Skip to content

Latest commit

 

History

History
64 lines (46 loc) · 3.17 KB

README_HH2016.md

File metadata and controls

64 lines (46 loc) · 3.17 KB

#3D annotation of Terrascale Biomedical Imaging Data in a 2D Web

##Health Hack 2016 - Brisbane

##Team Members

##The problem http://www.healthhack.com.au/challenges#c2

##The Pitch Biomedical Image segmentation is a major roadblock to advances in neuroscience. Data sizes are increasing but the tools to analyse them have not kept pace.

For example to trace a single high resolution 1200 slice mouse model it has taken over 8 months of intensive work using traditional tools. The tools are limited to 2D, there is no easy way to collaborate and data file size is huge. Typically in the GB range but sometimes up to the TB scale. Once complete these tracings need to be constantly updated, revised and annotated.

Even viewing the data is a problem, tools are limited and operating system bound. THere is an existing open source online tiled viewing platform (www.tissuestack.org) that allows us to share and collaboratively view datasets, this has revolutionised the way we work with data as it has allowed researchers to view all their data instead of being limited to a few views. But researchers are now requesting more. They want to collaborate on annotations and paintings, they want to share and they want to be able to do it from any platform.

Online viewing has also allowed Citizen Science by allowing us to disseminate images to both schools and individuals (CSIRO scientists in schools). Viewing of data in museums, old fossils, tracing/identification of fossils. The possibilities for online painting and collaboration are immense but we don't have the tools.

The problem of tracing isn't hard but the tools to do it collaboratively and in 3D don't exist.

The image below demonstrates the existing style of desktop interface on the left that is used for small files (4GB Max) and the current web based interface that supports view but not editing of information.

HH2016 image

##Design The POC online collaborative painting tool that has been developed can be used in many other fields that use blocks of data. Mining, weather, exploration, more broad areas of biomedical imaging (microscopy) and R+D. By allowing colalborative online editing.

This POC wil allow for citizen scientists to contribute to areas of science that they were previously liminted based upon the available tools. For example Zooniverse allows annotation but it is only 2D and not collaborative.

##Implementation This weekend we have delivered a POC implementation that allows for collaborative editing.

The solution tackles the most difficult part of the existing problem owners pitch. The solution is a working prototype within the presented problems codebase and is presented as a pull reqest on this fork.

##HH2016 and Open Knowledge values The proposed solution is Open Source (GPLv3)

Working POC here: HH2016 after ...