Systems, methods, and apparatuses for generating structured data from unstructured data using natural language processing to generate a secure medical dashboard

1 . A system for generating a secure medical dashboard using natural language processing (NLP), the system comprising:

at least one non-transitory storage device;

at least one processor coupled to the at least one non-transitory storage device, wherein the processing device is configured to execute computer program code comprising computer instructions configured to cause said at least one processor to perform the following operations:

identify at least one data input, wherein the at least one data input comprises unstructured data;

apply at least one trained NLP pipeline to the at least one data input, wherein the at least one trained NLP pipeline is trained using the following operations:

identifying at least one previous unstructured dataset, wherein the at least one previous unstructured dataset comprises an at least one previous term;

applying the at least one NLP pipeline to the at least one previous unstructured dataset;

determining, by the at least one NLP pipeline, the medical relevance attribute for the at least one previous term;

mapping, by the at least one NLP pipeline, the at least one previous term comprising the positive medical attribute to the medical entity title;

comparing the medical entity title for the at least one previous term to a pre-labeled medical entity title for the at least one previous term;

comparing the positive medical attribute for the at least one previous term to a pre-labeled medical relevance attribute;

determining whether to adjust at least one weight associated with the at least one NLP pipeline based on the comparison of the medical entity title to the pre-labeled medical entity title and the comparison of the positive medical attribute to the pre-labeled medical relevance attribute,

adjust, in an instance where at least one of the medical entity title does not match the pre-labeled medical entity title or the medical attribute does not match the pre-labeled medical relevance attribute, the at least one weight of the at least one NLP pipeline, or

maintain, in an instance where the medical entity title matches the pre-labeled medical entity title and the medical attribute matches the pre-labeled medical relevance attribute, the at least one weight of the at least one NLP pipeline;

training, in response to the determination of whether to adjust the at least one weight, the at least one NLP pipeline based on the at least one weight of the at least one NLP pipeline;

parse, by the at least one trained NLP pipeline, the unstructured data of the at least one data input to generate a parsed unstructured dataset, wherein the parsed unstructured dataset comprises a plurality of terms;

identify, by the at least one trained NLP pipeline, a medical relevance attribute for each term within the unstructured dataset, wherein the medical relevance attribute comprises at least one of a positive medical attribute, a medical modifier attribute, or a negative medical attribute;

generate, based on the medical relevance attribute for each term, a structured document comprising each term and associated medical relevance attribute comprising the positive medical attribute or the medical modifier attribute;

correlate, by the at least one trained NLP pipeline, each term comprising the positive medical attribute to a medical entity title;

generate a medical dashboard interface component comprising each term comprising the positive medical attribute and the medical entity title;

transmit the medical dashboard interface component to a user device to configure a graphical user interface of the user device; and

generate, based on the medical dashboard interface component, a secure medical dashboard on the user device.

2 . The system of claim 1 , wherein the pre-labeled medical entity title for the at least one previous term comprises at least one of an expert label or a proficient label.

3 . The system of claim 1 , wherein the at least one data input comprises identified data or de-identified data.

4 . The system of claim 1 , wherein the at least one data input comprises a plurality of data inputs from a plurality of data sources.

5 . The system of claim 1 , wherein the system further comprises:

identify at least one patient attribute based on the at least one data input, wherein the at least one patient attribute is associated with a patient;

receive an at least one inclusion requirement or at least one exclusion requirement, wherein the at least one inclusion requirement or the at least one exclusion requirement is associated with the at least one patient attribute; and

compare the at least one patient attribute and the at least one inclusion requirement or the at least one exclusion requirement.

6 . The system of claim 5 , wherein the at least one patient attribute comprises at least one of the medical entity title, a patient biological attribute, or a patient date.

7 . The system of claim 5 , wherein the system further comprises:

generate a record identifier associated with the patient;

generate, based on the comparison of the at least one patient attribute, an applicability rating of the patient; and

generate the medical dashboard interface component comprising the at least one term comprising the positive medical attribute, the record identifier, the applicability rating, the at least one patient attribute, and the at least one inclusion requirement or the at least one exclusion requirement.

8 . The system of claim 5 , wherein the system further comprises:

generate a provider identifier associated with the patient;

determine, based on the provider identifier, a plurality of patient attributes associated with a plurality of current patients, wherein the plurality of current patients are associated with a provider of the provider identifier;

receive the at least one inclusion requirement or the at least one exclusion requirement, wherein the at least one inclusion requirement or the at least one exclusion requirement is associated with at least one patient attribute of the plurality of patient attributes;

generate, based on the plurality of patient attributes associated with the plurality of current patients, a provider index for the provider identifier; and

generate the medical dashboard interface component comprising the at least one inclusion requirement or the at least one exclusion requirement, the provider identifier, and the provider index.

