Codes, Codes, Codes

With the use of Solr as our source for guideline text we are developing a process for importing text and assigning RxNorm, CPT, and ICD-10 codes. The original text will be normalized by running it through cTAKES. The selection of codes will be performed by taking the results of a word search of the UMLS database for coded definitions and comparing the text with the input text using cosine similarity measures.