Image Informatics

Bioimage informatics is an interdisciplinary field of research encompassing biology, information science, computer science, statistics, and engineering. Bioimage informatics strives to automate, simplify, and otherwise improve and reinvent techniques for the description, management, analysis, and preservation of biological image data.

Data Description

  1. What is better data description?
    1. Describing images and imaging experiments more accurately, more precisely, and more completely;
    2. Description keeping pace with the rate of data production by getting faster and less expensive.
  2. Why?
    1. To enable smarter searching and management of images
    2. To enable replication, reuse, and meta-analysis by enabling valid comparisons between separately collected images
    3. To enable automated analysis via semantic descriptions
  3. How?
    1. Metadata standards and Reporting Standards
      • Standardize the information that must recorded about an image and how it was created, as well as the format in which that information must be recorded.
    2. Ontology development
      • Standardize and disambiguate research vocabulary used for description
      • Harness the collective power of distributed annotation (like tagging)
    3. NLP
      • Description by proxy for images, using NLP of accompanying article text
    4. Visual recognition
      • Visual analysis of images to automate identification of image elements, textures, movement, etc.

Data Management

  1. What is better data management?
    1. Organizing images more efficiently and consistently.
    2. Enabling and controlling access to images more effectively and efficiently.
    3. Creating faster, clearer, and more innovative options for visualizing the data contained in an image.
  2. Why?
    1. To make it easier to search, sort, and find image data for a variety of purposes.
    2. To make image data from multiple labs and multiple timeframes interoperable.
    3. To increase access to more data, not just subsets of results featured in publications.
    4. To facilitate meta-analysis and reanalysis of data through interoperability.
    5. To facilitate preservation through good data planning.
  3. How?
    1. Software for visualization
    2. DBMS for organization and storage
    3. File formats and software for interoperability
    4. Federating data for findability and access.

Data Analysis

  1. What is better data analysis?
    1. Greater accuracy, precision, speed, and efficiency in extracting information from image data.
    2. Enabling new kinds of measurements and modeling that would not be possible without computers and other new technologies for detection.
    3. Extracting more information from less data.
  2. Why?
    1. Coming soon
  3. How?
    1. Data Mining and Reuse
    2. Automated visual analysis examples
      • Segmentation
      • Fluorescence decay

Data Preservation

  1. What is better data preservation?
    1. More of the most useful data of being available for longer periods of time.
  2. Why?
    1. To enable historical comparisons and other retrospective uses of data
    2. To keep data available for unforeseen future needs
    3. To maximize return on research investments
  3. How?
    1. Developing preservation criteria and priorities
    2. Developing failsafe storage strategies