3D vascular structure extraction method based on convolutional cycle network

Abstract

The invention provides a 3D vascular structure extraction method based on a convolutional cycle network. Main contents comprise experiment data, integration data, a convolutional neural network (CNN) framework and a convolutional long-term and short-period memory unit (ConvLSTM). The method comprises the following steps that clinic data and synthesis data are used to carry out experiment; actual data adopts a high-resolution fluorescence multiphoton microscope; in a mice, abdominal cavity model imaging is performed to acquire an image so as to describe a vascular system; a noise is added to simulate an actual condition; a CNN network and a convolutional LSTM network are combined; the convolutional long-term and short-period memory unit uses a deep configuration and shallow configuration combination and a training network and uses a weighted binary system cross entropy loss to test a training result. In the invention, from a complex microscope image, a real 3D vascular structure can be extracted, convolution complexity in three dimensions is reduced, a number of model parameters is greatly decreased, a wide view is possessed, and abundant characteristics can be extracted.

Description

technical field [0001] The invention relates to the field of image processing, in particular to a method for extracting 3D blood vessel structures based on a convolutional loop network. Background technique [0002] In the clinical practice of diagnosing and treating cerebrovascular diseases, the use of computers to process 3D cerebrovascular images provides doctors with a 3D structure that can observe blood vessels at any angle, helping doctors to make more accurate diagnoses and treatments for patients . Generally, the conditions of the vasculature and the like are observed by taking images of them, so the quality of the obtained images will affect the doctor's judgment. However, due to the numerous branches and small shapes of blood vessels, and as people's requirements for drawing blood vessel shapes become more and more detailed, how to obtain an accurate description of the blood vessel structure has become a very difficult problem. Previous methods for 3D reconstruct...

AI Technical Summary

Problems solved by technology

[0004] Aiming at the problem of low detail resolution ability, the purpose of the present invention is to provide a method for extracting 3D vessel structure based on convolutional loop network, using clinical data and synthetic data for experiments, combining CNN network with convolutional LSTM network, convolution Long short-term memory cells combine and train networks using deep and shallow configurations, and test training results using weighted binary cross-entropy loss

Images