Negation Issue
Description: TIES is not negating terms properly. Some of the results retrieved contain both the query keyword and the negation of that keyword. For example: Patient ID 163849 and Report ID 196647 is not negating the “was not seen”. Same issue has been reported from PITT and RPCI.
Annotation form not saving issue
Description: The content of the annotation form may not save the edits made by users. This only happens to Upenn and this only happens starting from the second annotation form.
Index/data gender count issue
Description: When searching for certain disease, the number of male patients plus the number of female patients does not equal to the sum of all patients with that disease.
TIES speed issue
Description: Launching and searching with TIES takes about 2 minutes. Such delay might arise from the communication with TIES node of old version.
Dynamically load reports of new document types based on content
Description: TIES currently looks for document types from the fixed position in the OBR section. The caties.properties file specifies which document type to load, too.
User from WFH suggests removing the configurations that specifies which document type the reports belong to.
We can design and develop software to load reports of any document type solely based on the content of the report.
Better logging invalid reports Issue
Description: We should output invalid reports into a new file with meaningful error messages.Currently such information is stored in logs.
Complaint from users:
Since the log contains hundreds of reports, some of them may fail due to different reasons.Read the TIES log file is so time consuming (thousands of lines); and cutting out these not-processed reports then pasting to a new file is also very time consuming.