Supercharge Your Innovation With Domain-Expert AI Agents!

Enhanced automatic medical diagnosis dialogue system based on deep learning

A medical diagnosis and deep learning technology, applied in the interdisciplinary field, can solve the problems of low accuracy of diagnosis results, medical vocabulary, and insensitivity of medical problems, and achieve the effect of improving the accuracy.

Pending Publication Date: 2022-02-01
HARBIN UNIV OF SCI & TECH
View PDF0 Cites 0 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

[0006] The purpose of the present invention is to solve the problem that the existing system is not sensitive to medical vocabulary and medical problems, which leads to the low accuracy of the output diagnosis result of the existing system, and proposes an enhanced automatic medical diagnosis dialogue system based on deep learning

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 more

Image

Smart Image Click on the blue labels to locate them in the text.
Viewing Examples
Smart Image
  • Enhanced automatic medical diagnosis dialogue system based on deep learning
  • Enhanced automatic medical diagnosis dialogue system based on deep learning
  • Enhanced automatic medical diagnosis dialogue system based on deep learning

Examples

Experimental program
Comparison scheme
Effect test

specific Embodiment approach 1

[0043] Specific implementation mode 1. Combination figure 1 This embodiment will be described. An enhanced automatic medical diagnosis dialogue system based on deep learning described in this embodiment, the system includes a data preprocessing module, a neural network module and a vocabulary-level fusion module; wherein:

[0044] The data preprocessing module is used to preprocess the medical dialog data to be processed to obtain a preprocessing result;

[0045] Find the embedding vector corresponding to each word of the medical dialogue data to be processed from the preprocessing result;

[0046] The neural network module is used to generate N×S text sequences of the medical dialogue data to be processed according to the embedding vector and the position code corresponding to each word in the data to be processed;

[0047] The vocabulary-level fusion module is used to fuse the word sequences generated by the neural network module, and output the fusion result as a dialogue...

specific Embodiment approach 2

[0048] Embodiment 2. The difference between this embodiment and Embodiment 1 is that the neural network module is composed of N decoder blocks and a softmax layer, wherein each decoder block includes a masked multi-head self-attention layer, a normalization layer and a feed-forward neural network layer that takes the output of the last decoder block as input to the softmax layer.

[0049] Other steps and parameters are the same as those in Embodiment 1.

specific Embodiment approach 3

[0050]Specific implementation mode three, this implementation mode is different from one of the specific implementation modes one or two in that: the training process of the neural network module is:

[0051] Step 1. Obtain medical dialogue material data, and then preprocess the acquired corpus data; wherein, the specific process of preprocessing is:

[0052] Step 11, perform data cleaning on the acquired corpus data,

[0053] Step 12. Describe the different types of sentence data after data cleaning under a unified framework;

[0054] Describing to a unified framework means: describing a statement as a triplet of statement, attribute, and attribute value;

[0055] Step 13. Semantically integrate different sentences in the same dialogue (integrate the relationship between the preceding and following sentences), and use the integrated embedding matrix as the semantic feature of the current dialogue;

[0056] In the same way, the semantic features of each dialogue are obtained...

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

No PUM Login to View More

Abstract

The invention discloses an enhanced automatic medical diagnosis dialogue system based on deep learning, and belongs to the field of discipline crossing combining natural language processing and deep learning. According to the invention, the problem that the existing system is insensitive to medical vocabularies and medical problems, so that the accuracy of the diagnosis result output by the existing system is low is solved. According to the method, pre-training and fine tuning are performed on the model by using the medical data set, so that the model can perform more effective modeling on the medical dialogue, and meanwhile, results of multiple systems are reordered through vocabulary-level system fusion so as to enhance the original dialogue result. According to the method, related information can be better captured for questions proposed by users, then corresponding answers are made, and the accuracy of outputting diagnosis results can be improved. The present invention can be applied to automated medical diagnostic conversations.

Description

technical field [0001] The invention belongs to the interdisciplinary field of combining natural language processing and deep learning, and specifically relates to an enhanced automatic medical diagnosis dialogue system based on deep learning. Background technique [0002] Deep learning is to learn the inherent laws and representation levels of data, and the information obtained during these learning processes is of great help to the interpretation of data such as text, images and sounds. The multi-hidden layer deep neural network has excellent feature learning ability, and the learned features have a more essential description of the data. The power of the deep neural network lies in its multi-layer structure that can automatically learn features, and can learn multiple levels of features. Its ultimate goal is to enable machines to have the ability to analyze and learn like humans, and to be able to recognize data such as text, images, and sounds. Deep learning has achiev...

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
no application Login to View More
Patent Type & Authority Applications(China)
IPC IPC(8): G16H80/00G06F40/35G06N3/04G06N3/08
CPCG16H80/00G06F40/35G06N3/08G06N3/047G06N3/044
Inventor 刘宇鹏林明豪杨锦锋张晓晨
Owner HARBIN UNIV OF SCI & TECH
Features
  • R&D
  • Intellectual Property
  • Life Sciences
  • Materials
  • Tech Scout
Why Patsnap Eureka
  • Unparalleled Data Quality
  • Higher Quality Content
  • 60% Fewer Hallucinations
Social media
Patsnap Eureka Blog
Learn More

Browse by: Latest US Patents, China's latest patents, Technical Efficacy Thesaurus, Application Domain, Technology Topic, Popular Technical Reports.

© 2025 PatSnap. All rights reserved.Legal|Privacy policy|Modern Slavery Act Transparency Statement|Sitemap|About US| Contact US: help@patsnap.com