-
Notifications
You must be signed in to change notification settings - Fork 0
/
Copy pathproblem_class_ground_truth.html
682 lines (631 loc) · 50.4 KB
/
problem_class_ground_truth.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
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
318
319
320
321
322
323
324
325
326
327
328
329
330
331
332
333
334
335
336
337
338
339
340
341
342
343
344
345
346
347
348
349
350
351
352
353
354
355
356
357
358
359
360
361
362
363
364
365
366
367
368
369
370
371
372
373
374
375
376
377
378
379
380
381
382
383
384
385
386
387
388
389
390
391
392
393
394
395
396
397
398
399
400
401
402
403
404
405
406
407
408
409
410
411
412
413
414
415
416
417
418
419
420
421
422
423
424
425
426
427
428
429
430
431
432
433
434
435
436
437
438
439
440
441
442
443
444
445
446
447
448
449
450
451
452
453
454
455
456
457
458
459
460
461
462
463
464
465
466
467
468
469
470
471
472
473
474
475
476
477
478
479
480
481
482
483
484
485
486
487
488
489
490
491
492
493
494
495
496
497
498
499
500
501
502
503
504
505
506
507
508
509
510
511
512
513
514
515
516
517
518
519
520
521
522
523
524
525
526
527
528
529
530
531
532
533
534
535
536
537
538
539
540
541
542
543
544
545
546
547
548
549
550
551
552
553
554
555
556
557
558
559
560
561
562
563
564
565
566
567
568
569
570
571
572
573
574
575
576
577
578
579
580
581
582
583
584
585
586
587
588
589
590
591
592
593
594
595
596
597
598
599
600
601
602
603
604
605
606
607
608
609
610
611
612
613
614
615
616
617
618
619
620
621
622
623
624
625
626
627
628
629
630
631
632
633
634
635
636
637
638
639
640
641
642
643
644
645
646
647
648
649
650
651
652
653
654
655
656
657
658
659
660
661
662
663
664
665
666
667
668
669
670
671
672
673
674
675
676
677
678
679
680
681
682
<!DOCTYPE html>
<html lang="en">
<head>
<title>BIAFLOWS NEUBIAS-WG5 : user guide documentation</title>
<!-- Meta -->
<meta charset="utf-8">
<meta http-equiv="X-UA-Compatible" content="IE=edge">
<meta name="viewport" content="width=device-width, initial-scale=1.0">
<meta name="description" content="">
<meta name="author" content="">
<link rel="shortcut icon" href="favicon.ico">
<link href='https://fonts.googleapis.com/css?family=Open+Sans:300italic,400italic,600italic,700italic,800italic,400,300,600,700,800' rel='stylesheet' type='text/css'>
<!-- FontAwesome JS -->
<script defer src="https://use.fontawesome.com/releases/v5.8.2/js/all.js" integrity="sha384-DJ25uNYET2XCl5ZF++U8eNxPWqcKohUUBUpKGlNLMchM7q4Wjg2CUpjHLaL8yYPH" crossorigin="anonymous"></script>
<!-- Global CSS -->
<link rel="stylesheet" href="assets/plugins/bootstrap/css/bootstrap.min.css">
<!-- Plugins CSS -->
<link rel="stylesheet" href="assets/plugins/prism/prism.css">
<link rel="stylesheet" href="assets/plugins/elegant_font/css/style.css">
<!-- Theme CSS -->
<link id="theme-style" rel="stylesheet" href="assets/css/styles.css">
</head>
<body class="body-pink">
<div class="page-wrapper">
<!-- ******Header****** -->
<header id="header" class="header">
<div class="container">
<div class="branding">
<h1 class="logo">
<a href="index.html">
<span aria-hidden="true" class="icon_documents_alt icon"></span>
<span class="text-highlight">BIA</span><span class="text-bold">FLOWS</span>
</a>
</h1>
</div>
</div><!--//container-->
</header><!--//header-->
<div class="doc-wrapper">
<div class="container" style="max-width:1600px">
<div id="doc-header" class="doc-header text-center">
<h1 class="doc-title">Problem class, ground truth annotations and reported metrics</h1>
<div class="meta"><i class="far fa-clock"></i> Last updated: November 18st, 2020</div>
</div>
<!--//doc-header-->
<div class="doc-body row">
<div class="doc-content col-md-9 col-12 order-1">
<div class="content-inner">
<section id="introduction-section" class="doc-section">
<h2 class="section-title">Introduction</h2>
<p class="section-block">
<p>
To perform benchmarking, ground truth annotations should be encoded in a format
that is specific to the associated problem class.
