BMI Course Information
BMI 516 Advanced Biomedical Data Analysis (3)
Instructor Information |
|
Fall 2009 |
Course not offered |
Spring 2009 |
Course not offered |
Catalog Description
Acquisition, conversion and organization of biological data into relevant diagnostic, therapeutic, and research information using information extraction and data mining.
Prerequisites
BMI 501 with a C or better and admission to any SCI graduate program
Textbook and Other Materials
An Introduction to Bioinformatics Algorithms (Computational Molecular Biology) by Neil C Jones (Author), Pavel A. Pevzner (Author), The MIT Press (August 1, 2004).
Course Learning Outcomes
Understand how to acquire, convert, and organize biological data into relevant diagnostic, therapeutic, or research information.
- Biological data acquisition and digitalization
- Imaging data acquisition
- Organization and reporting of data results
- Information extraction and data mining tools and techniques
Major Topics and Time Covered
Clinical, imaging and bioinformatics data will be utilized in the course. Advanced techniques on information extraction and data mining will be utilized for data analysis. Methods of data organization and data reporting will also be covered.
- Biological data types and acquisition: various types of high-throughput biological data will be discussed. Among them are genome sequences, expression microarrays and array-based CGH (comparative genomic hybridization), proteomic data such as mass spectroscopy, protein abundance array, etc, and other types of high throughput screening data such as siRNA perturbation data. In addition, medical data such as physiological data, histology, drug responsiveness, and other laboratory data that are relevant to biomedical informatics research will also be discussed. A brief introduction to biological sample acquisition and preparation for high throughput microarrays-based technologies will also be covered, including efficient signal extractions from the raw data.
- Medical imaging types and acquisition: medical image data such as X-ray, CT scans, MRI (Magnetic Resonance Image), Ultrasound image, and other nuclear data will be discussed. Some of the issues related to medical image acquisitions such as subject limitations – movement, radiation exposure, timing, basic body physiology issues, and effective resolution will also be discussed. This will be used to build an understanding of subsequent information extraction algorithms and data analysis.
- Data Organization: understanding and developing efficient methods to organize various types of biomedical data are critical to more streamlined data analysis, especially cutting edge high-throughput data. In more detail, the following subjects will be discussed:
- Data processing and transformation into clinical knowledge
- Algorithms for information extraction, data digitalizing and standardization, and the tradeoffs of digitalization
- Data reporting issues such as report formatting and the human computer interface with a focus on biomedical scientists and clinicians. Discussion of the challenges of institutional transactions and interoperability among highly heterogeneous computing systems.
- Data storage and archival of high throughput biological data and medical data
- Machine Learning and Data Processing: basics of various machine learning and data mining algorithms popular in biomedical informatics research will be introduced. A rule-based approach based on medical heuristics and/or biomedical knowledge will also be discussed to guide the development of better decision support systems. All students will be required to apply the methods to real-world biomedical data to solve biological questions in a semester project.