Spring, 2013


Course Syllabus

CPH 738-003

[Introduction to Biomedical Image Computing and Imaging Informatics]

[Spring 2013]

[738: for graduate students only]



Classroom and Meeting hours: Wed. 3-5.30PM, MDS 221



Contact information


Course Directors:                                [Lin Yang, Assistant Professor/DBI, CPH]

Web:                                                    []

Office:                                                 [230D MDS]

            Telephone:                                          [859-218-2248]

            E-mail:                                                 []

            Office Hours:                                      [Every Wed: 1.30-3.00PM]



Course description


This class aims to give students a broad overview of biomedical image analysis and imaging informatics. We will introduce the state-of-the-art knowledge to understand, develop, and apply existing methods and software to handle biomedical image data to extract quantitative matrices. The topics in this class will include:

1.    Introduction to basic imaging modalities such as MRI, CT, ultrasound, X-ray, Microscopy.

2.    Introduction to basic linear algebra that is required for understanding biomedical image analysis.

3.    Basic image processing, such as filtering, mathematical morphology, etc.

4.    Basic machine learning topics for biomedical image analysis (unsupervised learning: such as clustering using K-means and mean-shift, linear dimension reduction such as PCA, nonlinear dimension reduction, supervised learning such as naïve Bayesian, decision tree, support vector machine, etc)

5.    Image informatics (computer aided diagnosis, content based image retrieval, and basic knowledge for building an imaging informatics system)

6.    Selected hot topics: Multicore Processor, Graphic Processing Unit, cloud, and Grid for High Performance Computing


Course rationale:


We are living in a revolutionary age, witnessing the next-generation of medical image and information emerged in astounding volume and rich formats. Nowadays images and videos are widely used in biological research and medical clinical applications. Manual image analysis is extremely time consuming, labor intensive, prone to errors, and no reproducible. The end goal for this class is for you to not only learn the scope and importance of the fundamental principles of utilizing computers to analyze images automatically, but also get hands-on training in developing accurate image analysis algorithms and creating suitable imaging informatics tools and software to solve the computational image analysis problems in your research and practice. This class would be suitable for a broad spectrum of students from biostatistics, statistics, mathematics, informatics, computer science, electrical and computer engineering, biomedical engineering, and students, residents, fellows, and researchers from biology, pathology/radiology, and medicine, who are interested in learning how to utilize informatics and computational tools to analyze their images (multiple dimensions, multiple modalities) in a quantitative, automatic, objective, fast, and high throughput manner. 


Course prerequisites


Be familiar with Matlab and Linear Algebra.


Course objectives

Student learning outcomes:

Upon completion of this course, the learner will:


·         be able to understand a clear and theoretically grounded definition of imaging informatics;

·         be able to understand the widespread use of digital image with an emphasis on the areas of interest to the translational and clinical research community

·         be able to understand and use open source tools for radiology, pathology, and biology image processing, indexing, and retrieving,

·         be able to understand biomedical image analysis, basic computer vision, pattern recognition and machine learning terminologies with the focus of biomedical knowledge transfer;




No textbook is required for this class. Various readings from research articles and book chapters will be assigned based on the content being covered. Most readings will be available online for free of charge for UK students and staff. Readings can be found in course calendar. Students must have access to a computer with Internet connection that meets the university standard.


Recommended Reading List:


·         David Forsyth and Jean Ponce, "Computer Vision: A Modern Approach", Second Edition, Prentice Hall, 2011.

·         Richard Szeliski, "Computer Vision: Algorithms and Applications", Springer, PDF downloadable version at )

·         Christopher M. Bishop, “Pattern Recognition and Machine Learning", Springer, 2006

·         Michael S. Lew, “Principles of Visual Information Retrieval", Springer, 2001

·         Nadine Barrie Smith, “Introduction to Medical Imaging: Physics, Engineering, and Clinical Applications", Cambridge, 2010



Course requirements and learner evaluation


For those who take the class for credit, the following course assignments and midterm are required. Course grades will be based upon evaluation of the following activities:


Assignments and Class Presentation (30%)

·         Exercise #1 (10%): Image Processing

·         Exercise #2 (10%): Image Segmentation

·         Exercise #3 (10%): Machine Learning

·         Exercise #4 (10%): Class Presentation


Attendance (5%): 1 point deduction per missing class until 0.


Final exam (55%) will be a competitive team project. Students will be separated into several groups to work on one specific problem. Each group will test their biomedical image analysis algorithm based on the other group’s proposed testing data and instructor’s testing images. The training image will be provided in advance. Winner group will get extra credits to compensate what they lost during their exams. The winner team will be authorized a certificate and treated an honored pizza sponsored by the instructor.


Final project score distribution (20% competition and demo, 10% presentation, 25% final report): project will contains three stages. The winner will get pizza and honorable certificate from the instructor. The winning group will be able to ask for the instructor’s recommendation letter for future internship and permanent jobs.