<span style="color:#fe7f7f;">BIA</span> workflows are also expected to output
results in the same format.
</p>
<p>
Currently 9 problem classes are supported in
<span style="color:#fe7f7f;">BIA</span>FLOWS and their respective
annotation formats and computed benchmark metrics are described below.
</p>
<p>
<u>Note</u>: each problem class has a long name (explicit)
and short name (e.g. Object Segmentation / ObjSeg). The same hold for metrics (e.g. DICE / DC).
</p>
<p>
A description of each benchmark is available on the workflow runs result table by clicking on the
<i data-v-0503ad4e="" class="fas fa-info-circle"></i> symbol.
</p>
<div class="screenshot-holder">
<img class="img-fluid" src="assets/images/benchmark_01.png" alt="screenshot"></a>
</div>
</section><!--//doc-section-->
<!--
<section id="steps-section" class="doc-section">
<h2 class="section-title">Problem Class</h2>
<div id="ObjSeg" class="section-block">
<h4>Object Segmentation (ObjSeg)</h4>
<h6>Task</h6>
<p>Delineate objects or isolated regions</p>
<h6>Object Encoding</h6>
<p>2D/3D label masks with foreground > 0 (one unique ID per object), background = 0</p>
<h6>Reported metrics</h6>
<ul>
<li><b>Mean Average Precision</b> computed by <a href="https://www.kaggle.com/c/data-science-bowl-2018/overview/evaluation" target="_blank">Data Science Bowl 2018</a> Python code.</li>
<li>DICE (DC) and AVERAGE_HAUSDORFF_DISTANCE (AHD), computed by
<a href="http://www.visceral.eu/resources/evaluatesegmentation-software/" target="_blank">VISCERAL</a>
executable (archived
<a href="https://github.com/Neubias-WG5/neubiaswg5-utilities/blob/master/bin/Visceral" target="_blank" >here</a>)</li>
<li>Fraction overlap (FOVL) computed by <a href="https://github.com/Neubias-WG5/neubiaswg5-utilities" target="_blank">custom</a> Python code</li>
</ul>
</div>
<div id="SptCnt" class="section-block">
<h4>Spot / object counting (SptCnt)</h4>
<h6>Task</h6>
<p>Estimate the number of objects</p>
<h6>Object Encoding</h6>
<p>2D/3D binary masks, exactly 1 spot/object per non null pixel</p>
<h6>Reported metrics</h6>
<ul>
<li>RELATIVE_ERROR_COUNT (REC), computed by <a href="https://github.com/Neubias-WG5/neubiaswg5-utilities" target="_blank">custom</a> Python code.</li>
</ul>
</div>
<div id="ObjDet" class="section-block">
<h4>Spot / object detection (ObjDet)</h4>
<h6>Task</h6>
<p>Detect objects in an image (e.g. nucleus)</p>
<h6>Object Encoding</h6>
<p>2D/3D binary masks, exactly 1 object per non null pixel</p>
<h6>Reported metrics</h6>
<ul>
<li><b>F1_SCORE (F1)</b>, CONFUSION_MATRIX (TP, FN, FP),PRECISION (PR),
RECALL (RE), Distance RMSE (RMSE), computed by
<a href="http://bioimageanalysis.org/track/" target="_blank">Particle Tracking Challenge</a>
metric Java code (particle matching only, archived
<a href="https://github.com/Neubias-WG5/neubiaswg5-utilities" target="_blank">here</a>
in <code>bin/DetectionPerformance.jar</code>).</li>
</ul>
</div>
<div id="PixCla" class="section-block">
<h4>Pixel/Voxel Classification (PixCla)</h4>
<h6>Task</h6>
<p>Estimate pixels class</p>
<h6>Object Encoding</h6>
<p>2D/3D class masks, gray level encodes pixel/voxel class, background = 0</p>
<h6>Reported metrics</h6>
<ul>
<li>F1_SCORE (F1), ACCURACY (ACC), PRECISION (PR), RECALL (RE), computed by
<a href="https://github.com/Neubias-WG5/neubiaswg5-utilities" target="_blank">custom</a> Python code.</li>
</ul>
</div>
<div id="TreTrc" class="section-block">
<h4>Filament Tree Tracing (TreTrc)</h4>
<h6>Task</h6>
<p>Estimate the medial axis of a connected filament tree network (one per image)</p>
<h6>Object Encoding</h6>
<p>SWC file</p>
<h6>Reported metrics</h6>
<ul>
<li>UNMATCHED_VOXEL_RATE (UVR), computed by
