Predictive model for medical signatures

A system for printing, the system comprising:

• a memory;
• and
• one or more processors coupled to the memory, wherein the memory comprises non-transitory computer instructions that when executed by the one or more processors perform operations comprising:
• accessing a communication comprising a medicinal drug prescription associated with a patient, the medicinal drug prescription being provided in the communication by a healthcare professional;
• pre-processing the medicinal drug prescription by optimizing text comprising the medicinal drug prescription to generate optimized text;
• processing the optimized text to extract a plurality of entities from the medicinal drug prescription by
• determining if the optimized text includes a drug identifier and lacks a doctor instruction,
• if the optimized text includes the drug identifier and lacks the doctor instruction, selecting from a plurality of entity extraction engines, a drug entity extraction model, generating, with the drug entity extraction model, a set of drug entities for the optimized text, and storing the set of drug entities as the plurality of entities, and
• if the text includes the doctor instruction, selecting a doctor entity extraction model, generating, with the doctor entity extraction model, a set of entities for the optimized text, and storing the set of entities as the plurality of entities,
• wherein the plurality of entity extraction engines includes a first machine learning model;
• applying a second machine learning model to the plurality of entities to predict a medical signature, the predicted medical signature comprising directions for the patient to follow to take the medicinal drug;
• and
• printing, with a printer, a label with the medical signature to be applied to a container that includes the medicinal drug.

A non-transitory computer readable medium comprising non-transitory computer-readable instructions for performing operations associated with printing, the operations comprising:

• accessing a communication comprising a medicinal drug prescription associated with a patient, the medicinal drug prescription being provided in the communication by a healthcare professional;
• pre-processing the medicinal drug prescription by optimizing text comprising the medicinal drug prescription to generate optimized text;
• processing the optimized text to extract a plurality of entities from the medicinal drug prescription by
• determining if the optimized text includes a drug identifier and lacks a doctor instruction,
• if the optimized text includes the drug identifier and lacks the doctor instruction, selecting from a plurality of entity extraction engines, a drug entity extraction model, generating, with the drug entity extraction model, a set of drug entities for the optimized text, and storing the set of drug entities as the plurality of entities, and
• if the text includes the doctor instruction, selecting a doctor entity extraction model, generating, with the doctor entity extraction model, a set of entities for the optimized text, and storing the set of entities as the plurality of entities,
• wherein the plurality of entity extraction engines includes a first machine learning model;
• applying a second machine learning model to the plurality of entities to predict a medical signature, the predicted medical signature comprising directions for the patient to follow to take the medicinal drug;
• and
• printing, with a printer, a label with the medical signature to be printed and applied to a container that includes the medicinal drug.