Method and system for predicting medical re-seeking information based on cross-attention neural network
A neural network and diagnostic information technology, applied in the field of information processing, can solve problems that affect prediction results and interfere with patient judgment, and achieve the effects of improving prediction accuracy, ensuring integrity and independence, and good interpretability
Active Publication Date: 2020-11-20
浩睿智源山东人工智能有限公司
View PDF3 Cites 0 Cited by
- Summary
- Abstract
- Description
- Claims
- Application Information
AI Technical Summary
Problems solved by technology
Mixed diagnosis and treatment information will interfere with the judgment of the patient's current disease and affect the prediction effect
Method used
the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
View moreImage
Smart Image Click on the blue labels to locate them in the text.
Smart ImageViewing Examples
Examples
Experimental program
Comparison scheme
Effect test
Embodiment Construction
[0037] It should be noted that the following detailed description is exemplary and intended to provide further explanation of the present disclosure. Unless defined otherwise, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this disclosure belongs.
[0038] It should be noted that the terminology used herein is only for describing specific embodiments, and is not intended to limit the exemplary embodiments according to the present disclosure. As used herein, unless the context clearly dictates otherwise, the singular is intended to include the plural, and it should also be understood that when the terms "comprising" and / or "comprising" are used in this specification, they mean There are features, steps, operations, means, components and / or combinations thereof.
[0039] Separating diagnosis and treatment can better grasp the patient's condition changes and analyze the treatment process. ...
the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
Login to View More PUM
Login to View More Abstract
This disclosure proposes a method and system for predicting medical re-seeking information based on a cross-attention neural network, to obtain historical electronic health record data of patients; to split each patient's multi-variable time-stamped sequence of a medical-seeking information record into diagnostic information and treatment information. information, and perform dimensionality reduction representation on diagnostic information and treatment information respectively; use bidirectional neural network to process dimensionality-reduced diagnostic information to obtain the corresponding hidden state of diagnostic information, and use bidirectional neural network to process dimensionality-reduced treatment information to obtain corresponding treatment information Hidden state; use the cross-attention mechanism to integrate historical diagnosis information and historical treatment information into a context vector that can represent the patient's current state; after obtaining the patient's disease diagnosis information and treatment information, connect the two parts of the context vector The information is combined into a representation vector representing the overall information of the patient; the representation vector is put into the final output layer to predict medical information.
Description
technical field [0001] The present disclosure relates to the technical field of information processing, in particular to a method and system for predicting re-seeking information based on a cross-attention neural network. Background technique [0002] The purpose of analyzing patients' health information is to help people prevent diseases as early as possible and guide treatment methods. Therefore, it is an important task to predict the next medical coding (including disease and treatment) through the patient's historical Electronic Health Record (EHR) data. How to model the temporality and high dimensionality of continuous EHR data and interpret the prediction results is a key issue in accomplishing this task. [0003] The inventors found in the research that existing methods solve these problems by using a Recurrent Neural Network (RNN) to model EHR data and using an attention mechanism to provide interpretability. Previous models mixed treatment and diagnosis informatio...
Claims
the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
Login to View More Application Information
Patent Timeline
Login to View More Patent Type & Authority Patents(China)
IPC IPC(8): G16H50/20G16H10/60G16H50/70G06Q10/04G06K9/62G06N3/04G06N3/08
CPCG16H50/20G16H10/60G16H50/70G06Q10/04G06N3/049G06N3/084G06N3/045G06F18/213G06F18/2414
Inventor 郭伟葛伟任艺琴刘静崔立真
Owner 浩睿智源山东人工智能有限公司