<a href="https://github.com/Neubias-WG5/neubiaswg5-utilities" target="_blank">custom</a> Python code</li>
<li>NetMets metrics: Geometric False Negative rate (FNR),
Geometric False Positive rate (FPR) computed by
<a href="https://github.com/Neubias-WG5/neubiaswg5-utilities/blob/master/neubiaswg5/metrics/netmets_obj.py" target="_blank">NetMets Python code.</a></li>
</ul>
<h6>Metrics parameters</h6>
<ul>
<li>GATING_DIST (UVR): Maximum distance between skeleton voxels in reference
and prediction skeletons to be considered as matched (default = 5 pix)</li>
<li>Sigma (NetMets): tolerance in centreline position (default: 5 pix).</li>
</ul>
</div>
<div id="LooTrc" class="section-block">
<h4>Filament Networks Tracing (LooTrc)</h4>
<h6>Task</h6>
<p>Estimate the medial axis of one or several connected filament network(s)</p>
<h6>Object Encoding</h6>
<p>2D/3D skeleton binary masks with skeleton pixels > 0, background = 0</p>
<h6>Reported metrics</h6>
<ul>
<li>UNMATCHED_VOXEL_RATE (UVR), computed by
<a href="https://github.com/Neubias-WG5/neubiaswg5-utilities" target="_blank">custom</a> Python code</li>
<li>NetMets metrics: Geometric False Negative rate (FNR),
Geometric False Positive rate (FPR) computed by
<a href="https://github.com/Neubias-WG5/neubiaswg5-utilities/blob/master/neubiaswg5/metrics/netmets_obj.py" target="_blank">NetMets Python code.</a></li>
</ul>
<h6>Metrics parameters</h6>
<ul>
<li>GATING_DIST (UVR): Maximum distance between skeleton voxels in reference
and prediction skeletons to be considered as matched (default = 5 pix)</li>
<li>Sigma (NetMets): tolerance in centreline position (default: 5 pix)</li>
<li>Skeleton sampling distance (NetMets): skeletons are sampled to be converted
to SWC models. (default: 3 voxels, default Z Ratio: 1).</li>
</ul>
</div>
<div id="LndDet" class="section-block">
<h4>Landmark Detection (LndDet)</h4>
<h6>Task</h6>
<p>Estimate the position of specific feature points</p>
<h6>Object Encoding</h6>
<p>2D/3D class masks, exactly 1 landmark per non null pixel, gray level encodes
landmark class (1 to N, N is the number of landmarks)</p>
<h6>Reported metrics</h6>
<ul>
<li>Number of reference / predicted landmarks (NREF, NPRED)</li>
<li>Mean distance from
predicted landmarks to closest reference landmarks with same class (MRE).</li>
All metrics computed by
<a href="https://github.com/Neubias-WG5/neubiaswg5-utilities" target="_blank">custom</a> Python code.</li>
</ul>
</div>
<div id="PrtTrk" class="section-block">
<h4>Particle Tracking (PrtTrk)</h4>
<h6>Task</h6>
<p>Estimate the tracks followed by particles (no division)</p>
<h6>Object Encoding</h6>
<p>2D/3D label masks, exactly 1 particle per non null pixel, gray level encodes particle track ID</p>
<h6>Reported metrics</h6>
<ul>
<li>Normalized pairing score alpha (NPSA)</li>
<li>Full normalized pairing score beta (FNPSB)</li>
<li>Number of reference tracks (NRT)</li>
<li>Number of candidate tracks (NCT)</li>
<li>Jaccard Similarity Tracks (JST)</li>
<li>Number of paired tracks (NPT)</li>
<li>Number of missed tracks (NMT)</li>
<li>Number of spurious tracks (NST)</li>
<li>Number of reference detections (NRD)</li>
<li>Number of candidate detections (NCD)</li>
<li>Jaccard similarity detections (JSD)</li>
<li>Number of paired detections (NPD)</li>
<li>Number of missed detections (NMD)</li>
<li>Number of spurious detections (NSD)</li>
<p>
All metric computed by
<a href="http://bioimageanalysis.org/track/" target="_blank">Particle Tracking Challenge</a>
Java code (archived <a href="https://github.com/Neubias-WG5/neubiaswg5-utilities/blob/master/bin/TrackingPerformance.jar" target="_blank">here</a>).
