Present: Aslak, Hans, Øystein, Rune, Arezoo
Status
Rune:
We still don't have enough access to the data.
A: Remove Gender, Age, Identity.
AG: We have diagnosis code, and the number of days since day0.
We need to sign Non-Disclosure-Agrements.
We need informed consent from patients to work on free text.
Arezoo:
ICD-10 codes from Australia ...
(ICD-9 from America, convert it to ICD-10).
2012-version is much more detailed than the versions we map to.
Aslak: A81.09-level is deep enough
UMLS-semantic type: Ø: Pointing into the UMLS-ID? Maybe we can use the semantic types from UMLS.
The ICD-10 codes are from the upper level. Is that ok?
Summary: First look at one trajectories, then start splitting them into meaningful separate trajectories.
Øystein
talks to the people who translated ICD-10 into Norwegian.
Co-morbidity data from Karolinska (No answer yet)
Still not found any research on the temporal aspect of events, diagnoses, (Stefan Darmoni might know about this).
-We want to keep developing our own Norwegian ICD-10 ontology with chronic, temporal, etc...
A: Content of Ontology? ICD-10++ Example: Did Peter have this disease in 1940? Yes, if ....
A: Who validates the ontology? If they behave in a useful way, that's the validation.
Summary: We need to classify trajectories (but first we do just ONE big multi-year trajectory for each patient)
Aslak
How is co-morbidity related to trajectories?
Ø: We group events (for one individual) according to weak relationships (given by co-morbidity)
This will give us a classification system to split one patient into several trajectories. (Per Disease, Treatment, ...)
Ø: We need the background knowledge, can use data from Karolinska (Sjukehus, ikke institutt).
Ø: Segmentation based on clinical data (Japanese: Tanushi/Hercules).
There are existing classification about what is Chronic diseases or not. Ø: We have done some of this already.
-R: Where is this data? Ask Gro about where this data can be found in the research literature.
Summary: We need (ASAP) a graph showing That Patient, visiting That Department. Next Step: Show this in PE. (Rune, just do it ;-)
Hans
What about mapping between text/ontology? Ø: We don't have any text... Just classifying chronic or not, etc.