基于多b值DWI的乳腺癌新型影像表型筛选及多参数联合诊断、分型模型的构建

Heterogeneity of breast cancer is emerging as one of the major reasons for failed therapy and tumor recurrence. Qualitative diagnosis and determination of molecular subtypes are the key points in diagnosis and treatment of breast cancer. Puncture biopsy is the gold standard of qualitative diagnosis and molecular subtyping, but it is inadequate at capturing the entire range of the intratumor heterogeneity. Thus, it is necessary to explore new imaging phenotype and establish effective system to assist diagnosis and subtyping of breast cancer. The diffusion-weighted imaging (DWI) with multi-b values is a kind of new technique of functional MRI (fMRI). Our preliminary experiments demonstrated that DWI with multi-b values (Mb-DWI) had high sensitivity and specificity in the differential diagnosis of benign and malignant breast lesions. Also, different molecular subtypes of breast cancer had distinct Mb-DWI parameters. These results indicated that Mb-DWI parameters might be new imaging phenotypes of breast cancer. According to radiogenomics strategy, this project is to further evaluate the diagnostic efficiency of Mb-DWI by bioinformatic techniques. By using molecular biological and histopathological techniques, we also intend to clarify the molecular pathological basis of Mb-DWI. Multi-parameter model based on Mb-DWI phenotypes and traditional MR features will be constructed and tested. The aim of current project is to produce powerful imaging supports and theoretical basis for the diagnosis, making therapeutic schedule, and prognostic evaluation of breast cancer.

肿瘤的异质性是乳腺癌治疗失败和复发的重要原因之一，因此乳腺癌的准确诊断和分型是诊治的关键。穿刺活检是定性诊断及分子分型的金标准，但其对肿瘤异质性的反映作用有限，有必要探索乳腺癌新型影像表型并构建高效的影像学诊断分型体系，作为重要的补充。多b值弥散加权成像（DWI）是一项新的fMRI技术。我们的预实验发现，多b值DWI鉴别良、恶性乳腺病变有较好的敏感性与特异性，不同分子亚型乳腺癌的多b值DWI参数值有所不同，提示多b值DWI参数可能成为乳腺癌新型影像表型。本项目拟采用放射基因组学的思路，应用生物信息学技术，深入评价多b值DWI参数的诊断效率，确立乳腺癌新型影像表型，并结合分子病理学技术，阐明其可能的分子病理学基础，最终将新的影像表型与传统MR特征结合，构建多参数联合的诊断分型模型，为乳腺癌的诊治和预后评估提供有力的影像学支持和理论基础。

肿瘤的异质性是乳腺癌治疗失败和复发的重要原因之一，因此乳腺癌的准确诊断和分型是诊治的关键。穿刺活检是定性诊断及分子分型的金标准，但其对肿瘤异质性的反映作用有限，有必要探索乳腺癌新型影像表型并构建高效的影像学诊断分型体系，作为重要的补充。多b值弥散加权成像（DWI）是一项新的fMRI技术。本研究优化了多b值DWI序列，评估了其参数的稳定性和可重复性，保证了该技术临床应用的可操作性；其次，发现多b值DWI的单参数及多参数联合具有良好的诊断效率，并初步构建并测试了多参数联合的模型；初步探讨了多b值DWI可能的组织病理学基础。该研究为多b值DWI技术的临床应用奠定了基础，并提供了循证医学证据，为乳腺癌的诊治和预后评估提供了有力的影像学支持和理论基础。