</p>
</ul>
<h6>Metrics parameters</h6>
<ul>
<li>GATING_DIST (default = 5, maximum distance between particle detections
in reference / prediction tracks to be considered as matching)</li>
</ul>
</div>
<div id="ObjTrk" class="section-block">
<h4>Object Tracking (ObjTrk)</h4>
<h6>Task</h6>
<p>Estimate object tracks and segmentation masks (with possible divisions)</p>
<h6>Object Encoding</h6>
<p>2D/3D TIFF label masks, gray level encodes object ID + division text file
(see Cell Tracking Challenge format)</p>
<h6>Reported metrics</h6>
<ul>
<li>Segmentation measure (SEG), implementation archived <a href="https://github.com/Neubias-WG5/neubiaswg5-utilities/blob/master/bin/SEGMeasure" target="_blank">here</a></li>
<li>Tracking measure (TRA), implementation archived <a href="https://github.com/Neubias-WG5/neubiaswg5-utilities/blob/master/bin/TRAMeasure" target="_blank">here</a></li>
All computed from
<a href="http://celltrackingchallenge.net/evaluation-methodology/" target="_blank">Cell Tracking Challenge</a> metric command-line executables.
</ul>
</div>
</section>
-->
<!--//doc-section-->
<section id="steps-section" class="doc-section">
<h2 class="section-title">Problem Class</h2>
<div class="table-responsive">
<table class="table">
<thead>
<tr>
<th scope="col">Problem Class</th>
<th scope="col">Tasks</th>
<th scope="col">Shortname</th>
<th scope="col">Annotation</th>
<th scope="col">Example</th>
<th scope="col">Metrics</th>
<th scope="col">Tools</th>
</tr>
</thead>
<tbody>
<tr>
<th class="theme-bg-light" id="ObjSeg">Object Segmentation</th>
<td>Delineate objects or isolated regions</td>
<td>ObjSeg</td>
<td>
<figure class="figure docs-figure py-3">
<a href="assets/images/class-01-label_mask.png" data-title="2D/3D label masks with foreground > 0 (one unique ID per object), background = 0" data-toggle="lightbox"><img class="figure-img img-fluid shadow rounded" src="assets/images/class-01-label_mask.png" alt="Label masks" style="width: 600px;"></a></figure></br>Label masks
</td>
<td style="white-space: nowrap;">
<a class="theme-link" href="assets/images/01-Pixel-classification.zip" target="_blank">sample <svg class="svg-inline--fa fa-external-link-alt fa-w-16" aria-hidden="true" focusable="false" data-prefix="fas" data-icon="external-link-alt" role="img" xmlns="http://www.w3.org/2000/svg" viewBox="0 0 512 512" data-fa-i2svg=""><path fill="currentColor" d="M432,320H400a16,16,0,0,0-16,16V448H64V128H208a16,16,0,0,0,16-16V80a16,16,0,0,0-16-16H48A48,48,0,0,0,0,112V464a48,48,0,0,0,48,48H400a48,48,0,0,0,48-48V336A16,16,0,0,0,432,320ZM488,0h-128c-21.37,0-32.05,25.91-17,41l35.73,35.73L135,320.37a24,24,0,0,0,0,34L157.67,377a24,24,0,0,0,34,0L435.28,133.32,471,169c15,15,41,4.5,41-17V24A24,24,0,0,0,488,0Z"></path></svg> </a>
</td>
<td>
<ul>
<li><b>Mean Average Precision</b> computed by <a href="https://www.kaggle.com/c/data-science-bowl-2018/overview/evaluation" target="_blank">Data Science Bowl 2018</a> Python code</li>
<li>DICE (DC), computed by
<a href="http://www.visceral.eu/resources/evaluatesegmentation-software/" target="_blank">VISCERAL</a>
executable (archived
<a href="https://github.com/Neubias-WG5/neubiaswg5-utilities/blob/master/bin/Visceral" target="_blank" >here</a>)</li>
<li>AVERAGE_HAUSDORFF_DISTANCE (AHD), computed by
<a href="http://www.visceral.eu/resources/evaluatesegmentation-software/" target="_blank">VISCERAL</a>
executable (archived
