Systems and methods for natural language processing-based classification of electronic medical records

A method, comprising:

• receiving, by at least one processor, patient data comprising at least one written report associated with a patient;
• accessing, by the at least one processor, a dictionary of terminology associated with a disease;
• wherein the terminology comprises a plurality of descriptors;
• wherein each descriptor is indicative of at least one category of a plurality of categories associated with the disease;
• wherein each descriptor is associated with at least one descriptor-specific score representative of a relevance to the at least one category;
• inputting, by the at least one processor, the at least one written report into a tokenization function to output a set of tokens, wherein the tokenization function is configured to:
• parse at least one word pattern in the at least one written report, and
• generate the set of tokens from the at least one word pattern;
• determining, by the at least one processor, at least one identified descriptor, from the plurality of descriptors, that is present in the at least one written report based at least in part on the set of tokens associated with the at least one written report;
• determining, by the at least one processor, a category-specific score of a plurality of category-specific scores associated with each category of the plurality of categories based at least in part on:
• the at least one identified descriptor of the plurality of descriptors and
• the at least one descriptor-specific score of each descriptor representative of the relevance of each descriptor to the at least one category;
• determining, by the at least one processor, at least one recommended category score indicative of at least one particular category of the disease based at least in part on the category-specific score associated with each category;
• generating, by the at least one processor, at least one category recommendation representing the at least one particular category for the at least one patient based at least in part on the at least one recommended category score;
• and
• rendering, by the at least one processor, an output to a graphical user interface (GUI) associated with the at least one patient to present to a user at least one category recommendation associated with the at least one category recommendation score so as to provide clinical decision support,
• wherein the GUI is rendered with at least one filter user interface (UI) element having a plurality of selections associated with the plurality of categories of the disease;
• wherein each selection of the plurality of selections is configured to cause the GUI of the display to selectively show an associated one of the plurality of categories of the disease;
• and
• wherein user selection, via the GUI, of a respective selection associated with the at least one category recommendation causes the GUI to present the at least one written report associated with the patient.