The National Library of Medicine provides tables with the UMLS CUI and the native ontology code (e.g. ICD, LOINC, etc.). You can download the UMLS tables from the NLM website and then use them for the SNOMED – ICD mapping.
Hope this helps. --Guergana From: Prasanna Bala [mailto:[email protected]] Sent: Tuesday, April 15, 2014 7:32 AM To: [email protected] Subject: Re: ICD9 code resolving.. Hi, Can you please suggest me some ideas how can we go for UMLS concept id to ICD10 tagging ? For eg: when I run the UMLS module for the disease "Diabetic nephropathy", I am getting the CUI id as "C0011881" for Tui T047. How can I match the ICD10 tagging for this ? If someone can give me ideas for simple mapping from then I can for it. Also I get SNOMED code for "diabetic nepropathy" I searched for both query in SNOMED CT browser. But I also need to relate with ICD10 tagging. I am new to this tagging. So I need some help on this. Thanks for response. Regards, Prasanna. On Sat, Apr 12, 2014 at 11:51 AM, Savova, Guergana <[email protected]<mailto:[email protected]>> wrote: cTAKES uses UMLS codes that are subset to sources like SNOMED CT and RxNORM augmented with synonyms. If you are interested in ICD-9 codes, you might be able to use the UMLS CUI to map to the ICD-9 code. Hope this helps. --Guergana From: Prasanna Bala [mailto:[email protected]<mailto:[email protected]>] Sent: Friday, April 11, 2014 3:41 AM To: [email protected]<mailto:[email protected]> Subject: Re: ICD9 code resolving.. Hi, I was reading a mail from the cTakes user group the other day where it was mentioned that cTakes uses ICD9 code internally. But then write the output for SNOMED CT code. I have certain entities for which I need to find out ICD9/10 codes. For eg: Consider the disease "DIABETES MELLITUS". There are so many ICD codes present for this illness. I need to narrow down to fewer ICD codes. Is it possible to automate this type of problem using cTakes. I also found that SNOMED CT code can be mapped to ICD codes. How reliable is this ? Also I have implemented another version of lookup table for matching UMLS semantic types using trie structure. I found this to be pretty fast although I implemented in python. I would like to contribute this to cTakes because this can improve the efficiency of the lookup process too. Please let me know your thoughts. Thanks for response. Regards, Prasanna On Thu, Apr 10, 2014 at 4:02 PM, Miller, Timothy <[email protected]<mailto:[email protected]>> wrote: Prasanna, I'm not aware of ICD9 codes being used within cTAKES. Is there a specific piece of code you're referencing when you say: "I found that ICD9 code is used internally by the cTakes for finding the entities." Tim ________________________________ From: Prasanna Bala [[email protected]<mailto:[email protected]>] Sent: Thursday, April 10, 2014 2:15 AM To: [email protected]<mailto:[email protected]> Subject: ICD9 code resolving.. Hi, I have some clarifications regarding ICD9 code. I found that cTakes uses different mapping technique to resolve the entities from the diagnosis. I want to know how they are doing this for ICD9. When I run the code, in the output it is showing the various tagged entities with SNOMED CT code. But my question, why I am not able to access the ICD9 code. I found that ICD9 code is used internally by the cTakes for finding the entities. How can I find the IC9 code for the entities ? Can you please tell me if there is any module which is using ICD9 code in the cTakes. I need to pipeline the ICD9 code also separately. Regards, Prasanna