<a href="https://github.com/Neubias-WG5/neubiaswg5-utilities/blob/master/bin/Visceral" target="_blank" >here</a>)</li>
<li>Fraction overlap (FOVL) computed by <a href="https://github.com/Neubias-WG5/neubiaswg5-utilities" target="_blank">custom</a> Python code</li>
</ul>
</td>
<td>
<ul>
<li><a href="https://github.com/saalfeldlab/paintera">Paintera (3D)</a></li>
<li><a href="http://www.itksnap.org/pmwiki/pmwiki.php">ITK-snap (3D)</a></li>
<li><a href="https://doc.cytomine.org/AnnotationsV2?structure=UsersV2"</a>Cytomine (2D)</li>
<li><a href="https://qupath.readthedocs.io/en/latest/docs/starting/annotating.html">QuPath (2D)</a></li>
<li><a href="http://www.3d-cell-annotator.org/index.html">3D Cell annotator</a></li>
<li><a href="https://www.ilastik.org/index.html">Ilastik (3D)</a></li>
</ul>
</td>
</tr>
<tr>
<th class="theme-bg-light" id="PixCla">Pixel/Voxel classification</th>
<td>Estimate pixels class</td>
<td>PixCla</td>
<td>
<figure class="figure docs-figure py-3">
<a href="assets/images/class-01-label_mask.png" data-title="2D/3D class masks, gray level encodes pixel/voxel class, background = 0" data-toggle="lightbox"><img class="figure-img img-fluid shadow rounded" src="assets/images/class-01-label_mask.png" alt="Label masks" style="width: 600px;"></a></figure></br>Label masks
</td>
<td style="white-space: nowrap;">
<a class="theme-link" href="assets/images/00-Pixel-classification.zip" target="_blank">sample <svg class="svg-inline--fa fa-external-link-alt fa-w-16" aria-hidden="true" focusable="false" data-prefix="fas" data-icon="external-link-alt" role="img" xmlns="http://www.w3.org/2000/svg" viewBox="0 0 512 512" data-fa-i2svg=""><path fill="currentColor" d="M432,320H400a16,16,0,0,0-16,16V448H64V128H208a16,16,0,0,0,16-16V80a16,16,0,0,0-16-16H48A48,48,0,0,0,0,112V464a48,48,0,0,0,48,48H400a48,48,0,0,0,48-48V336A16,16,0,0,0,432,320ZM488,0h-128c-21.37,0-32.05,25.91-17,41l35.73,35.73L135,320.37a24,24,0,0,0,0,34L157.67,377a24,24,0,0,0,34,0L435.28,133.32,471,169c15,15,41,4.5,41-17V24A24,24,0,0,0,488,0Z"></path></svg> </a>
</td>
<td>
<ul>
<li><b>F1_SCORE (F1)</b> computed by
<a href="https://github.com/Neubias-WG5/neubiaswg5-utilities" target="_blank">custom</a> Python code</li>
<li>RECALL (RE) computed by
<a href="https://github.com/Neubias-WG5/neubiaswg5-utilities" target="_blank">custom</a> Python code</li>
<li>ACCURACY (ACC) computed by
<a href="https://github.com/Neubias-WG5/neubiaswg5-utilities" target="_blank">custom</a> Python code</li>
<li>PRECISION (PR) computed by
<a href="https://github.com/Neubias-WG5/neubiaswg5-utilities" target="_blank">custom</a> Python code</li>
</ul>
</td>
<td>
<ul>
<li><a href="https://github.com/saalfeldlab/paintera">Paintera (3D)</a></li>
<li><a href="http://www.itksnap.org/pmwiki/pmwiki.php">ITK-snap (3D)</a></li>
<li><a href="https://doc.cytomine.org/AnnotationsV2?structure=UsersV2"</a>Cytomine (2D)</li>
<li><a href="https://qupath.readthedocs.io/en/latest/docs/starting/annotating.html">QuPath (2D)</a></li>
<li><a href="http://www.3d-cell-annotator.org/index.html">3D Cell annotator</a></li>
<li><a href="https://www.ilastik.org/index.html">Ilastik (3D)</a></li>
</ul>
</td>
</tr>
<tr>
<th class="theme-bg-light" id="SptCnt">Spot/Object Counting</th>
<td>Estimate the number of objects</td>
<td>SptCnt</td>
<td>
<figure class="figure docs-figure py-3">
