A method to summarize Disease Based Temporal state of human organ using laboratory test data and UMLS knowledge

摘要

A novel concept of Disease Based Temporal Score (DT-Score) is introduced to efficiently represent periodic laboratory test data. Through this score, temporal state of an organ can be represented by summarizing periodic laboratory test data. The score can be used to indicate early trend for chronic abnormalities and thus results in an effective wellness measure. Many of these chronic abnormalities have a late manifestation and are major contributors for healthcare cost and mortality. Resources of Unified Medical Language System (UMLS) are introduced for automatic generation of relational tree between laboratory test, disease and organs with relative rank. Doctor's annotations are used to create reference score and data mining techniques are employed in deriving a mathematical model for estimating the DT-Score. A novel human body based summarization is employed for an intuitive view of the DT-Score and resultant temporal state of the associated organ. The proposed method enables an efficient temporal summarization of high volume of laboratory data and eventually reduces the cognitive load on physician. This method has potential to impact larger population as this can be effectively built over low cost regular laboratory test.