Bonus points: 20% will be offered during your final project if you can build a live demo to demonstrate all your developed algorithms for your final project.


Grading Scale:

·         100-80=A

·         79-70=B

·         69-60=C

·         0-59=F


Class Schedule:









Read Matlab tutorial: pdf


Lin Yang: How to give professional scientific presentations?


Vector Space and Linear Algebra



Frank Appiah: Computer vision in cell biology

Recommend reading: Low rank matrix approximation with weights or missing data is NP hard.  


Guest Lecture: Medical Imaging by Dr. Jie Zhang





Image Processing: Filter and Edge Detection

Read Matlab Image Processing Toolbox




Segmentation: Clustering



Sean Michael Hamlet: A computational approach to edge detection

Recommend reading: Optimization algorithm on matrix manifold


Segmentation: Graph Cut and Deformable Model

Assignment 1 and final project training data released


Manish Sapkota: Mean-shift: A Robust Approach toward feature space


Final Project Group:


Group 1: Sean Michael Hamlet, Chao Du, Daniel Irwin (L)

Group 2: Fujun Liu, Feiyu Shi, Yuchen Yang (M)

Group 3: Fuyong Xing, Hai Su, Hongyuan Wang (L)

Group 4: Manish Sapkota, Mingjun Zhao, Changzhe Jiao, Frank Appiah (M)


Annocement about final project:



-       In your shared folder, it contains two set of images. You can tell by checking your postfix (L or M) to know which dataset you will work on. Group 1 and 3 will work on the same dataset, Group 2 and 4 will work on another dataset. I have included both the original images and the ground-truth annotations. Please keep in mind that, your final testing data might not look quite similar to your training data. Therefore be careful do not over-fit your models if you choose learning based method.

-       I recommend you start to think about how to get the results closest to the ground truth annotations. Please let me know if you have any questions. At least you should have one/two group meetings at a weekly basis to discuss how your group decides to proceed.

-       Any questions, don’t hesitate to contact me, the earlier, the better!



Instructor expectations


1.    I expect you to attend every class session.  The components are highly interrelated; missing a class will detract from the learning potential of subsequent sessions.

2.    I expect you to be in the classroom and prepared to begin work at the scheduled starting time for each session.

3.    I expect you to actively participate in the discussions.  This is not the type of class where you can “sit back and listen.”

4.    I expect you to submit papers using proper English grammar, syntax, and spelling.  You are encouraged to use spell check and grammar check prior to submitting your written work.  The Writing Laboratory is available to anyone who may need assistance.  Grammar, syntax, and spelling will account for 10% of the grade for written work.

5.    I expect (and encourage) you to provide honest and timely feedback regarding the content and process of this course throughout the semester.

6.    I expect you during the semester to interactively engage via Blackboard with the other students and the instructor.

7.    I expect you to share in the responsibility for making this course an enjoyable and beneficial learning experience.

8.    This is a graduate-level course that requires you to study at least 3 hours a week for one credit course like this one. You may need less time, but be prepared for the fact that some weeks may be busier than others.

9.    I expect you to log into BB course homepage to access course announcement, course information, assignment submission, and communication with your fellow classmates.


Academic honesty

Academic honesty is highly valued at the University.  You must always submit work that represents your original words or ideas.  If any words or ideas used in a class assignment submission do not represent your original words or ideas, you must cite all relevant sources and make clear the extent to which such sources were used.  Words or ideas that require citation include, but are not limited to, all hard copy or electronic publications, whether copyrighted or not, and all verbal or visual communication when the content of such communication clearly originates from an identifiable sources.  Please see the University’s policies concerning the consequences for plagiarism. Source:  




If you have a documented disability that requires academic accommodations, please see me as soon as possible during scheduled office hours.  In order to receive accommodations in this course, submit to me a Letter of Accommodation from the Disability Resource Center (  If you have not already done so, please register with the Disability Resource Center for coordination of campus disability services available to students with disabilities.


Religious Observances

Students will be given the opportunity to make up work (typically, exams or assignments) when students notify their instructor that religious observances prevent the student from doing their work at its scheduled time. Students must notify the course instructor at least two weeks prior to such an absence and propose how to make up the missed academic work.


Inclement weather

The University of Kentucky has a detailed policy for decisions to close in inclement weather.  The snow policy is described in detail at or you can call (859) 257-5684.


Late work policy

Assignments that are turned in late will be marked one letter grade lower unless prior approval from the instructor has been obtained. It will be based on the time stamp provided by Blackboard. (NOTE: Assignments more than one week past the original due date will not be graded.)


Excused absences policy

Attendance, excused absences and make-up opportunities for this course will conform to the course policies established by the Office of Academic Ombud Services as found at