Cancer cell recognition method, device and system

A recognition method and a technology of a recognition device, which are applied in the medical field, can solve problems such as low recognition rate, out-of-control division, and difficulty in curing malignant tumors, and achieve the effect of reducing background noise and difficulty

Pending Publication Date: 2019-09-20
SOUTH UNIVERSITY OF SCIENCE AND TECHNOLOGY OF CHINA
View PDF9 Cites 15 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

In addition to the uncontrolled division of cancer cells (capable of multipolar division), cancer cells can also partially invade the surrounding normal tissues and even transfer to other parts of the body through the circulatory system or lymphatic system in the body, resulting in malignant tumors that are difficult to cure in the human body; the cure of existing cancer One of the main reasons for the low rate is that the period of cancer discovery is too late. Therefore, if the signs of cancer can be found in advance, and then reasonable treatment methods can be adopted, the treatment rate of cancer can be significantly improved
[0003] The existing algorithms for cancer cell identific

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
  • Cancer cell recognition method, device and system
  • Cancer cell recognition method, device and system
  • Cancer cell recognition method, device and system

Examples

Experimental program
Comparison scheme
Effect test

Embodiment 1

[0032] With the development of global high-tech, smart medical care and safe treatment have become the focus of every country. One of the key points in the infrastructure required to realize this concept is that the ubiquitous artificial intelligence network will likely be embedded in Every medical device, including positron-emission CT, MRI, and electron microscopy systems, is then used to detect abnormal cells in the pictures. By detecting various forms of cells to distinguish cell types, the detected targets are simple cell tissues, for example, there are few cell types, and there is not much cell overlap, but in the face of more cell types, cell overlap is not easy to judge When it comes to cell tissue, the recognition rate is not high. In order to avoid misjudgment and accidents, we not only need to detect the size of the cells, but also predict the shape of the cells and the distribution density in the tissue. When the doctor gets the picture, he can see more prompt info...

Embodiment 2

[0051] The purpose of this embodiment is to further illustrate the principles / steps of the pre-order training work involved in the embodiment and the subsequent actual application process.

[0052] Establishment and training of recognition network for identifying cancer cells, including:

[0053] S01. Manually process the microscope video of the tissue section frame by frame, manually calibrate the area where cells appear in each picture, and obtain a calibration map for each cell, and each pixel in the calibration map can only be 0 or 1 ( 0 means that the pixel does not belong to the cell, and 1 means that the pixel belongs to the cell), the video image file number of the same tissue slice sample is V, and the number of the cell is N, then the corresponding calibration map is named V_N, and finally Obtain the calibration image set Ti for the image pi, that is, establish a training sample for a frame of image

[0054] S02, repeat S01, calibrate M-frame images, for example, M>...

Embodiment 3

[0074] This embodiment provides as figure 2 A cancer cell identification device shown includes: a feature processing unit 1, configured to process a cell image to be detected based on a residual network to obtain a first feature map; a cell identification unit 2, configured to process the image based on a region proposal network The first feature map obtains a set of elliptical cell candidate frames; the cancer cell labeling unit 3 is configured to process the set of cell candidate frames based on ROIAlign to obtain a second feature map, and process the first feature map through a preset fully connected network. Two feature maps to obtain cancer cell detection frames and cancer cell codes for distinguishing cancer cells; cancer cell identification unit 4, for processing the cancer cell detection frames based on a preset deep learning network to output for marking cancer cell types type information.

[0075] The cancer cell labeling unit 3 is configured to process the second ...

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 a cancer cell recognition method, device and system. The method comprises the steps of processing a cell image to obtain a first feature map; processing the first feature map to obtain a set; processing the set to obtain a second feature map, and processing the second feature map to obtain a cancer cell detection frame and a cancer cell code; processing the cancer cell detection frame is processed based on a deep learning network to output category information. According to the method, the cell image is processed to obtain the collection of the elliptical cell candidate frames, and the elliptical cell candidate frames can reduce background noise generated during subsequent processing compared with a normal rectangular frame; processing the set is processed to obtain a cancer cell detection frame and cancer cell codes, the segmenting cancer cells are segmented for type analysis, and reducing the subsequent image processing difficulty is reduced through the cancer cell codes; a cancer cell detection frame is processed to output type information, the type of cancer cells can be recognized through a mature deep learning network, and compared with an existing recognition mode, the cancer cell recognition method can better adapt to cancer cell recognition in a complex cell environment.

Description

technical field [0001] The invention relates to the field of medical technology, in particular to a cancer cell identification method, device and system. Background technique [0002] Cancer cells are mutated cells and the source of cancer. Cancer cells are different from normal cells in that they have three characteristics: infinite proliferation, transformation and easy transfer. They can infinitely proliferate and destroy normal cell tissues. In addition to the uncontrolled division of cancer cells (capable of multipolar division), cancer cells can also partially invade the surrounding normal tissues and even transfer to other parts of the body through the circulatory system or lymphatic system in the body, resulting in malignant tumors that are difficult to cure in the human body; the cure of existing cancer One of the main reasons for the low rate is that the period of cancer discovery is too late. Therefore, if the signs of cancer can be found in advance, and then reas...

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
IPC IPC(8): G06K9/00G06K9/32G06K9/34
CPCG06V20/695G06V20/698G06V10/25G06V10/267
Inventor 刘亮希黄骏史玉回马思清
Owner SOUTH UNIVERSITY OF SCIENCE AND TECHNOLOGY OF CHINA
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