-
Notifications
You must be signed in to change notification settings - Fork 0
/
Copy pathindex.html
103 lines (89 loc) · 8.08 KB
/
index.html
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
<html xmlns="http://www.w3.org/1999/xhtml">
<head><meta http-equiv="Content-Type" content="text/html; charset=UTF-8">
<title>XPRESS: Xray Projectomic Reconstruction - Extracting Segmentation with Skeletons</title>
<!-- Meta tags for search engines to crawl -->
<meta name="robots" content="index,follow">
<meta name="description" content="">
<meta name="keywords" content="X-ray microscopy; connectomics; white matter axons; skeletons; 3D instance segmentation">
<!-- Fonts and stuff -->
<link href="./src/css" rel="stylesheet" type="text/css">
<link rel="stylesheet" type="text/css" href="./src/project.css" media="screen">
<link rel="stylesheet" type="text/css" media="screen" href="./src/iconize.css">
<script async="" src="./src/prettify.js"></script>
<style>
.highlight {
padding: 1.5rem;
margin-right: 0;
margin-left: 0;
}
</style>
</head>
<body>
<div id="content">
<div id="content-inner">
<div class="section head">
<h1>XPRESS: Xray Projectomic Reconstruction - Extracting Segmentation with Skeletons</h1>
<br/>
<div class="authors">
Mukul Narwani<sup>1</sup>
Mark Larson<sup>2</sup>
<a href="https://trivoldus28.github.io/">Tri Nguyen</a><sup>1</sup>
Yicong Li<sup>3</sup>
Shuhan Xie<sup>2</sup>
<a href="https://vcg.seas.harvard.edu/people/hanspeter-pfister">Hanspeter Pfister</a><sup>3</sup>
<a href="https://people.csail.mit.edu/donglai/">Donglai Wei</a><sup>4</sup> <br>
<a href="https://people.csail.mit.edu/shanir/">Nir Shavit</a><sup>5</sup>
<a href="https://people.csail.mit.edu/lumi/">Lu Mi</a><sup>5</sup>
<a href="https://www.pacureanu.com/">Alexandra Pacureanu</a><sup>6</sup>
<a href="https://brain.harvard.edu/?people=wei-chung-allen-lee">Wei-Chung Lee</a><sup>1,7</sup>
<a href="https://scholar.harvard.edu/akuan/home">Aaron Kuan</a><sup>1</sup>
</div>
<div class="affiliations">
<sup>1</sup>Department of Neurobiology, Harvard Medical School
<sup>2</sup>College of Science, Northeastern University
<sup>3</sup>John A. Paulson School of Engineering and Applied Sciences, Harvard University
<sup>4</sup>Department of Computer Science, Boston College <br/>
<sup>5</sup>Computer Science & Artificial Intelligence Laboratory, MIT
<sup>6</sup>ESRF, The European Synchrotron <br/>
<sup>7</sup>F.M. Kirby Neurobiology Center, Boston Children's Hospital, Harvard Medical School
</div>
</div>
<center>
<font size=4>
<a href="https://xpress.grand-challenge.org/">Grand Challenge</a><br/>
Dataset Visualization<br/>
[<a href="https://neuroglancer-demo.appspot.com/#!%7B%22dimensions%22:%7B%22x%22:%5B3.3e-8%2C%22m%22%5D%2C%22y%22:%5B3.3e-8%2C%22m%22%5D%2C%22z%22:%5B3.3e-8%2C%22m%22%5D%7D%2C%22position%22:%5B600.5%2C600.5%2C600.5%5D%2C%22crossSectionScale%22:0.5886049696783552%2C%22projectionScale%22:1571.2377855505524%2C%22layers%22:%5B%7B%22type%22:%22image%22%2C%22source%22:%22precomputed://gs://lee-pacureanu_data-exchange_us-storage/ls2892_LTP/2102/s22/s22_WM_100nm_rec_db27_400_upscaled_cutout5_3x.tif%22%2C%22tab%22:%22source%22%2C%22name%22:%22s22_WM_100nm_rec_db27_400_upscaled_cutout5_3x.tif%22%7D%2C%7B%22type%22:%22segmentation%22%2C%22source%22:%22precomputed://gs://lee-pacureanu_data-exchange_us-storage/ls2892_LTP/2102/s22/gt_labels_cutout5_train%22%2C%22tab%22:%22source%22%2C%22name%22:%22gt_labels%22%7D%2C%7B%22type%22:%22segmentation%22%2C%22source%22:%22precomputed://https://catmaid3.hms.harvard.edu/cb2o2/staged_alignment_v3/delete_me/ng_skeletons/cutout5%22%2C%22tab%22:%22source%22%2C%22segments%22:%5B%221%22%5D%2C%22segmentQuery%22:%221%22%2C%22name%22:%22gt_skeletons%22%7D%5D%2C%22selectedLayer%22:%7B%22visible%22:true%2C%22layer%22:%22s22_WM_100nm_rec_db27_400_upscaled_cutout5_3x.tif%22%7D%2C%22layout%22:%22yz%22%7D">Train</a>]
