课程大纲

课程介绍

《基于生物医学统计的信号处理》介绍数字信号处理和统计理论和技术在生物医学信号和系统中的应用。课程主要介绍运用统计学、信号处理理论以及机器学习和优化理论等知识对生物医学信号进行表示和分析。通过实例分析，介绍经典的生物医学信号的统计处理方法，以及最新的基于深度学习、多模态学习和图神经网络等方法，让学生了解各种方法优缺点，掌握分析和解决实际问题能力。《基于生物医学统计的信号处理》是一门学科交叉课程，介绍信号和图像处理基本方法（傅立叶变化、图像滤波和增强、图像恢复与重构、小波与多分辨率处理、边缘检测和图像分割、聚类和识别等）和新兴技术（深度学习、多模态学习、图神经网络等），在生物医学信号处理中的应用（例如生物电信号、X射线影像、断层扫描成像、磁共振成像以及病理图像等），及统计分析理论和方法（假设检验和方差分析、广义线性回归、生存分析、组学数据归一、基因检测注释、疾病分型分类等）。作为学科交叉课程，通过介绍信号处理、机器学习、统计学和生物医学信息学等学科的交叉应用，使学生获得更广阔的学术和应用视野，能够综合多学科知识解决实际问题。

Course Description

“Signal Processing Based on Biomedical Statistics” introduces the applications of digital signal processing and statistical theories and techniques into biomedical signals and systems. This course will present the analysis and representation of biomedical signals using the intersection of biostatistics, signal processing, machine learning and optimization theory. It aims at elaborating the benefits and limitations of various approaches (e.g., state-of-the-art statistical processing methods and latest techniques based on deep learning, multimodality machine learning and graph neural networks) and specifying the way to identify proper solutions to biomedical applications through case study. “Signal Processing Based on Biomedical Statistics” is an interdisciplinary course that incorporates classical methods and latest techniques for signal and image processing and theories and techniques for statistical analysis in a variety of biomedical signals, including bioelectrical signals (ECG, EEG and EMG), medical images (X-ray images and tomography, magnetic resonance imaging (MRI) and pathological/histology images), biomedical omics data and clinical data. We will introduce the classical signal and image processing methods (i.e., Fourier transform, image filtering and enhancement, image restoration and reconstruction, wavelet and multiresolution processing, edge detection and image segmentation, and clustering and classification) and latest techniques (i.e., deep learning, multimodality machine learning and graph neural networks), as well as theories and methods for statistical analysis (i.e., hypothetical testing and analysis of variance, generalized linear regression, survival analysis, omics data normalization, genetic testing and annotation, and disease classification).

学习目标（Learning outcomes）

1. 了解大数据医疗和精准医疗背景下，信号处理、机器学习、统计学、生物医学信息等学科交叉的趋势，获得跨学科视野，提升专业热情。

Understand the tendency of intersection of signal processing, machine learning, statistics and biomedical informatics under the background of medical big data and precision medicine, develop inter-disciplinary viewpoints and improve professional insights.

2. 掌握生物医学信号的统计和处理中的基本概念，认识从信号处理、表示识别到统计分析的全过程。

Understand the basic concepts in processing and statistical analysis of biomedical signals and the whole process of process, representation, classification and statistical analysis.

3. 掌握生物医学信号的统计和处理理论知识，描述处理方法的数学模型和求解过程，理解统计结果的实际意义。

Demonstrate the knowledge of statistical and processing theories for biomedical signals, describe the mathematical formulation and solving process for processing, and interpret the statistical results.

4. 独立实现对于典型生物医学信号（生物电信号、医学影像、组学数据、临床数据）处理和统计的基本方法。

Implement the basic algorithms for statistical analysis and processing of typical biomedical signals (e.g., bioelectrical signals, medical images, biomedical omics data and clinical data).

5. 掌握生物医学影像（X射线影像、断层扫描成像、磁共振成像、超声影像等）处理的各种经典和最新方法，清楚各种方法优缺点，能够针对实际问题选择合适方法。

Understand the practical benefits and limitations of various biomedical image processing approaches (state-of-the-arts and most recent methods), and identify the best solution for specific practical problems.

6. 针对涉及生物医学信号的相关任务/系统，培养处理、统计和分析的综合能力，通过协作完成课题设计。

Co-operate to analyze, propose, carry out and present the project designs for specific task/system related to biomedical signals with a comprehensive capacity in processing and statistical analysis.

参考书

参考文献

教学日历

教学日历

具体详见：

基于生物医学统计的信号处理-课程教学大纲.doc