三维生物发光断层成像系统搭建与算法研究
发布时间：2018-12-18 14:12
【摘要】：生物发光断层成像作为光学分子影像学的重要分支,能够从细胞和分子水平对生物体内肿瘤的生长和转移,特定基因的表达等诸多生物学过程进行分析和检测,可广泛应用于肿瘤检测、基因治疗及药物研发等领域。目前,生物发光断层成像系统成熟产品不多,美国精诺真IVIS系列产品在市场上占主要份额,但价格昂贵。因此,研究具有自主知识产权,高性能指标的生物发光断层成像系统势在必行。本文研究了生物发光断层成像的重建算法,并设计和搭建了成像仪硬件系统。主要工作如下:1、研究了生物发光断层成像中光在组织内的传输模型。针对辐射传输方程的复杂性,将光在生物组织内的传输简化为扩散近似方程,并用有限元方法求解方程的系数矩阵。2、将光在空气与透镜系统中的传输简化为小孔模型,利用CCD成像原理推导出图像灰度值与光辐射功率的线性关系,并根据朗伯余弦定律推导出像点与物点间的能量关系,避免传统积分球校准方法的高成本及复杂性。通过CCD采集发光图像,实现了成像物体体表光辐照度的重建。仿体和猪肉组织实验验证了本文提出方法的有效性。3、研究了生物发光断层成像体内光源重建算法。本文提出了自适应选取可行区的策略以降低重建算法的欠定性和病态性。该方法不但可以有效避免传统手动选取可行区带来的系统不稳定性,还可以提高体内光源的重建精度。可行区划定后,针对目前系数矩阵抽取方法会破坏矩阵正定性的问题,本文提出了模型降阶算法,可确保矩阵的正定性。最后,结合Tihonov正则化和共轭向量基算法快速准确实现了体内光源的重建。4、设计和搭建了生物发光断层成像仪的软硬件系统。通过设计的旋转平台采集八视角的发光图像以增加先验信息,可有效降低系统的欠定性。利用仿体和小鼠实验对成像系统综合性能指标进行验证,仿体实验光源重建的定位精度在1mm以内,小鼠实验光源重建的定位精度在2mm以内。上述研究表明,论文研究的内容可有效降低生物发光断层成像系统的欠定性及病态性问题,提高体内光源重建的定位精度。
[Abstract]:As an important branch of optical molecular imaging, bioluminescence tomography can analyze and detect many biological processes, such as tumor growth and metastasis, specific gene expression and so on. Can be widely used in tumor detection, gene therapy and drug development and other fields. At present, the bioluminescence tomography system mature products are not many, the United States Jinnuzhen IVIS series products in the market share, but the price is expensive. Therefore, it is imperative to study bioluminescence tomography system with independent intellectual property and high performance index. In this paper, the reconstruction algorithm of bioluminescence tomography is studied, and the hardware system of the imaging instrument is designed and built. The main work is as follows: 1. The light transmission model in bioluminescence tomography is studied. In view of the complexity of the radiation transfer equation, the propagation of light in biological tissues is simplified as a diffusion approximation equation, and the coefficient matrix of the equation is solved by using the finite element method. 2. The transmission of light in the air and lens system is simplified as a small hole model. The linear relationship between image gray value and optical radiation power is derived by using CCD imaging principle, and the energy relationship between image point and object point is deduced according to Lambert's cosine law, which avoids the high cost and complexity of traditional integrated sphere calibration method. The light irradiance of the body surface of the imaging object is reconstructed by collecting the luminous image by CCD. Experiments of simulated body and pork tissue verify the effectiveness of the proposed method. 3. The reconstruction algorithm of bioluminescence tomography in vivo is studied. In this paper, a strategy of adaptive selection of feasible region is proposed to reduce the ill-nature and ill-condition of the reconstruction algorithm. This method can not only effectively avoid the instability caused by the traditional manual selection of feasible areas, but also improve the reconstruction accuracy of the internal light source. In view of the problem that the method of coefficient matrix extraction can destroy the positive definiteness of matrix after the feasible area is delineated, a model reduction algorithm is proposed to ensure the positive definiteness of matrix. Finally, combining with Tihonov regularization and conjugate vector basis algorithm, the reconstruction of internal light source is realized quickly and accurately. 4. The hardware and software system of bioluminescence tomography is designed and built. In order to increase the priori information the luminous images of eight angles of view are collected by the designed rotating platform which can effectively reduce the undercharacterization of the system. The synthetic performance indexes of the imaging system were verified by imitating body and mouse experiments. The localization accuracy of the reconstruction was within 1mm and the localization accuracy of mouse experimental light source was within 2mm. The above results show that the content of this paper can effectively reduce the problem of ill-quality and ill-condition of bioluminescence tomography system and improve the positioning accuracy of the reconstruction of the light source in vivo.
