Time-series image-based prediction method and device

An image and timing technology, applied in the field of image recognition, can solve problems such as inaccurate prediction of fundus image results, difficult features, and unfixed sampling intervals, and achieve the effect of overcoming uniform timing sampling

Active Publication Date: 2022-06-03
SHENZHEN NEW IND MATERIAL OF OPHTHALMOLOGYCO
View PDF8 Cites 0 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

[0007] The embodiment of the present invention provides a prediction method and device based on time-series images to at least solve the problem of fundus image result prediction caused by difficult extraction of features, unbalanced sample ratio and unfixed sampling interval in the fundus image prediction process in the related art. inaccurate technical issues

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
  • Time-series image-based prediction method and device
  • Time-series image-based prediction method and device
  • Time-series image-based prediction method and device

Examples

Experimental program
Comparison scheme
Effect test

Embodiment 1

[0031] According to an embodiment of the present invention, a method for predicting fundus images based on time-series images is provided, such as figure 1 As shown, the method includes:

[0032] S102, acquiring a fundus image sequence, wherein the fundus image sequence includes a plurality of fundus images sorted by time;

[0033] S104, input the fundus image sequence into the pre-trained fundus image prediction model to obtain the prediction result, wherein the fundus image prediction model is used to determine the prediction result based on the image features and time series features respectively corresponding to the fundus image sequence, and the fundus image The predictive model is trained on a dataset of fundus image sequences with eigenvalues,

[0034] In a specific application scenario, in the fundus image sequence, the time intervals between adjacent fundus images of multiple fundus images can be the same or different. For example, fundus image X 1 、X 2 , X 3 、X ...

Embodiment 2

[0093] According to an embodiment of the present invention, there is also provided a time-series image-based fundus image prediction device for implementing the above-mentioned time-series image-based fundus image prediction method, such as Figure 5 As shown, the device includes:

[0094] 1) acquisition unit 50, used to acquire a sequence of fundus images, wherein the sequence of fundus images includes a plurality of fundus images sorted by time;

[0095] 2) Prediction unit 52, configured to input the fundus image sequence into the pre-trained fundus image prediction model to obtain a prediction result, wherein the fundus image prediction model is used to respectively correspond to The image features and time series features of the fundus image prediction model are obtained by training according to the data set of the fundus image sequence with eigenvalues.

[0096] Optionally, for a specific example in this embodiment, refer to the example described in Embodiment 1 above, ...

Embodiment 3

[0098] According to an embodiment of the present invention, a time-series image-based fundus image prediction model is also provided. Preferably, in this embodiment, the fundus image model is obtained by training a training data set composed of fundus image sequences containing multiple groups of fundus images , a model for predicting fundus image sequences with different time series, such as figure 2 As shown, the fundus image prediction model includes: an image processing unit 20, a time processing unit 22, and a classification unit 24, wherein:

[0099] 1) Image processing unit 20, for obtaining the corresponding spatial features of the fundus image according to the image features of the fundus image sequence, wherein the fundus image sequence includes a plurality of fundus images sorted by time;

[0100]2) time processing unit 22, for obtaining the corresponding spatio-temporal feature of the fundus image according to the time difference value of the fundus image of the s...

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

Embodiments of the present invention relate to a prediction method and device based on time-series images. Wherein, the method includes: taking a fundus image sequence, wherein the fundus image sequence includes a plurality of fundus images sorted by time; inputting the fundus image sequence into a pre-trained fundus image prediction model to obtain a prediction result, wherein, The fundus image prediction model is used to determine the prediction results based on the corresponding image features and time series features of the fundus image sequences, and the fundus image prediction model is trained according to the data set of the fundus image sequences with eigenvalues. The invention solves the technical problem of inaccurate prediction of fundus image results due to the low accuracy and sensitivity of the fundus image prediction network in the related art and the different temporal intervals of the fundus images.

Description

technical field [0001] The present invention relates to the field of image recognition, in particular to a prediction method and device based on time-series images. Background technique [0002] As the most important blinding irreversible ophthalmic disease, glaucoma has an incidence rate of about 3.5% among people aged 45 and over. Aging, an estimated 110 million people suffer from glaucoma. Early detection of diseases is a very important part of medical diagnosis. Statistics show that 11% of hospital deaths are due to lack of timely diagnosis and treatment. Therefore, for disease screening and prevention, future disease prediction algorithms based on time series information are more important. In recent years, some work has tried to predict the disease, but it mainly faces the following three problems: difficult feature extraction, unbalanced sample ratio and unfixed sampling interval. [0003] For disease prediction tasks, there have also been many disease prediction ...

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 Patents(China)
IPC IPC(8): G06Q10/04G06K9/62G06N3/04G06T7/90G06V10/764G06V10/774G06V10/82G06V10/25
CPCG06Q10/04G06T7/90G06V10/25G06N3/045G06F18/24G06F18/2411G06F18/24155G06F18/214
Inventor 徐迈李柳
Owner SHENZHEN NEW IND MATERIAL OF OPHTHALMOLOGYCO
Who we serve
  • R&D Engineer
  • R&D Manager
  • IP Professional
Why Eureka
  • Industry Leading Data Capabilities
  • Powerful AI technology
  • Patent DNA Extraction
Social media
Try Eureka
PatSnap group products

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

© 2024 PatSnap. All rights reserved.Legal|Privacy policy|Modern Slavery Act Transparency Statement|Sitemap