Proposal aims

Purpose: Refine ImageJ’s core design to accommodate a broader range of needs in the scientific community.

Imaging is one of the most powerful tools available to the modern biologist and recent advances in quantitative microscopy and image analysis have greatly accelerated our understanding of many complex and dynamic processes. The NIH’s open-source ImageJ software has been enormously useful in many of these projects since its release in 1997. Any successful software project, after a period of sustained growth and the addition of functionality outside the original intended scope, will benefit from a subsequent period of scrutiny and refactoring, and ImageJ is no exception. Such review helps the program to remain accessible to newcomers, powerful enough for experts, and relevant to an evolving community. Furthermore, the pressing unmet needs of the existing ImageJ community, as well as of researchers who are hindered from building on ImageJ due to its limitations, lead us to propose the following work to improve ImageJ’s functionality and interoperability:

Aim I – Improve the ImageJ core architecture

Improvements in core architecture are required for the development and stability of the ImageJ project, its interoperability with other software, and its ability to support new features and applications:

1. Separate the data model from the user interface – The current ImageJ is tightly coupled to its GUI, hindering its use in many applications. We will decouple the ImageJ data model from its display. This will enable alternate views of data, features like dynamic charts and real-time linked displays, and most importantly, the ability to use ImageJ in headless mode on a computing cluster or as a library of independent functions. See Grant Harris's IjX whitepaper for a technical overview of our proposed interface-driven design, and the source code for a proof of concept.
2. Introduce an extensions framework for algorithms – To enable more widespread interoperability between independently developed algorithms and to enable the recording and writing of modular macros in multiple scripting languages, we will introduce a framework with clear declaration of inputs and outputs—including data types beyond images such as segmentation results, numerical statistics, and other metadata—modeled after successful analysis workflow software.
3. Broaden the image data model – We will extend ImageJ to handle important image types, including large numbers of image planes, dimensionality beyond 5D, high-resolution image planes, and images stored in databases. The data model will also leverage image metadata, which is required as imaging experiments become more complex.

Aim II – Expand functionality by interfacing ImageJ with existing open-source programs

To ensure that development proceeds in a practical direction that maximizes interoperability, and to improve ImageJ’s functionality, we will interface the improved ImageJ framework with two existing open source biology applications, VisBio and CellProfiler. We will rework VisBio as a suite of ImageJ extensions, taking advantage of the improved data model to provide robust 3D visualization and analysis of massive multidimensional image. We will link CellProfiler with ImageJ to allow users to execute analysis workflows between the two, such that algorithms can be shared.

Aim III – Grow community-driven development while maintaining compatibility

ImageJ has a strong, established user base, with thousands of plugins and macros for performing a wide variety of tasks; the proposed changes to ImageJ will be done in a manner that preserves the functionality of existing code. To foster participation, understanding and enthusiasm from a growing community, we propose the adoption of several “best practices” in line with other modern, successful open source projects, which when taken together will build on ImageJ's solid foundation of community-driven development. This will include establishing a public source code repository with a standard location for extensions, deploying unit tests and a continuous integration system, increasing usage of modern language features, and developing an integrated online help system.

Conclusion

Together, these improvements will enhance the functionality and interoperability of ImageJ, solidifying and expanding the efforts of the community of researchers who rely on it for important research across all areas of biomedicine.