Cleavage ball cleavage state intelligent detection method and system

Abstract

The invention discloses a cleavage ball cleavage state intelligent detection method and system. The method comprises the following steps: collecting an embryo image in periods D1 to D3 and recording ashooting time point; inputting the embryo image and the time point thereof into a cleavage ball number prediction network model for prediction; calculating a cleavage globule division index F, and judging whether the embryo is normally divided in the cleavage stage or not; and outputting the predicted number of the cleavage balls in each embryo image, the cleavage ball cleavage curve and the cleavage state judgment result of the embryo in the cleavage stage as a detection report. According to the detection method provided by the invention, the whole process of the cleavage ball development ismonitored; the number of the cleavage balls and cleavage ball cleavage points can be accurately detected; a doctor can rapidly and visually observe and analyze the whole cleavage process of the cleavage ball, time and money cost are saved for the doctor and a patient, the prediction accuracy of the number of embryos of the cleavage ball number prediction network model reaches 93%, and the accuracy is far higher than that of the traditional image processing method.

Description

technical field [0001] The invention relates to the technical fields of embryo morphology and artificial intelligence, in particular to an intelligent detection method and system for blastomere division state. Background technique [0002] The quality of embryo development directly affects the result of pregnancy rate. Embryologists rely on embryo morphology and genetics to distinguish the quality of embryos. Among them, the use of genetic methods to determine the quality of embryos requires extremely high experimental conditions. , and using embryo morphological information to complete embryo evaluation is a simple, fast and effective method. At present, most embryologists judge the quality of embryos by obtaining some important morphological feature changes in the embryo development process based on their long-term embryo observation experience, and use it as an important predictor of the success rate of embryo transfer in the later stage. in accordance with. Doctors usu...

AI Technical Summary

Problems solved by technology

[0006] The purpose of the present invention is to overcome the problem of low accuracy and poor real-time performance of the traditional image analysis method in the prior art by extracting simple artificial features to complete the prediction of the number of blastomeres, and propose an intelligent detection method for the state of blastomere division And the system, by building an effective AI model for blastomere number prediction, it can automatically complete the high-level feature extraction, model training and prediction of blastomere in different periods, and its accuracy and recognition efficiency have been greatly improved

Images