[<a href="https://neuroglancer-demo.appspot.com/#!%7B%22dimensions%22:%7B%22x%22:%5B3.3e-8%2C%22m%22%5D%2C%22y%22:%5B3.3e-8%2C%22m%22%5D%2C%22z%22:%5B3.3e-8%2C%22m%22%5D%7D%2C%22position%22:%5B600.5%2C600.5%2C600.5%5D%2C%22crossSectionScale%22:1.6989323086185508%2C%22projectionScale%22:2048%2C%22layers%22:%5B%7B%22type%22:%22image%22%2C%22source%22:%22precomputed://gs://lee-pacureanu_data-exchange_us-storage/ls2892_LTP/2102/s22/s22_WM_100nm_rec_db27_400_upscaled_cutout4_3x.tif%22%2C%22tab%22:%22source%22%2C%22name%22:%22s22_WM_100nm_rec_db27_400_upscaled_cutout4_3x.tif%22%7D%2C%7B%22type%22:%22segmentation%22%2C%22source%22:%22precomputed://https://catmaid3.hms.harvard.edu/cb2o2/staged_alignment_v3/delete_me/ng_skeletons/cutout4%22%2C%22tab%22:%22segments%22%2C%22segments%22:%5B%221%22%5D%2C%22segmentQuery%22:%221%22%2C%22name%22:%22skeletons_gt%22%7D%5D%2C%22layout%22:%224panel%22%7D">Validation</a>]
</font>
</center>
<br/>
<br/>
<br/>
<center><img src="./src/xpress_teaser.png" border="0" width="80%"></center>
<div class="section abstract">
<h2>Abstract</h2>
<br>
<p> The wiring and connectivity of neurons forms a critical structural basis for the function of the nervous system. Advances in volume electron microscopy (EM) and image segmentation have enabled mapping of circuit diagrams (connectomics) within local regions of the mouse brain. However, applying volume EM over the whole brain is not currently feasible due to technological challenges. As a result, comprehensive maps of long-range connections between brain regions are lacking. <br>
Recently, we demonstrated that X-ray holographic nano-tomography (XNH) can provide high-resolution images of brain tissue at a much larger scale than EM. In particular, XNH is well-suited to resolve large, myelinated axon tracts (white matter) that make up the bulk of long-range connections (projections) and are critical for inter-region communication. Thus, XNH provides an imaging solution for brain-wide projectomics. However, because XNH data is typically collected at lower resolutions and larger fields-of-view than EM, accurate segmentation of XNH images remains an important challenge that we present here. <br>
In this task, we provide volumetric XNH images of cortical white matter axons from the mouse brain at 100 nm per voxel isotropic resolution. Additionally, we provide ground truth annotations for axon trajectories. Manual voxel-wise annotation of ground truth is a time-consuming bottleneck for training segmentation networks. On the other hand, skeleton-based ground truth is much faster to annotate, and sufficient to determine connectivity. Therefore, we encourage participants to develop methods to leverage skeleton-based training. To this end, we provide two types of training (validation) sets: a small volume of voxel-wise annotations and a larger volume with skeleton-based annotations. The participants will have the flexibility to use either or both of the provided annotations to train their models, and are challenged to submit an accurate voxel-wise prediction on the test volume. Entries will be evaluated on how accurately the submitted segmentations agree with the ground-truth skeleton annotations.
</p>
</div>
<br>
<div class="section materials">
<h2>Data & Evaluation</h2>
<center><img src="./src/xpress_seg.png" border="0" width="80%"></center>
<p>(a) Evaluation metric - Expected Run Length (ERL) for baseline model evaluated on the validation set.
(b)XNH image data.
(c) Example axon segmentation from the baseline model.
(d) Example of split error and merge error.
</p>
</div>
<br>
<br>
<div class="section materials">
<h2>Acknowledgement</h2>
<p>
This project received funding from the NIH (EB032217) to ATK and (MH128949) to WCAL and AP, and the European Research Council (ERC) under the European Union's Horizon 2020 research and innovation programme (grant agreement n°852455) to AP. We acknowledge the ESRF for granting beamtime for proposal LS2892.
</p>
</div>
</div></div></body></html>