连续波近红外无创脑功能检测中的干扰抑制方法研究及脑力疲劳评估分析

The in-depth study of brain activity measurement will be helpful to understand the relationship between the function of the brain structure and neural information processing, the operating mechanism of brain activities, generation mechanism and representation forms of brain disease and injury, which has great practical significance and scientific value. As an emerging measurement technique in brain activity measurement, continuous wave functional near infra-red spectroscopy has a lot of advantages such as simple structure, non-invasive, easily implementation, operational safety, suitable for real environment, high time resolution and appropriate spatial resolution. However, the interference factors exist in practical applications, such as gross errors, cross talk, physiological interference, layered inhomogeneity and measurement conditions, seriously affect the accuracy of brain activity measurement and restrict its applications in other fields. To solve the above problems, research areas of this project include theoretical analysis, signal processing method implementation, innovative design of near infra-red spectroscopy measurement system, complex issues exploration and frontier application research. The research results of the project will improve the accuracy and reliability of brain activity measurement based on continuous wave near infra-red spectroscopy and promote its application in the field of mental fatigue assessment.

对脑功能活动检测的深入研究，有助于了解脑部结构功能与神经信息处理之间的关系，理解脑功能响应活动的运行机制，掌握脑疾病、脑损伤产生机理与表征形式，具有重大现实意义与科学价值。连续波功能近红外光谱技术作为新兴的脑功能活动检测手段，具有结构简单、易实现、非侵入、操作安全、适于真实环境测试、时间分辨率高、空间分辨率合理等诸多优点。但在应用中，由于粗差干扰、串扰问题、生理噪声干扰、脑组织非均匀性、测量条件影响等干扰因素的存在，严重影响了脑功能活动检测的精度，也制约了相关应用领域的拓展与深入研究。针对以上问题，本项目拟通过基础理论分析、数据处理方法研究、创新性测试系统设计、复杂真实问题探索和前沿应用研究来探索各自的应对思路与解决方法。项目的研究成果将有助于提高连续波近红外无创脑功能检测技术在应用中的准确性和可靠性，也将有助于推动该技术在脑力疲劳评估分析领域的发展。

本项目基于连续波近红外光学检测技术，开展多波长多距离近红外脑功能检测技术研究，针对测量中存在的粗差干扰、生理噪声干扰、测量条件复杂性影响等问题提出了相应的信号处理方法，并开发研制了基于多波长多距离的连续波近红外检测系统样机，该系统设计能提供更多的冗余波长组合、增强对深层脑组织的生理干扰抑制、适用于复杂测试条件。本项目已完成以下研究内容：（1）针对传统自适应滤波方法难以抑制深层脑组织中的生理噪声干扰，提出基于最小二乘支持向量机的近红外脑功能信号提取方法；（2）针对实际应用中的差分路径因子参考值偏差干扰，提出基于粒子群算法的差分路径因子估算方法；（3）针对光强测量信号中存在噪声干扰及差分路径因子偏差共同存在影响，提出一种基于总体最小二乘法的连续波近红外脑功能活动信号处理方法；（4）针对光强测量中存在的粗差问题，提出基于改进鲁棒最小二乘支持向量机的近红外脑功能信号粗差处理方法；（5）针对测量中存在的运动伪迹问题，提出了基于集合经验模态分解及最小二乘线性拟合的近红外脑功能信号运动伪迹剔除方法；（6）基于近红外测量系统样机，开展脑功能活动测量研究并测量分析不同工作状态下的脑组织血液动力学变化情况。.基于项目研究成果，共发表论文7篇，其中SCI 收录4篇，EI收录2篇；申请国家发明专利7项。