| CPC G16H 40/20 (2018.01) [G16H 10/60 (2018.01); G16H 40/67 (2018.01); G16H 50/30 (2018.01)] | 20 Claims |
|
1. A method, comprising:
executing, by at least one processing device of a plurality of computing devices, a plurality of heart disease triage (HDT) microservices operating on a scalable data container orchestration platform;
wherein the scalable data container orchestration platform, comprises at least one load balancer, and is configured to:
dynamically scale a number of computing devices based on data traffic assessed by the at least one load balancer for the plurality of HDT microservices to maximize computational and load efficiencies in the plurality of computing devices,
interact with a plurality of backend computing devices associated with a plurality of providers through an application programming interface (API) layer comprising a plurality of APIs, and
relay data associated with a plurality of patients between any of the plurality of backend computing devices and the plurality of HDT microservices;
wherein the scalable data container orchestration platform is programmed to:
continuously receive over a communication network, from a plurality of electronic resources, a plurality of patient-specific data files associated with the plurality of patients;
wherein each of the plurality of patient-specific data files comprises at least one patient-specific data record associated with a patient from the plurality of patients;
wherein the at least one patient-specific data record comprises:
a medical history record of the patient, and
at least one date of at least one visit to at least one provider that provided cardiac medical care to the patient;
assign each patient-specific data file from the plurality of patient-specific data files to any of the plurality of HDT microservices based on scaling of the number of computing devices;
analyze, via a corresponding HDT application of the plurality of HDT microservices, each patient-specific data file from the plurality of patient-specific data files to:
identify patient-specific data files from the plurality of patient-specific data files comprising at least one patient-specific echocardiogram data record for generating a plurality of identified patient-specific data files respectively associated with a plurality of identified patients, or
ignore the patient-specific data files without the at least one patient-specific echocardiogram data record;
generate for each of the plurality of identified patient-specific data files, a patient-specific tokenized diagnostic feature set using a feature-space transformation software module applied to at least one patient-specific diagnostic text;
wherein the at least one patient-specific diagnostic text in each of the plurality of identified patient-specific data files with text associated with:
the at least one patient-specific echocardiogram data record of the patient,
the at least one patient-specific data record of the patient, or
any combination thereof;
wherein the patient-specific tokenized diagnostic feature set comprises at least one text tokenized feature of the at least one patient-specific diagnostic text based on a feature-space transformation of each word, a phrase of words, or both in the at least one patient-specific diagnostic text;
generate a heart disease (HD) severity score from an output of at least one trained HD severity determination machine learning model based on inputting the patient-specific tokenized diagnostic feature set for each of the plurality of identified patients into the at least one trained HD severity determination machine learning model;
output a triage score from a triage scoring algorithm for each of the plurality of identified patients based at least in part on:
the HD severity score, and
a provider-visit weight based on the at least one date of the at least one visit to the at least one provider;
generate a triage-score patient severity list ranked from a highest triage score to a lowest triage score, for each identified patient associated with each identified patient-specific data file from the plurality of identified patient-specific data files;
wherein the triage-score patient severity list indicates a set of the plurality of identified patients that should first receive urgent care; and
transmit, via the API layer, at least one displaying instruction, to at least one specific backend computing devices from the plurality of backend computing devices associated with at least one specific provider from the plurality of providers that provided cardiac medical care to any of the plurality of identified patients in the set, that causes to display on a graphical user interface (GUI) on at least one display of the specific backend computing devices:
the triage-score patient severity list,
an urgent medical care alert for each identified patient in the set,
or both.
|