9 . The system of claim 8 , wherein the provider index comprises a dynamic patient total for the at least one inclusion requirement or the at least one exclusion requirement, wherein the dynamic patient total is based on the plurality of patient attributes compared to the at least one inclusion requirement or the at least one exclusion requirement.

10 . The system of claim 1 , wherein the system further comprises:

determine a user identifier or an entity identifier associated with the user device; and

dynamically generate, based on the user identifier or the entity identifier, the medical dashboard interface component,

wherein, in an instance the entity identifier is a trial provider entity, generate the medical dashboard interface component comprising at least one inclusion requirement, at least one exclusion requirement, a patient total, a medical provider identifier, or an applicability rating of one or more patients or one or more medical provider identifiers,

wherein, in an instance the user identifier is a patient identifier, generate the medical dashboard interface component comprising patient identifying data associated with the patient identifier or a record identifier, or

wherein, in an instance the entity identifier is a medical provider, generate the medical dashboard interface component comprising the record identifier for one or more patients associated with the medical provider and patient identifying data for the one or more patients.

11 . A computer program product for generating a secure medical dashboard using natural language processing (NLP), wherein the computer program product comprises at least one non-transitory computer-readable medium having computer-readable program code portions embodied therein, the computer-readable program code portions which when executed by a processing device are configured to cause the processor to:

identify at least one data input, wherein the at least one data input comprises unstructured data;

apply at least one trained NLP pipeline to the at least one data input, wherein the at least one trained NLP pipeline is trained using the following operations:

identifying at least one previous unstructured dataset, wherein the at least one previous unstructured dataset comprises an at least one previous term;

applying the at least one NLP pipeline to the at least one previous unstructured dataset;

determining, by the at least one NLP pipeline, the medical relevance attribute for the at least one previous term;

mapping, by the at least one NLP pipeline, the at least one previous term comprising the positive medical attribute to the medical entity title;

comparing the medical entity title for the at least one previous term to a pre-labeled medical entity title for the at least one previous term;

comparing the positive medical attribute for the at least one previous term to a pre-labeled medical relevance attribute;

determining whether to adjust at least one weight associated with the at least one NLP pipeline based on the comparison of the medical entity title to the pre-labeled medical entity title and the comparison of the positive medical attribute to the pre-labeled medical relevance attribute,

adjust, in an instance where at least one of the medical entity title does not match the pre-labeled medical entity title or the medical attribute does not match the pre-labeled medical relevance attribute, the at least one weight of the at least one NLP pipeline, or

maintain, in an instance where the medical entity title matches the pre-labeled medical entity title and the medical attribute matches the pre-labeled medical relevance attribute, the at least one weight of the at least one NLP pipeline;

training, in response to the determination of whether to adjust the at least one weight, the at least one NLP pipeline based on the at least one weight of the at least one NLP pipeline;

parse, by the at least one trained NLP pipeline, the unstructured data of the at least one data input to generate a parsed unstructured dataset, wherein the parsed unstructured dataset comprises a plurality of terms;

identify, by the at least one trained NLP pipeline, a medical relevance attribute for each term within the unstructured dataset, wherein the medical relevance attribute comprises at least one of a positive medical attribute, a medical modifier attribute, or a negative medical attribute;

generate, based on the medical relevance attribute for each term, a structured document comprising each term and associated medical relevance attribute comprising the positive medical attribute or the medical modifier attribute;

correlate, by the at least one trained NLP pipeline, each term comprising the positive medical attribute to a medical entity title;

generate a medical dashboard interface component comprising each term comprising the positive medical attribute and the medical entity title;

transmit the medical dashboard interface component to a user device to configure a graphical user interface of the user device; and

generate, based on the medical dashboard interface component, a secure medical dashboard on the user device.

12 . The computer program product of claim 11 , wherein the processing device is configured to cause the processor to:

identify at least one patient attribute based on the at least one data input, wherein the at least one patient attribute is associated with a patient;

receive an at least one inclusion requirement or an at least one exclusion requirement, wherein the at least one inclusion requirement or the at least one exclusion requirement is associated with the at least one patient attribute; and

compare the at least one patient attribute and the at least one inclusion requirement or the at least one exclusion requirement.

