Login| Sign Up| Help| Contact|

Patent Searching and Data


Title:
NOVEL CORONAVIRUS PATIENT CONDITION CLASSIFICATION SYSTEM BASED ON ARTIFICIAL INTELLIGENCE
Document Type and Number:
WIPO Patent Application WO/2021/179514
Kind Code:
A1
Abstract:
A novel coronavirus patient condition classification system based on artificial intelligence, relating to the field of condition classification and comprising: a classification model obtaining module, configured to train one or more dichotomy models for classifying patient conditions according to patient data, obtain a dichotomy model with the highest accuracy from the dichotomy models as a target model, and determine interpretable features in the patient data; a preprocessing module, configured to extract the interpretable features in the patient data to be classified and then preprocess the extracted features to fill missing values in the extracted features and replace abnormal values in the extracted features, so that features to be classified are obtained after the preprocessing; and a condition classification module, configured to complete condition classification of patients to be classified using the target model by taking said features as input of the target model. The classification system satisfies the requirements of stability, robustness and accuracy, and the problem of novel coronavirus patient condition classification can be solved.

Inventors:
YUAN YE (CN)
SUN CHUAN (CN)
YAN LI (CN)
XU HUI (CN)
WANG MAOLIN (CN)
GUO YUQI (CN)
TANG XIUCHUAN (CN)
ZHANG HAITAO (CN)
XIAO YANG (CN)
Application Number:
PCT/CN2020/105477
Publication Date:
September 16, 2021
Filing Date:
July 29, 2020
Export Citation:
Click for automatic bibliography generation   Help
Assignee:
UNIV HUAZHONG SCIENCE TECH (CN)
International Classes:
G16H50/80
Foreign References:
CN111446002A2020-07-24
CN109785976A2019-05-21
CN110825819A2020-02-21
CN109119167A2019-01-01
CN107403072A2017-11-28
US20140270493A12014-09-18
Attorney, Agent or Firm:
WUHAN HUAZHIYU INTELLECTUAL PROPERTY AGENCY CO., LTD. (CN)
Download PDF: