, cTAKES identified these instances of "Chemotherapy" using
codes 363688001 and 367336001.

The UMLS CUI C3665472, with TUI T061 (Therapeutic or Preventive Procedure),
includes the SNOMED codes:

   - Chemotherapy (367336001)
   - Administration of antineoplastic agent (363688001)

> However, the specific SNOMED codes that accurately represent the
diagnosis of "Chemotherapy"

What's your reason for treating "Chemotherapy" as a diagnosis, rather than
a procedure?

> This particular report has been diagnosed with ICD10 code Z51.11.

The ICD10CM code Z51.11 code description says "Encounter for antineoplastic
chemotherapy".
Like many ICD10CM Z-codes, it's a diagnostic label indicating a procedure
was (to be) performed.

> ...are 373832009, 373833004, 373835006, and 373836007.

Your four SNOMED codes are all closely related, and semantically loaded:
they mean more than just the concept of chemotherapy.
The last three are children of the first, and to me they look mutually
exclusive.

   - 373832009: Reason for change in planned chemotherapy treatment (UMLS
   CUI: C1298666)
  - 373833004: Chemotherapy changed - progressive disease during
  chemotherapy (UMLS CUI: C1298667)
  - 373835006: Chemotherapy changed due to technical or organizational
  problems (UMLS CUI: C1300033)
  - 373836007: Chemotherapy changed - patient choice (UMLS CUI:
  C1298669)

> Any resources I can use to help me with other similar questions?

UMLS Metathesaurus Browser: https://uts.nlm.nih.gov/uts/umls/home
Signup is free.
_
*Kean Kaufmann*
NLP Architect

RecordsOne
nSight Driven | *Priority. Clarity. Integrity. *


On Fri, Nov 17, 2023 at 7:40 PM Akram  wrote:

> Hi AllI wonder if someone can help me answer these questions:
> I am a PhD student, my research involves utilizing cTAKES. I'm using the
> desktop version, cTAKES 4.0.0.1 , for my research. I built a dictionary
> using the most recent SNOMED codes 2023 sourced from UMLS. My research
> involves using the MIMIC VI database, within which resides report
> #5419. This particular report has been diagnosed with ICD10 code Z51.11.
> Upon manual inspection of the report, I observed the occurrence of the
> term "Chemotherapy" appearing 13 times. Interestingly, cTAKES identified
> these instances of "Chemotherapy" using codes 363688001 and
> 367336001. However, the specific SNOMED codes that accurately represent the
> diagnosis of "Chemotherapy" are 373832009, 373833004, 373835006, and
> 373836007.
> This discrepancy prompts the question: Why did cTAKES opt for codes
> 363688001 and 367336001 instead of employing 373832009, 373833004,
> 373835006, and 373836007 to represent the concept of "Chemotherapy"? Any
> resources I can use to help me with other similar questions?
> Thanks
>