<a href="assets/images/class-03-spot_detection.png" data-title="2D/3D binary masks, exactly 1 spot/object per non null pixel" data-toggle="lightbox"><img class="figure-img img-fluid shadow rounded" src="assets/images/class-03-spot_detection.png" alt="2D/3D binary masks, exactly 1 spot/object per non null pixel" style="width: 600px;"></a></figure></br>Binary masks
</td>
<td style="white-space: nowrap;">
<a class="theme-link" href="assets/images/03-spot-detection.zip" target="_blank">sample <svg class="svg-inline--fa fa-external-link-alt fa-w-16" aria-hidden="true" focusable="false" data-prefix="fas" data-icon="external-link-alt" role="img" xmlns="http://www.w3.org/2000/svg" viewBox="0 0 512 512" data-fa-i2svg=""><path fill="currentColor" d="M432,320H400a16,16,0,0,0-16,16V448H64V128H208a16,16,0,0,0,16-16V80a16,16,0,0,0-16-16H48A48,48,0,0,0,0,112V464a48,48,0,0,0,48,48H400a48,48,0,0,0,48-48V336A16,16,0,0,0,432,320ZM488,0h-128c-21.37,0-32.05,25.91-17,41l35.73,35.73L135,320.37a24,24,0,0,0,0,34L157.67,377a24,24,0,0,0,34,0L435.28,133.32,471,169c15,15,41,4.5,41-17V24A24,24,0,0,0,488,0Z"></path></svg> </a>
</td>
<td>
<ul>
<li><b>RELATIVE_ERROR_COUNT (REC)</b>, computed by <a href="https://github.com/Neubias-WG5/neubiaswg5-utilities" target="_blank">custom</a> Python code.</li>
</ul>
</td>
<td>
<ul>
<li><a href="https://github.com/saalfeldlab/paintera">Paintera (3D)</a></li>
<li><a href="http://www.itksnap.org/pmwiki/pmwiki.php">ITK-snap (3D)</a></li>
<li><a href="https://doc.cytomine.org/AnnotationsV2?structure=UsersV2"</a>Cytomine (2D)</li>
<li><a href="https://qupath.readthedocs.io/en/latest/docs/starting/annotating.html">QuPath (2D)</a></li>
<li><a href="https://www.ilastik.org/index.html">Ilastik (3D)</a></li>
<li><a href="https://imagej.nih.gov/ij/docs/guide/146-19.html#toc-Subsection-19.6">ImageJ with Multi-point Tool</a></li>
</ul>
</td>
</tr>
<tr>
<th class="theme-bg-light" id="ObjDet">Spot/Object Detection</th>
<td>Detect objects in an image (e.g. nucleus)</td>
<td>ObjDet</td>
<td>
<figure class="figure docs-figure py-3">
<a href="assets/images/class-03-spot_detection.png" data-title="2D/3D binary masks, exactly 1 object per non null pixel" data-toggle="lightbox"><img class="figure-img img-fluid shadow rounded" src="assets/images/class-03-spot_detection.png" alt="Label masks" style="width: 600px;"></a></figure></br>Binary masks
</td>
<td style="white-space: nowrap;">
<a class="theme-link" href="assets/images/03-spot-detection.zip" target="_blank">sample <svg class="svg-inline--fa fa-external-link-alt fa-w-16" aria-hidden="true" focusable="false" data-prefix="fas" data-icon="external-link-alt" role="img" xmlns="http://www.w3.org/2000/svg" viewBox="0 0 512 512" data-fa-i2svg=""><path fill="currentColor" d="M432,320H400a16,16,0,0,0-16,16V448H64V128H208a16,16,0,0,0,16-16V80a16,16,0,0,0-16-16H48A48,48,0,0,0,0,112V464a48,48,0,0,0,48,48H400a48,48,0,0,0,48-48V336A16,16,0,0,0,432,320ZM488,0h-128c-21.37,0-32.05,25.91-17,41l35.73,35.73L135,320.37a24,24,0,0,0,0,34L157.67,377a24,24,0,0,0,34,0L435.28,133.32,471,169c15,15,41,4.5,41-17V24A24,24,0,0,0,488,0Z"></path></svg> </a>
</td>
<td>
<ul>
<li><b>F1_SCORE (F1)</b> computed by <a href="http://bioimageanalysis.org/track/" target="_blank">Particle Tracking Challenge</a>
metric Java code (particle matching only, archived <a href="https://github.com/Neubias-WG5/neubiaswg5-utilities" target="_blank">here</a>
in <code>bin/DetectionPerformance.jar</code></li>
<li>CONFUSION_MATRIX (TP, FN, FP) computed by
