Posted on Feb 16, 2017 by inkycatz

We’re not done yet on our latest batch of ALM mice neurons, but we’re very close. Have you checked back in on your traces lately? Look now or read on for spoilers!
Several of the comments on ALM-03 and ALM-04 indicate the pathways may be faint, but still have room to explore.
Mouse-ALM-05 is doing pretty well on that front!

(read more)
Posted on Feb 13, 2017 by inkycatz

Mouse-ALM-05 is available. We currently are estimating ALM-03 and ALM-04 challenges will be returning to the scientists for additional processing and commentary around mid-week.


(read more)
Posted on Feb 9, 2017 by inkycatz

Mouse-ALM-04 is here, as originally referred to earlier this week. You can give it a try now.

(read more)
Posted on Feb 9, 2017 by inkycatz

We have a fresh batch of feedback courtesy of scientist “alexh” to share with the community today. Have you been regularly checking in on your traces? We recommend checking daily if you can, since the community goes through neurons so quickly these days you don’t want to miss a chance to boost your score. alexh keeps us on our toes, and often points out areas that need tracing, and occasional course corrections if we’ve veered off in the wrong direction. This is the latest in a series of summaries of “how we did” afterwards.

Scientist notes on 03: Mouse Neuron V1 is our oldest available feedback.
Read more from the Gold Standard Comparisons V1-04 through V1-07
Check out Gold Standard Comparisons V1-08 through V1-10

For those new to the feedback series, the "gold standard" label is given to things that have already been currently reconstructed by experts (either elsewhere or at the Allen Institute for Brain Science) that we use to evaluate against to measure the effectiveness of our tools and help identify areas that could use more education efforts.
In all the images for this post, the gold standard is in red (axon) and blue (dendrites), while the consensus is green.
Estimated Percent Complete:

Three dendrite tips were missed in the consensus, and much of the faintest axon was missed. There is very little false positive, which is great. This is the kind of result that is particularly useful as a starting point for us. It would allow us to quickly complete the dendrites, and then do a sweep to capture any missed axon.

Estimated Percent Complete:

The dendrites are perfect. There is almost zero positive. The 30% of missed signal was all faint axon. This would also be another good starting point for us.

Estimated Percent Complete:

The dendrites are excellent. There is some false positive, likely due to a nearby patch attempt. As in the other cases, the majority of missed signal is faint axon. This would also be another good starting point for us.

Estimated Percent Complete: 72%

The dendrites are excellent and this neuron contains very few false positives. Most of the missed length was faint axon. The edge of the tissue appears to have been interpreted as signal. This would also be another good starting point for us.

Estimated Percent Complete: 38%

This is an example of very dense axon, which can be understandably difficult to capture. The dendrites are all captured, and look good. On the axon side, lots of weak axon signal was missed and we noted a number of false positives in the bottom of the image. This is likely to be axon from another nearby patch attempt. Using this as a starting point would still be better for us than starting from scratch or using an autotrace.

Bring us your science questions on the
General Discussion Forum.
Good job, everyone, and thank you alexh for taking time out to send post-challenge notes and your continued feedback so we can keep learning.

(read more)