Looking for breakthrough ideas for innovation challenges? Try Patsnap Eureka!

A detection method of urine formed components based on deep learning and context-sensitive

A component detection and deep learning technology, applied in neural learning methods, image analysis, image enhancement and other directions, can solve the problems of re-inspection, easy change, missed detection, etc., to achieve accurate positioning and tracking, fast speed, and reduce labor costs. Effect

Active Publication Date: 2021-11-05
BEIJING XIAOYING TECH CO LTD
View PDF6 Cites 0 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

[0011] The present invention aims to solve the problem of missed detection and re-inspection caused by the easy change of the position of the formed components in the microscope field of view, and provides a method for detecting the formed components of urine based on deep learning and context. Based on the Transformer technology, the Based on the Kalman filter target tracking technology applied to urine cell counting, fully considering the long-tail distribution, light and other interference factors, to a certain extent to solve the above problems to achieve a trade-off between speed and accuracy, has a strong clinical application value

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
  • A detection method of urine formed components based on deep learning and context-sensitive
  • A detection method of urine formed components based on deep learning and context-sensitive
  • A detection method of urine formed components based on deep learning and context-sensitive

Examples

Experimental program
Comparison scheme
Effect test

Embodiment 1

[0104] like figure 1 As shown, a deep learning-based, context-sensitive urine formed component detection method includes the following steps:

[0105] S1. Establish a data set of formed components: collect video of the urine sample for labeling under the field of view of the microscopic imaging equipment, and obtain the urine video for labeling; extract images from the urine video for labeling, and obtain a time series n single-frame image data; mark the position and category of the formed elements on the single-frame image data, obtain the formed element data set, and divide the formed element data set into a training set and a test set;

[0106] S2. Establish a detection and classification model for urine formed components: build a detection and classification model for urine formed components with the functions of detection, tracking, prediction, contextual correlation matching, positioning and comparison recognition, and use training sets and test sets for model training;...

Embodiment 2

[0109] like figure 1 , 2 , 3, a deep learning-based, context-sensitive urine formed component detection method, comprising the following steps:

[0110] S1. Establish a data set of formed components: collect video of the urine sample for labeling under the field of view of the microscopic imaging equipment, and obtain the urine video for labeling; extract images from the urine video for labeling, and obtain a time series n single-frame image data; mark the position and category of the formed elements on the single-frame image data, obtain the formed element data set, and divide the formed element data set into a training set and a test set;

[0111] S2. Establish a detection and classification model for urine formed components: build a detection and classification model for urine formed components with the functions of detection, tracking, prediction, contextual correlation matching, positioning and comparison recognition, and use training sets and test sets for model traini...

Embodiment 3

[0146] like Figure 2-3 As shown, a deep learning-based, context-sensitive urine formed component detection method includes the following steps:

[0147] 1. Real-time collection of urine dynamic video under the microscope

[0148] First, each field of view is photographed and collected under the microscope imaging device to generate a time series of single field of view images.

[0149] 2. Urine Formed Component Data Labeling

[0150] Professional doctors use specific labeling tools to mark the position and category information of the formed components on the form. After the amount of labeling reaches a certain scale, the labeled data is divided into a training set and a test set according to a certain proportion to prepare data for building a neural network model.

[0151] 3. Transformer-based urine formed component detection, tracking, and recognition model construction

[0152] Construct a context-sensitive urine formed component detection, tracking, and identification a...

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 present invention provides a method for detecting urine formed components based on deep learning and context-related, which collects urine samples in the field of view of a microscopic imaging device, extracts images from urine videos, and obtains time-series images. Single-frame image data, marking the position and category of the formed components of the single-frame image data, and dividing it into a training set and a test set; constructing a urine effective detection, tracking, prediction, context matching, positioning, and comparison recognition functions. Form a component detection and classification model, use the training set and test set for model training; then detect the urine sample to be detected. The invention solves the problem of missed detection and repeated detection due to the easy change of the position of the formed components of urine under the microscope field of view. Based on the Transformer technology, the target tracking technology based on the Kalman filter is applied to urine cell counting, and the long-tail distribution is fully considered. Interfering factors such as illumination can solve the above problems while achieving a balance between speed and accuracy, which has clinical application value.

Description

technical field [0001] The invention relates to the technical fields of calculation, calculation and counting, in particular to a method for detecting urine formed components based on deep learning and context-dependent. Background technique [0002] Urine formed components refer to the general term for substances that come from the urinary system and are formed by exudation, excretion, shedding, and concentrated crystallization in a visible form. Urine formed element analysis refers to the examination of urine formed elements with a microscope or special equipment to identify various pathological elements in urine such as cells, casts, crystals, bacteria, parasites, etc., to assist in the diagnosis of urinary system diseases, Localization, differential diagnosis and prognosis. [0003] The detection of urine formed components belongs to the detection of flowing liquid, and the position of formed components in each frame may change under the same microscopic imaging field o...

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): G06T7/00G06T7/246G06T7/66G06T7/73G06K9/62G06N3/04G06N3/08
CPCG06T7/0012G06T7/246G06T7/66G06T7/73G06N3/08G06T2207/10016G06T2207/10056G06T2207/30004G06T2207/20081G06T2207/20084G06N3/045G06F18/241
Inventor 李柏蕤连荷清
Owner BEIJING XIAOYING TECH CO LTD
Who we serve
  • R&D Engineer
  • R&D Manager
  • IP Professional
Why Patsnap Eureka
  • Industry Leading Data Capabilities
  • Powerful AI technology
  • Patent DNA Extraction
Social media
Patsnap Eureka Blog
Learn More
PatSnap group products

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

© 2024 PatSnap. All rights reserved.Legal|Privacy policy|Modern Slavery Act Transparency Statement|Sitemap|About US| Contact US: help@patsnap.com