Mesoscopic fluorescence molecular imaging three-dimensional reconstruction method based on 3D-Unet and system thereof

Abstract

The invention discloses a mesoscopic fluorescence molecular imaging three-dimensional reconstruction method based on 3D-Unet and a system thereof, and the method comprises the steps: obtaining the surface photon flux of a biological tissue, and obtaining a surface fluorescence photon intensity measurement data set of the biological tissue; obtaining a photon intensity simulation data set based ona phantom with an embedded fluorescent biological label generated by computer simulation; mixing the surface fluorescence photon intensity measurement data set of the biological tissue with the simulation data set, and selecting one part as a training data set; training a 3D-Unet network model based on the training data set to obtain model parameters; and carrying out three-dimensional reconstruction on the mesoscale biological label in the biological tissue. Compared with a traditional three-dimensional reconstruction method, the method is higher in reconstruction efficiency and robustness, the obtained reconstruction result has higher precision and resolution, and the method is more suitable for three-dimensional reconstruction of mesoscale biological tags.

Description

technical field [0001] The invention belongs to the technical field of mesoscopic fluorescence molecular tomography, and in particular relates to a method and system for three-dimensional reconstruction of mesoscopic fluorescence molecular imaging based on 3D-Unet. Background technique [0002] The statements in this section merely provide background information related to the present disclosure and do not necessarily constitute prior art. [0003] Optical molecular imaging is a new imaging modality that has developed rapidly in the past decade. With the continuous development of bioluminescent protein technology, fluorescent molecular imaging technology can more accurately intervene in biological and physiological processes for tracking and observation. Among the many imaging modalities of molecular imaging, each has different performance due to different working principles and working methods, and is suitable for different application fields. Optical molecular imaging te...

AI Technical Summary

Problems solved by technology

[0004] However, the mesoscopic fluorescence molecular tomography system can only detect the photon signals after multiple scattering, reflection and absorption inside the biological tissue on the surface of the biological tissue, and cannot directly obtain the internal spatial information of the biological tissue, so it must be reconstructed through the corresponding three-dimensional Algorithms can restore the three-dimensional distribution of fluorescent labels or tumor targets inside biological tissues

However, since the number and intensity of fluorescent photons decay exponentially with depth, it becomes more difficult to obtain 3D reconstruction of fluorophore distribution at deeper positions, and the 3D reconstruction process is an ill-posed and ill-conditioned problem.

At the same time, due to the high spatial resolution of mesoscopic scale reconstruction, a large amount of data is required, which makes the traditional iterative reconstruction algorithm not only time-consuming, but also the reconstruction accuracy and resolution will be limited.

Images