面向早期乳腺肿瘤诊断的MRI导引扩散光层析成像先进方法

扩散光层析技术（DOT）作为一种新型功能成像方式，在乳腺肿瘤检测、良恶肿瘤判别，病情分期等方面具有重要的临床应用前景。然而由于近红外光在组织体传输过程中固有的散射作用以及测量数据个数的有限性，导致成像具有较低的空间分辨率和定量性。本研究以实现早期乳腺肿瘤在体定量检测为目的，搭建扫描式锁相-光子计数型全并行稳态测量系统，以满足乳腺组织对测量系统高灵敏度、高空间分辨率、大动态范围和高信噪比等特性的要求；基于MRI解剖图像的高分辨特性，发展无需图像分割的非均匀组织光子模型及其基础上的MRI导引DOT直接正则化策略，并进一步引入基于压缩传感技术的稀疏正则化方法，在提高图像重建空间分率、定量化和鲁棒性的同时，克服目前多模态DOT成像方法中解剖图像分割中可能存在的人为误差、降低对影像医师的依赖性等问题，为DOT进一步临床应用提供新方法。另外，该方法也对具有相似成像理论的其它多模态成像具有重要的借鉴意义

Diffuse optical tomography is a new functional imaging and has important clinical potential in many aspects such as benign and malignant breast tumor detection, tumor staging criteria and so on. However, the natural scattering of the photon propagation in the tissue and the limited number of detection lead to the low spatial resolution and quantification of DOT. For in vivo quantitative detection of early breast tumor, a fully parallelized CW prototype system based on scan style, lock in technique and photon counting technique is established with the high sensitivity, high spatial resolution, large dynamic range and signal-to-noise ratio; Combined with the high spatial resolution of MRI, a direct regularization from co-registered anatomical images for MRI-guided DOT image reconstruction is proposed. In this novel approach, structural image segmentation is not required, which can avoid of the cases that the process of segmenting anatomical images into tissue types is prone to errors and can be extremely time consuming, requiring anywhere from a few minutes to over an hour, depending on the users’ expertise. Moreover the sparse regularization of compressed sensing is introduced to further improve the spatial resolution, the quantification and the robustness. In addition, this novel approach is general and could be adapted to other high resolution.anatomical images for guiding functional imaging modalities which require inversion similar to DOT.