Extracting entities, relations, and contextual attributes

Palmyra-Med uses context-aware models to extract medical entities, relations, and contextual attributes. Each text entity is extracted into a medical dictionary entry.


The maximum size of the medical text is 10,000 unicode characters.

**Palmyra-Med supports the following medical knowledge categories:

BM_UNITBody measurement unit
BM_VALUEValue of a body measurement
BODY_FUNCTIONFunction carried out by the human body
BODY_MEASUREMENTA normal measurement of the human body, such as a vital sign
LABORATORY_DATAResults of testing a bodily sample
LAB_RESULTQualitative description of laboratory data, such as increased, decreased, positive, or negative
LAB_UNITUnit of measurement for a laboratory value
LAB_VALUEValue of an instance of laboratory data
MEDICAL_DEVICEPhysical or virtual instrument
MEDICINEDrug or other preparation for treatment or prevention of disease
MED_DOSEMedication dose
MED_DURATIONMedication duration
MED_FORMPhysical characteristics of the specific medication
MED_FREQUENCYHow often the medication is taken
MED_ROUTEBody location at which the medication is administered
MED_STATUSFor an existing medication, status can be a modifier such as "continue", "start", "restart", "stop", "switch", "increase", "decrease".
MED_STRENGTHAmount of active ingredient in a dose of medication
MED_TOTALDOSEQuantity of medication to take at one time
MED_UNITUnit of measurement for the active ingredient in a medication
PROBLEMMedical condition, including findings and diseases
PROCEDURE_RESULTResults of a procedure
PROCEDUREDiagnostic or treatment procedure
PROC_METHODMethod in which a procedure is conducted
SEVERITYSeverity of the medical condition
SUBSTANCE_ABUSEAbuse of a psychoactive substance