Institute Launches ‘Hackathons’ To Improve Brain Neuron Research
April 2, 2015 in News
The Allen Institute for Brain Science has launched a series of hacker challenges as part of an effort to improve a computer’s ability to differentiate between healthy and unhealthy neurons, NPR’s “Shots” reports.
Allen Institute Executive Director Jane Roskams said highly trained researchers currently are better than computers at identifying healthy and unhealthy neurons by looking at the neurons’ shape.
Details of BigNeuron Project
The institute’s BigNeuron project aims to find the best computer algorithms to electronically turn neuron images from microscopes into three-dimensional digital images that can be analyzed. The institute then plans to share the improved algorithms with researchers across the globe.
According to Roskams, developing a standardized, automated computer-based system to analyze neurons could:
- Help address unsolved questions, such as how a neuron’s shape changes during an individual’s lifetime;
- Improve researchers’ ability to compare neurons and analyze how neuron shapes are changed by various factors, such as Alzheimer’s disease or gaining new knowledge; and
- Speed up the research process.
In addition, Roskams said a long-term goal of the BigNeuron project is to allow the analysis of neuron shapes to occur on more standard computers. Currently, such research often must be conducted on a supercomputer (Hamilton, “Shots,” NPR, 3/31).
Details of ‘Hackathons’
The project will include several “hackathons” that will bring together brain scientists and programmers to test their algorithms (Oran, MedCity News, 4/1). According to Roskams, the hackathons typically have 15 to 20 participants, and participants do not necessarily need to have extensive knowledge of neurology in order to succeed.
The first hackathon occurred last month in China, and others are scheduled to take place in the U.S. and the United Kingdom (“Shots,” NPR, 3/31).