Biology 727: Image processing for bioscientists

Images are essential in many areas of research.  Images of many forms need to be captured, processed, displayed and analyzed in scientifically meaningful ways.  This class aims to give a broad introduction into the principles behind image-based data and build skills to allow you to confidently deal with the images you use as a scientist.  Quantification of images is a major focus of the course.

The class covers images generally but examples will be based on the types of image data commonly used in biological research such as fluorescence, luminescence, light or electron micrographs, gel images, satellite images or scientific photographs.  The course is aimed at graduate students actively working in biological research labs but assumes no prior experience with programming or any computational background beyond that of a typical PC/Mac user.  

Aims: At the end of the course you will have the understanding and technical skills to process and analyze any image based data you need. With improved understanding you will feel confident in preserving the integrity and accuracy of images through any suitable processing steps. You will have skills and ideas to perform any quantitative analysis needed for your PhD work.

Formats: Lecture material will cover fundamental principles and background information to help you understand the processes. Most of the time will be spent on exercises and projects in and beyond class-time to allow acquisition of practically relevant skills in processing, displaying and analyzing images of many kinds that will help with your research. The class aims to provide general understanding of image processing rather than specific software instruction (so you can use whatever you need) but FIJI/ImageJ, some Matlab, and other software will be used for exercises.

Requirements: Need to bring a laptop computer (PC or Mac) to the classes. Software and example images will be provided. No previous experience with any software is required.

 

Topics and syllabus outline

Introduction Course aims, formats and approaches
Capture: sources of images A brief overview of what produces digital images, how they are made and how we perceive them. Context for accurate processing and analysis.
Image fundamentals The bits and pixels from which images are made. Colour spaces. Introduction to the spatial and frequency domains.
Processing and enhancement What you can do to images to make them easier to see or analyze. Contrast adjustments, histogram operations, look-up tables, filters for denoising, sharpening, finding edges
Quantification: fundamentals and  basic operations A suite of core functions for getting useful numbers from images - Intensity and spatial measures, manual and threshold-based measurements, ROIs, counting, binary images, improving accuracy with pre-processing. Using and understanding the numbers.
Restoration and recovery Correcting flaws - background, illumination, geometric corrections. Deconvolution - how it works and how to use it. Try Huygens with your images.
Transformation, registration and assembly Stitching and alignment of data.
Multi-dimensional data processing and display Coping with nDimensions - hyperstacks, projections and visualization tools.
Presentation of images and video Formats and operations for showing your data to others - figures, talks, web. Image compression options and when to use them.
Advanced quantification Building on the core quantification tools. . . Image calculations - ratio, differences, FRET and FRAP calculations. Colocalization, Tracking and motion estimation. Advanced segmentation.
Storage, retrieval and computation Safe keeping and retrieval. Tips and understanding to help if you have lots of data. Future possibilities.
Review and Image integrity Summary of core concepts and what you should be able to do. Making sure your new understanding helps you avoid accidentally doing anything inappropriate to your data. What not to do!
Computational customization Improving and building tools for your needs - writing and adapting macros/scripts in FIJI, Matlab operations and suitabilities. Automation of image analysis.
User projects Using all of the above for some applications that are useful for you. Either real or practice data. Get help and feedback in class. Aim to finish the class with skills and/or tools to help you in your PhD.

 

Enrollment:

 

  • If you are interested in taking this course please request enrollment with this form with some details of what you are doing and why you want to do the course. Since the numbers are limited, places are given to those who would seem to benefit most from the course.
  • Official enrollment for graduate student is through ACES with a permission number which I will email to you if it sounds like the course is a good fit for your needs.
  • A limited number of places are available for postdocs, faculty and staff
  • 3 credits
  • Credit/no-credit basis
Instructor:
Sam Johnson
sam.johnson@duke.edu
613-8216
Fall 2015
Thursdays 1:25 to 3:55
August 27 to November 19
Bio Sci 144

 

 

Back to LMCF guides