September 2010 - Results Announced

The Diadem Challenge has concluded and the results were announced at the Scientific Conference on September 1, 2010 at Janelia Farm. Click here for more details, and here for an audio slideshow of the conference.

The History

The DIADEM Challenge began as a competition launched in April 2009. The goals of the competition were to raise awareness of the problem of automated neuronal reconstruction, spur development of automated and semi-automated algorithms, and to gauge the state of the art in the field. Competitors had one year in order to develop their algorithm(s) and submit reconstructions of the data sets that were provided for the Qualifier Round. The reconstructions were evaluated by a team of judges who chose five finalists out of 120+ registrants to compete in the Final Round, held from August 29th-September 1st 2010 at the HHMI Janelia Farm Research Campus.

The competition and conference were successful in accelerating the field of automated reconstruction, driving development and innovation as well as bringing together and strengthening collaboration between neuroscientists and computer scientists. This includes the creation of the DIADEM Consortium for the purposes of preserving the momentum produced by the competition (link to site or mailing list?). Indeed, the algorithms presented in the Final Round showed great potential in reaching the threshold set at the inception of the competition (a 20x speed-up to reach gold standard quality in terms of human interaction in a majority of the data sets). This potential was recognized in the form of commendation and jury awards to:

Badrinath Roysam Team, $25,000
"for the better overall generality of their program in producing robust reconstructions by integration of human and machines interactions."

Armen Stepanyants Team, $25,000
"for the better overall biological results in the spirit of pure automation."

Eugene Myers Team, $15,000
"for the excellent quality and strength of their algorithm."

German Gonzalez Team, $10,000
"for their deeper potential, more original approach, and ultimate scalability of their proposed solution."

Deniz Erdogmus Team
"for elevating themselves above the current state of automated reconstructions...with a deep understanding of the technical and scientific problems."

To provide a more complete historical context, the competition framework, its rules and evaluation methods, the conference, and the output (reconstruction SWC files) from the Qualifier and Final Round of competition are available here on this website.