13 . The computer program product of claim 12 , wherein the processing device is configured to cause the processor to:

generate a record identifier associated with the patient;

generate, based on the comparison of the at least one patient attribute, an applicability rating of the patient; and

generate the medical dashboard interface component comprising the at least one term comprising the positive medical attribute, the record identifier, the applicability rating, the at least one patient attribute, and the at least one inclusion requirement or the at least one exclusion requirement.

14 . The computer program product of claim 12 , wherein the processing device is configured to cause the processor to:

generate a provider identifier associated with the patient;

determine, based on the provider identifier, a plurality of patient attributes associated with a plurality of current patients, wherein the plurality of current patients are associated with a provider of the provider identifier;

receive an at least one inclusion requirement or an at least one exclusion requirement, wherein the at least one inclusion requirement or the at least one exclusion requirement is associated with at least one patient attribute of the plurality of patient attributes;

generate, based on the plurality of patient attributes associated with the plurality of current patients, a provider index for the provider identifier; and

generate the medical dashboard interface component comprising the at least one inclusion requirement or the at least one exclusion requirement, the provider identifier, and the provider index.

15 . A computer-implemented method for generating a secure medical dashboard using natural language processing (NLP), the computer-implemented method comprising:

identifying at least one data input, wherein the at least one data input comprises unstructured data;

applying at least one trained NLP pipeline to the at least one data input, wherein the at least one trained NLP pipeline is trained using the following operations:

identifying at least one previous unstructured dataset, wherein the at least one previous unstructured dataset comprises an at least one previous term;

applying the at least one NLP pipeline to the at least one previous unstructured dataset;

determining, by the at least one NLP pipeline, the medical relevance attribute for the at least one previous term;

mapping, by the at least one NLP pipeline, the at least one previous term comprising the positive medical attribute to the medical entity title;

comparing the medical entity title for the at least one previous term to a pre-labeled medical entity title for the at least one previous term;

comparing the positive medical attribute for the at least one previous term to a pre-labeled medical relevance attribute;

determining whether to adjust at least one weight associated with the at least one NLP pipeline based on the comparison of the medical entity title to the pre-labeled medical entity title and the comparison of the positive medical attribute to the pre-labeled medical relevance attribute,

adjust, in an instance where at least one of the medical entity title does not match the pre-labeled medical entity title or the medical attribute does not match the pre-labeled medical relevance attribute, the at least one weight of the at least one NLP pipeline, or

maintain, in an instance where the medical entity title matches the pre-labeled medical entity title and the medical attribute matches the pre-labeled medical relevance attribute, the at least one weight of the at least one NLP pipeline:

training, in response to the determination of whether to adjust the at least one weight, the at least one NLP pipeline based on the at least one weight of the at least one NLP pipeline;

parsing, by the at least one trained NLP pipeline, the unstructured data of the at least one data input to generate a parsed unstructured dataset, wherein the parsed unstructured dataset comprises a plurality of terms;

identifying, by the at least one trained NLP pipeline, a medical relevance attribute for each term within the unstructured dataset, wherein the medical relevance attribute comprises at least one of a positive medical attribute, a medical modifier attribute, or a negative medical attribute;

generating, based on the medical relevance attribute for each term, a structured document comprising each term and associated medical relevance attribute comprising the positive medical attribute or the medical modifier attribute;

correlating, by the at least one trained NLP pipeline, each term comprising the positive medical attribute to a medical entity title;

generating a medical dashboard interface component comprising each term comprising the positive medical attribute and the medical entity title;

transmitting the medical dashboard interface component to a user device to configure a graphical user interface of the user device; and

generating, based on the medical dashboard interface component, a secure medical dashboard on the user device.

16 . The computer-implemented method of claim 15 , further comprising:

identifying at least one patient attribute based on the at least one data input, wherein the at least one patient attribute is associated with a patient;

receiving an at least one inclusion requirement or an at least one exclusion requirement, wherein the at least one inclusion requirement or the at least one exclusion requirement is associated with the at least one patient attribute; and

comparing the at least one patient attribute and the at least one inclusion requirement or the at least one exclusion requirement.

17 . The computer-implemented method of claim 16 , further comprising:

generating a provider identifier associated with the patient;

determining, based on the provider identifier, a plurality of patient attributes associated with a plurality of current patients, wherein the plurality of current patients are associated with a provider of the provider identifier;

receiving an at least one inclusion requirement or an at least one exclusion requirement, wherein the at least one inclusion requirement or the at least one exclusion requirement is associated with at least one patient attribute of the plurality of patient attributes;

generating, based on the plurality of patient attributes associated with the plurality of current patients, a provider index for the provider identifier; and

generating the medical dashboard interface component comprising the at least one inclusion requirement or the at least one exclusion requirement, the provider identifier, and the provider index.