| Budget Amount *help |
¥2,000,000 (Direct Cost: ¥2,000,000)
Fiscal Year 1997: ¥600,000 (Direct Cost: ¥600,000)
Fiscal Year 1996: ¥1,400,000 (Direct Cost: ¥1,400,000)
|
|---|
| Research Abstract |
Object analysis were applied to prescription, clinical lab test results, and diagnostic disease classifications of clinical case data. Methods and semantic axis of analysis were investigated so that they can be used for intelligent data retrieval for clinical supports. As the results, prescriptions should be categorized in effects, medication routes, and medicine styles. Clinical laboratory test results in test methods, and specimen varieties. Diagnostic disease classifications in anatomical sites, and pathophysiological states.
Then, the researcher made a conversion method to change all daily transaction data of these three kinds, arose in Hamamatsu University Hospital information systems, into HL7 message data types. And they are sent to object oriented data-base, where these data are categorized and storec as object representations. This is the first case in Japan, that this huge amount of daily hospital information data transmitted into clinical support data-base in standardized format. All the day's data are sent to Object-oriented data-base in the midnight of the day. By doing this, diagnostic disease registered by ICD-9-CM based classification, are converted to SNOMED description of disease states.he researcher also created this auto-converter program.
The clinical support object-oriented data-base, with its rich representation of semantics of data in hierarchies, can allow abstract level retrieval, such as "any patient who were prescribed xxx series antibiotics, with at least one time high record of GOT being more than 100." The researcher believes that this level of abstractness in clinical information retrieval has never before realized.
The researcher hopes that HL7 implementation in Japanese hospitals would increase, following by this pioneer experiment, and that real semantic clinical support would realized in very near future.
|
