Tissue blood flow and oxyhemoglobin saturation measuring method based on deep learning diffusion correlation spectrum

Abstract

The invention discloses a tissue blood flow and oxyhemoglobin saturation measuring method based on a deep learning diffusion correlation spectrum. The method is used for solving the problem that in the current diffusion correlation spectrum tissue blood flow quantification process, the obtained tissue parameter is unique. The method specifically comprises that light intensity and light intensity autocorrelation function data of a measured tissue are acquired through measurement of a plurality of wavelengths by utilizing a diffusion correlation spectrum technology, an LSTM network model is constrcuted through a deep learning method, time sequence characteristics of light intensity I(rho, lambda) and a light intensity autocorrelation function g2(tau) are extracted through the deep learning method, and the relationship between the time sequence characteristics and the blood flow and the blood oxygen saturation is constructed based on the time sequence characteristics, so that the measurement of the tissue blood flow index and the tissue blood oxygen saturation is achieved. According to the method, the fitting speed is greatly increased while the quantification precision is guaranteed, favorable conditions can be provided for dynamic longitudinal measurement of tissue blood flow, and continuous monitoring of two physiological parameters of blood flow and blood oxygen saturation is achieved.

Description

technical field [0001] The invention relates to the technical field of biomedical engineering, in particular to a method for measuring tissue blood flow and blood oxygen saturation based on deep learning diffusion correlation spectroscopy. Background technique [0002] Tissue blood oxygen metabolism is a very important parameter in biological tissues. This parameter depends on changes in blood flow and blood oxygen saturation in the tissue. Simultaneous detection of both is crucial for the clinical diagnosis of cardiovascular and cerebrovascular diseases. Diffuse Correlation Spectroscopy (DCS) uses long-term coherent light to detect the dynamic characteristics of scattering particles in tissues to provide blood flow changes. It is non-invasive, has a wide range of applications, has low detection requirements, and is suitable for long-term bedside detection. However, traditional DCS needs to quantify blood flow with an iterative fitting method, which is slow and cannot obtain...

AI Technical Summary

Problems solved by technology

[0009] Aiming at the problem that traditional DCS needs to quantify blood flow by iterative fitting method, the speed is low, and the change of tissue blood oxygen saturation cannot be obtained, the present invention proposes a tissue blood flow and blood oxygen saturation based on deep learning diffusion correlation spectrum Measurement methods

Images