Title:
COMPUTER VISUAL FEATURE-BASED OCT IMAGE CLASSIFICATION METHOD, SYSTEM AND DEVICE
Document Type and Number:
WIPO Patent Application WO/2022/160118
Kind Code:
A1
Abstract:
The present invention provides a computer visual feature-based OCT image classification method, system and device, and a storage medium. Said method comprises: generating a first mask image in an OCT image, and performing an operation on same and the OCT image to acquire a first image of interest; processing the first image of interest to determine a foreground, and converting the foreground into a second mask image and processing same; performing an operation on the processed second mask image and the first image of interest, and extracting a second image of interest; performing feature enhancement processing on the second image of interest to obtain a third image of interest; and classifying the third image of interest to obtain a classification result, acquiring a confidence of the classification result, and displaying the final result on the OCT image. By means of said method, a computer visual preprocessed part is added, and noise of an OCT image is effectively removed, so as to increase the signal-to-noise ratio; and feature extraction and feature enhancement are performed on a sensitive region, and deep learning recognition is performed, so as to improve the overall recognition accuracy, and improve the confidence of the image determination result.
Inventors:
XIANG SHAOHUA (CN)
WEN HUAJIE (CN)
XIAO ZHIYONG (CN)
ZHAO JIAN (CN)
WEN HUAJIE (CN)
XIAO ZHIYONG (CN)
ZHAO JIAN (CN)
Application Number:
PCT/CN2021/073936
Publication Date:
August 04, 2022
Filing Date:
January 27, 2021
Export Citation:
Assignee:
UNIV SHENZHEN TECHNOLOGY (CN)
International Classes:
G06T7/00; G06N3/04; G06T5/00; G06T7/136; G06T7/194
Foreign References:
| CN110428421A | 2019-11-08 | |||
| CN109493954A | 2019-03-19 | |||
| CN111369572A | 2020-07-03 | |||
| CN107392130A | 2017-11-24 |
Other References:
YIBIAO RONG ET AL.: "Surrogate-Assisted Retinal OCT Image Classification Based on Convolutional Neural Networks", IEEE JOURNAL OF BIOMEDICAL AND HEALTH INFORMATICS, vol. 23, no. 1, 31 January 2019 (2019-01-31), XP011695774, ISSN: 2168-2194, DOI: 10.1109/JBHI.2018.2795545
Attorney, Agent or Firm:
HENSEN INTELLECTUAL PROPERTY FIRM (CN)
Download PDF: