低强度脉冲超声激励间充质干细胞移植治疗脊髓损伤技术研究

发布时间：2018-09-15 19:07

【摘要】：随着现代交通、建筑、工业的迅猛发展,高能量损伤患者数量在外伤患者中所占比例逐年增加,脊髓损伤(Spinal Cord Injury,SCI)的发生率呈逐年上升趋势。脊髓损伤是骨科领域常见的疾病之一,且具有高发性,高致残率和多损伤严重等特点,所带来的后果可能是灾难性的,轻者丧失劳动能力,重者不能活动、小便失禁、丧失生活自理能力。据统计分析发达国家的SCI发病率为28.3～45人/百万人/年,我国发病率6.7人/百万人/年。损伤发生后患者的神经功能恢复率极低,因此对于脊髓损伤的后续治疗,如何提高治愈率是现在医学界面临的难题之一,也是重点研究领域之一。目前针对脊髓损伤已开展大量的研究,但其修复主要面临的问题是:神经元的凋亡,轴突再生能力差以及损伤所造成的抑制性微环境。而随着研究的深入,尤其是细胞移植研究结果表明,细胞移植可以促进轴突的再生和髓鞘化,进而替代凋亡的神经细胞,从而促进脊髓损伤的修复,为感觉及运动功能的恢复创造条件。对于细胞移植治疗脊髓损伤而言,如何提高细胞的活力和分化能力是移植成功的关键。而现有的研究表明,低强度脉冲超声波(low intensity pulsed ultrasound stimulation, LIPUS)干预种子细胞能够明显的提高细胞的增殖和分化能力。然而怎样选取合适的干预参数使得治疗效果最佳是难题之一。对于多目标寻优,基于遗传算法改进的BP神经网络能够通过全局搜索,得出最优解。本实验首先搭建一套超声激励系统,包括信号发生器、放大器、电源和检测模块。其中放大器模块采用THS4062芯片搭建而成的。然后通过医学实验得出实验结果。将BP神经网络与遗传算法相结合,利用MATLAB软件对实验结果进行预测仿真分析,得出最佳实验条件,即对BMSCs的增殖、活性影响最佳激励参数的组合。且经过验证实验,证明预测结果的正确性,为后期的移植实验做准备。本文基于低强度脉冲超声波对骨髓间充质干细胞的生物学效应,利用不同参数组合的LIPUS刺激细胞,并通过采取基于遗传算法改进的BP网络模型进行预测。通过对神经网络计算出的实验条件进行验证实验,得出此时的细胞活性值与比对照组提高了 24.9%,效果显著,且预测的最佳条件下得到的细胞活性值优于其他预设实验条件下所得到的。
[Abstract]:With the rapid development of modern traffic, construction and industry, the proportion of high-energy injury patients in trauma patients increased year by year, and the incidence of spinal cord injury (Spinal Cord Injury,SCI) increased year by year. Spinal cord injury is one of the most common diseases in orthopedic field, and it has the characteristics of high incidence, high disability rate and severe injury. The result may be disastrous, the light person is incapacitated to work, the heavy person is unable to move, the urine is incontinence, and so on. Loss of ability to take care of oneself. According to statistics and analysis, the incidence of SCI in developed countries is 28.30.45 people / million people / year, and the incidence rate in China is 6.7 people / million people / year. The recovery rate of neurological function in patients with spinal cord injury is very low. Therefore, how to improve the cure rate of spinal cord injury is one of the difficult problems in the medical field and one of the key research fields. At present, a lot of researches have been carried out on spinal cord injury, but the main problems in repairing spinal cord injury are neuron apoptosis, poor axon regeneration ability and the inhibitory microenvironment caused by injury. With the development of research, especially the results of cell transplantation, cell transplantation can promote axon regeneration and myelination, and then replace apoptotic nerve cells, thus promoting the repair of spinal cord injury. To create conditions for the recovery of sensory and motor functions. For the treatment of spinal cord injury by cell transplantation, how to improve cell viability and differentiation ability is the key to successful transplantation. However, current studies have shown that low intensity pulsed ultrasound (low intensity pulsed ultrasound stimulation, LIPUS) can significantly improve the proliferation and differentiation of seed cells. However, how to select the appropriate intervention parameters to make the best treatment is one of the problems. For multi-objective optimization, the improved BP neural network based on genetic algorithm can obtain the optimal solution by global search. In this experiment, a set of ultrasonic excitation system including signal generator, amplifier, power supply and detection module is built. The amplifier module is made of THS4062 chip. Then the experimental results are obtained through medical experiments. By combining BP neural network with genetic algorithm, the experimental results are predicted and simulated by MATLAB software, and the optimal experimental conditions are obtained, that is, the combination of the best excitation parameters affecting the proliferation of BMSCs and the activity of BMSCs. The experimental results proved the correctness of the prediction, and prepared for the later transplantation experiment. Based on the biological effects of low-intensity pulsed ultrasound on bone marrow mesenchymal stem cells (BMSCs), different parameter combinations of LIPUS were used to stimulate the cells, and the improved BP neural network model based on genetic algorithm was adopted to predict the effects. According to the experimental conditions calculated by neural network, the results show that the cell activity at this time is 24. 9% higher than that of the control group, and the effect is remarkable. The predicted values of cell activity under the optimal conditions are better than those obtained under other preset experimental conditions.
【学位授予单位】：天津科技大学
【学位级别】：硕士
【学位授予年份】：2017
【分类号】：R651.2

【参考文献】