【学位授予单位】：南京航空航天大学
【学位级别】：硕士
【学位授予年份】：2014
【分类号】：TP391.41
，
本文编号：2385964
[Abstract]:As an important branch of optical molecular imaging, bioluminescence tomography can analyze and detect many biological processes, such as tumor growth and metastasis, specific gene expression and so on. Can be widely used in tumor detection, gene therapy and drug development and other fields. At present, the bioluminescence tomography system mature products are not many, the United States Jinnuzhen IVIS series products in the market share, but the price is expensive. Therefore, it is imperative to study bioluminescence tomography system with independent intellectual property and high performance index. In this paper, the reconstruction algorithm of bioluminescence tomography is studied, and the hardware system of the imaging instrument is designed and built. The main work is as follows: 1. The light transmission model in bioluminescence tomography is studied. In view of the complexity of the radiation transfer equation, the propagation of light in biological tissues is simplified as a diffusion approximation equation, and the coefficient matrix of the equation is solved by using the finite element method. 2. The transmission of light in the air and lens system is simplified as a small hole model. The linear relationship between image gray value and optical radiation power is derived by using CCD imaging principle, and the energy relationship between image point and object point is deduced according to Lambert's cosine law, which avoids the high cost and complexity of traditional integrated sphere calibration method. The light irradiance of the body surface of the imaging object is reconstructed by collecting the luminous image by CCD. Experiments of simulated body and pork tissue verify the effectiveness of the proposed method. 3. The reconstruction algorithm of bioluminescence tomography in vivo is studied. In this paper, a strategy of adaptive selection of feasible region is proposed to reduce the ill-nature and ill-condition of the reconstruction algorithm. This method can not only effectively avoid the instability caused by the traditional manual selection of feasible areas, but also improve the reconstruction accuracy of the internal light source. In view of the problem that the method of coefficient matrix extraction can destroy the positive definiteness of matrix after the feasible area is delineated, a model reduction algorithm is proposed to ensure the positive definiteness of matrix. Finally, combining with Tihonov regularization and conjugate vector basis algorithm, the reconstruction of internal light source is realized quickly and accurately. 4. The hardware and software system of bioluminescence tomography is designed and built. In order to increase the priori information the luminous images of eight angles of view are collected by the designed rotating platform which can effectively reduce the undercharacterization of the system. The synthetic performance indexes of the imaging system were verified by imitating body and mouse experiments. The localization accuracy of the reconstruction was within 1mm and the localization accuracy of mouse experimental light source was within 2mm. The above results show that the content of this paper can effectively reduce the problem of ill-quality and ill-condition of bioluminescence tomography system and improve the positioning accuracy of the reconstruction of the light source in vivo.
【学位授予单位】：南京航空航天大学
【学位级别】：硕士
【学位授予年份】：2014
【分类号】：TP391.41
，
本文编号：2385964
本文链接：https://www.wllwen.com/falvlunwen/zhishichanquanfa/2385964.html
