突发感音神经性耳聋脑网络动态演化与预后评估研究

Sudden sensorineural hearing loss (SSHL) is one of the acute onset, rapid progress, severe hearing loss of ENT emergency with a high incidence. The cause of SSHL is still unknown and it is very difficult to find significant lesions in the routine clinical and imaging examination, which resulting in poor prognosis of some patients. The pathogenesis of SSHL is still unclear and some studies reported a relationship with the brain network alterations caused by hearing impairment. Our previous study found an alteration of the functional connectome in unilateral SSHL patients which is characterized by a shift toward small-worldization. However, the mechanisms of these alterations and the dynamic evolution of the brain network are still unknown. Based on the previous research, this project explores the dynamic evolution of the brain network (structural network and functional network) of patients with SSHL by multimodal magnetic resonance imaging. Furthermore, the brain network progression of patients with SSHL was predicted based on the dynamic evolution model of brain network as well as the prognosis of the treatment. It is hoped that this project will elucidate the mechanisms of the progress of SSHL from the perspective of the overall dynamic evolution of the brain network, which will ultimately assist in providing a precision diagnosis and effective treatment of SSHL.

突发感音神经性耳聋是一种发病急、进展迅速、听力损伤严重的耳鼻喉科急症之一。发病率高，病因不明，在常规临床和影像检查中很难发现明显的病变，导致部分病人预后不佳。目前对其发病机制还不清楚，有研究表明与突发感音神经性耳聋病人听觉损伤后的脑网络改变有关。我们前期基于脑连接组的研究发现，单侧突发感音神经性耳聋病人脑功能网络存在“小世界化”的改变。但突发感音神经性耳聋病人脑网络为什么出现这些改变，以及其动态演化机制仍不清楚。本项目在先前研究基础上，通过多模态磁共振成像方法，探究突发感应神经性耳聋病人脑网络（结构网络与功能网络）随疾病进展的动态演化过程。并依据脑网络动态演化模型结果对突发感音神经性耳聋病人大脑网络进展进行预测，从而对其治疗效果进行一定程度的预后评估。从大脑网络的整体动态演化角度，阐明突发感音神经性耳聋的疾病进展机制，为突发感音神经性耳聋疾病的准确诊断和有效治疗提供帮助。

突发感音神经性耳聋是一种发病急、进展迅速、听力损伤严重的耳鼻喉科急症之一。发病率高，病因不明，在常规临床和影像检查中很难发现明显的病变，导致部分病人预后不佳。目前对其发病机制还不清楚，有研究表明与突发感音神经性耳聋病人听觉损伤后的脑网络改变有关。本项目收集了165例突发感音神经性耳聋病人，185例听觉正常的对照组被试的多模态功能磁共振数据，通过图论及仿真模拟的方法，探究突发感应神经性耳聋病人脑网络（结构网络与功能网络）的改变。并基于此提出突发感应神经性耳聋动态演化规则，对突发感音神经性耳聋病人治疗效果进行了预后预测评估。结果表明突发感音神经性耳聋病人各脑网络无论是全局还是局部拓扑属性参数在急性期内都已经存在显著的改变。与对照组相比，病人组白质结构网络向“随机化”方向偏移，而其功能连接网络则向“小世界”化方向偏移。基于此提出提高聚类系数（同时减小最短特征路径长度）的脑网络概率方法仿真演化模型，有效刻画了健康对照向突发感音神经性耳聋演化过程。同时，利用脑网络参数构建的分类模型，有效区分了突发感音神经性耳聋病人与健康对照，验证了基于功能磁共振数据进行疾病分类的可行性，为其他疾病基于多模态磁共振的分类研究，提供了方法学框架。在构建预测模型评估突发感音神经性耳聋患者的临床预后效果的中，发现在不同模态的MR数据参数中，有17个特征指标组成的模型表现出良好的性能，这为今后突发感音神经性耳聋患者听觉水平恢复的预测，提供了脑网络生物学标记。