Breast cancer molecular typing method, device and system based on unsupervised learning

An unsupervised learning and molecular typing technology, applied in the field of computer-aided medicine, can solve problems such as reducing the timeliness of molecular typing detection, achieve the effects of improving accuracy, reducing negative transfer, and improving prediction accuracy

Pending Publication Date: 2021-11-12
CENT HOSPITAL TAIAN CITY +1
View PDF0 Cites 4 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

However, in this type of method, in order to ensure accurate and effective delineation of the region of interest and optimization of fe

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
  • Breast cancer molecular typing method, device and system based on unsupervised learning
  • Breast cancer molecular typing method, device and system based on unsupervised learning
  • Breast cancer molecular typing method, device and system based on unsupervised learning

Examples

Experimental program
Comparison scheme
Effect test

Embodiment 1

[0050] refer to figure 1 As shown, the present embodiment provides a method for molecular typing of breast cancer based on unsupervised learning, which includes the following steps:

[0051] Step S1, obtaining the DCE-MRI image of the breast to be predicted, and extracting the region of interest of sequence images of various specifications in the image. In this embodiment, three sequence images of DCE-MRI TPs 1, TPs 2, and TPs 3 are used;

[0052]Step S2, using a molecular typing prediction model obtained by unsupervised learning training to predict and obtain the corresponding molecular subtype classification probabilities of various sequence images;

[0053] Step S3, using ensemble learning fusion to obtain the final corresponding molecular subtype classification results.

[0054] In the above step S2, the training process of the molecular typing prediction model specifically includes the following steps:

[0055] Step S201, obtain breast DCE-MRI images for training, and f...

Embodiment 2

[0081] This embodiment provides a breast cancer molecular classification prediction device based on unsupervised learning, including: a training data set acquisition module, which acquires breast DCE-MRI images for training, and forms mutually disjoint information according to the division of benign and malignant lesions in the images. A source domain data set and a target domain data set, the source domain data set contains unlabeled samples, and the target domain data set contains labeled samples; the region of interest extraction module is used to extract multiple specifications The region of interest of the breast DCE-MRI sequence image; the unsupervised learning pre-training module uses the region of interest of multiple sequence images obtained in the source domain data set to perform unsupervised learning on the constructed molecular typing prediction model Pre-training to obtain model weights; transfer learning fine-tuning module, using the regions of interest of multip...

Embodiment 3

[0083] The present embodiment provides a computer system for predicting breast cancer molecular typing based on unsupervised learning, including: one or more command processors and memory associated with the processor; wherein, the command processing When the device is executed, the program instructions in the memory are called to implement the steps in the method described in Embodiment 1.

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 relates to a breast cancer molecular typing method, device and system based on unsupervised learning. The method comprises the following steps: obtaining a to-be-predicted breast DCE-MRI image, and extracting regions of interest of sequence images of various specifications in the image; utilizing a molecular subtype prediction model obtained by adopting unsupervised learning training to predict and obtain molecular subtype classification probabilities corresponding to various sequence images; adopting ensemble learning fusion to obtain a final corresponding molecular subtype classification result. When the molecular typing prediction model is trained, through the thought of an unsupervised learning pre-training network and a transfer learning fine tuning network, a breast benign tumor image is fully utilized to construct an unlabeled source domain data set in the previous stage, and the feature extraction capability of the model is enhanced; and in the later stage, a target domain data set with a label is constructed by adopting the breast malignant tumor image to carry out fine tuning on the model with the pre-training weight. Compared with the prior art, the prediction precision of breast cancer molecular typing is remarkably improved.

Description

technical field [0001] The invention relates to the field of computer-aided medicine, in particular to a breast cancer molecular typing method, device and system based on unsupervised learning. Background technique [0002] According to the latest data from the International Agency for Research on Cancer of the World Health Organization in 2021, breast cancer has replaced lung cancer as the cancer with the highest incidence in the world. As a highly heterogeneous malignant tumor, even if patients have the same clinical stage and pathological type, there are great differences in treatment efficacy and prognosis. Currently, immunohistochemical marker technology or gene expression profiling has become the main way to accurately determine the molecular subtype of breast cancer. However, this operation method is complicated and invasive, and fails to achieve "early detection, early diagnosis, and early treatment". The vigorous development of modern medical imaging has created fa...

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): G06T7/00G06K9/32G06K9/62G06N3/04G06N3/08
CPCG06T7/0012G06N3/08G06T2207/10096G06T2207/20081G06T2207/20084G06T2207/20104G06T2207/30068G06N3/045G06F18/2415G06F18/214
Inventor 谢元忠聂生东孙榕李秀娟孔雪
Owner CENT HOSPITAL TAIAN CITY
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