| CPC G16H 80/00 (2018.01) [G06F 40/40 (2020.01); G16H 10/60 (2018.01); G16H 20/00 (2018.01); G16H 70/00 (2018.01)] | 16 Claims |
|
1. A method for generating a medical knowledge packet in an active conversation session, the method comprising:
receiving, by a processor, a message associated with a conversation thread;
monitoring, by the processor, the conversation thread to identify one or more keywords indicating user's intent to seek medical knowledge whereby upon identification, triggering the processor to determine a context of the message associated with the conversation thread;
determining, by the processor, the context of the message using a machine learning model based on identification of the one or more keywords in the conversation thread, wherein the machine learning includes a large language model, and wherein the large language model is fine-tuned by providing a domain-specific dataset comprising a medical text, a medical conversation, a medical knowledge source, medical question answering data, a medical record and a medical report;
generating, by the processor, one or more medical knowledge packets, from one more sources based on the message and the context,
wherein the one or more data sources comprises a prestored medical library comprising a collection of medical resources and external data sources including third-party software, communicatively coupled with the processor, maintaining patient history records, and wherein the prestored medical library comprises medical books, medical guidelines, medical calculator, medical procedures reference, medical skills information, video, and therapeutic algorithm, and wherein the medical knowledge packet comprises at least one of a medical condition, a medication, a treatment plan, a diagnostic protocol, and a medical definition, and wherein the prestored medical library is updated based on at least one of:
when a new medical knowledge packet is available; and
a feedback associated with the one or more factors of previously displayed medical knowledge packets;
assigning, by the processor, a confidence score to the medical knowledge packet from the one or more medical knowledge packets based on one or more factors comprising relevance, accuracy, the one or more sources, and recency of the medical knowledge packet;
modifying, by the processor, the medical knowledge packet with a highest confidence score by at least one of formatting, highlighting, and simplifying by using one or more text analysis algorithms; and
displaying, by the processor, the modified medical knowledge packet in the active conversation session.
|