Quantitative Detection And Study Of Biological Tissue Using Molecular Hyperspectral Imaging

The hyperspectral imaging technology can provide information of image and spectra about biological tissue samples simultaneously, and give qualitative, quantitative and positioning description aim at the detection target, and achieve early diagnosis of certain pathological changes. Compared to other medical imaging technology, it has unique advantages. In this paper, a molecular hyperspectral imaging (MHSI) system was designed with hyperspectral imaging technology. Since the system can detect sample without damage and contact detection, it has been used for biological detection and quantitative assessment of living tissue. By depth comprehensive study the self-developed MHSI system about molecular hyperspectral imaging principle and biomedical imaging applications, the following innovative results were acquired:1. On the basis of the first generation of microscopic pushbroom hyperspectral imaging system, a novel MHSI system based on AOTF was presented. Compared with the first generation of imaging systems, the second generation has superior performance with fast electrical tuning, no moving parts, relatively high transmission rate, good image quality and real-time imaging. By studying the key components, we analyzed the system’s overall specifications and performance in order to provide the basis for preprocessing and quantification of molecular hyperspectral imaging data.2. Combination of the system performance and technical indicators of various components, we developed system software. The system software can effectively run with the system hardware and capture the molecular hyperspectral imaging data by different ways, including fixed-band continuous acquisition of non-integral approach, single-frame capture mode of integral approach, continuous variable-band approach collection of integral approach and continuous variable-band collection of non-integration approach, etc. These ways can be flexibly chosen. System software has friendly interface to flexibly configurate parameters.3. In the analysis of optical model and data model of MHSI system, the molecular hyperspectral imaging data of biological issue are preprocessed referring performance of MHSI system. A new gray correction coefficient algorithm with spatial dimension and spectral dimension was provided by using reference blank images of biological tissue. The algorithm can effectively eliminate the influence of system light. It was achieved in ENVI and IDL software platform. The data was normalized so that it was comparable with the biological tissue sample data.4. We are taking vigorous action to explore a new way of classify motor and sensory nerve by MPHI system. During nerve repair, only the distal and proximal nerve fibers are matched with corresponding to the mirror, in order to ensure the effect of nerve repair. We captured the nerve biopsy samples by MHSI system, and found four different types of the spectral characteristics of motor and sensory nerve, respectively from the outer membrane of nerve fibers, the nerve fibers, the core of nerve fibers, the outer ring of nerve fiber. We also analysed their spectral characteristics and classification and explored feasibility to classification of motor and sensory nerve by MHSI system.5. Base on the research of diabetic retinopathy(DR) by the microscopic pushbroom hyperspectral imaging system(MPHI), the relative work wa still go on to evaluate the protective effects of erythropoietin(EPO) and Adeno-Associated Virus Type 2 erythropoietin (AAV2-EPO) on early diabetic retinopathy in rats byMHSI system. Molecular hyperspectral imaging data are collected and preprocessed from six groups:normal(N),diabetic(D),EPO-treated(E0),AAV2-EPO(low-dose)-treated(E1), AAV2-EPO(middle-dose)-treated(E2) and AAV2-EPO(high-dose)-treated(E3). Not only by qualitative analysis of images and spectra curves, but also by quantitative analysis of the thickness of ONL, the comparision of relative errors and the comparision of the spectral similarity, the results suggest EPO and AAV2-EPO may prevent cell apoptosis and has protective effect on rat DR at a certain level.