<a href="http://bioimageanalysis.org/track/" target="_blank">Particle Tracking Challenge</a>
metric Java code (particle matching only, archived
<a href="https://github.com/Neubias-WG5/neubiaswg5-utilities" target="_blank">here</a>
in <code>bin/DetectionPerformance.jar</code></li>
<li>PRECISION (PR) computed by
<a href="http://bioimageanalysis.org/track/" target="_blank">Particle Tracking Challenge</a>
metric Java code (particle matching only, archived
<a href="https://github.com/Neubias-WG5/neubiaswg5-utilities" target="_blank">here</a>
in <code>bin/DetectionPerformance.jar</code></li>
<li>RECALL (RE) computed by
<a href="http://bioimageanalysis.org/track/" target="_blank">Particle Tracking Challenge</a>
metric Java code (particle matching only, archived
<a href="https://github.com/Neubias-WG5/neubiaswg5-utilities" target="_blank">here</a>
in <code>bin/DetectionPerformance.jar</code></li>
<li>Distance RMSE (RMSE) computed by
<a href="http://bioimageanalysis.org/track/" target="_blank">Particle Tracking Challenge</a>
metric Java code (particle matching only, archived
<a href="https://github.com/Neubias-WG5/neubiaswg5-utilities" target="_blank">here</a>
in <code>bin/DetectionPerformance.jar</code></li>
</ul>
</td>
<td>
<ul>
<li><a href="https://github.com/saalfeldlab/paintera">Paintera (3D)</a></li>
<li><a href="http://www.itksnap.org/pmwiki/pmwiki.php">ITK-snap (3D)</a></li>
<li><a href="https://doc.cytomine.org/AnnotationsV2?structure=UsersV2"</a>Cytomine (2D)</li>
<li><a href="https://qupath.readthedocs.io/en/latest/docs/starting/annotating.html">QuPath (2D)</a></li>
<li><a href="https://www.ilastik.org/index.html">Ilastik (3D)</a></li>
<li><a href="https://imagej.nih.gov/ij/docs/guide/146-19.html#toc-Subsection-19.6">ImageJ with Multi-point Tool</a></li>
</ul>
</td>
</tr>
<tr>
<th class="theme-bg-light" id="TreTrc">Filament Tree Tracing</th>
<td>Estimate the medial axis of a connected filament tree network (one per image)</td>
<td>TreTrc</td>
<td>
<figure class="figure docs-figure py-3">
<a href="assets/images/class-07-filament_tracing.png" data-title="SWC file" data-toggle="lightbox"><img class="figure-img img-fluid shadow rounded" src="assets/images/class-07-filament_tracing_small.png" alt="SWC" style="width: 600px;"></a></figure></br>
SWC
</td>
<td style="white-space: nowrap;">
<a class="theme-link" href="assets/images/07-Filament-tracing-tree.zip" target="_blank">sample <svg class="svg-inline--fa fa-external-link-alt fa-w-16" aria-hidden="true" focusable="false" data-prefix="fas" data-icon="external-link-alt" role="img" xmlns="http://www.w3.org/2000/svg" viewBox="0 0 512 512" data-fa-i2svg=""><path fill="currentColor" d="M432,320H400a16,16,0,0,0-16,16V448H64V128H208a16,16,0,0,0,16-16V80a16,16,0,0,0-16-16H48A48,48,0,0,0,0,112V464a48,48,0,0,0,48,48H400a48,48,0,0,0,48-48V336A16,16,0,0,0,432,320ZM488,0h-128c-21.37,0-32.05,25.91-17,41l35.73,35.73L135,320.37a24,24,0,0,0,0,34L157.67,377a24,24,0,0,0,34,0L435.28,133.32,471,169c15,15,41,4.5,41-17V24A24,24,0,0,0,488,0Z"></path></svg> </a></br>
<a class="theme-link" href="http://www.neuronland.org/NLMorphologyConverter/MorphologyFormats/SWC/Spec.html" target="_blank">SWC format <svg class="svg-inline--fa fa-external-link-alt fa-w-16" aria-hidden="true" focusable="false" data-prefix="fas" data-icon="external-link-alt" role="img" xmlns="http://www.w3.org/2000/svg" viewBox="0 0 512 512" data-fa-i2svg=""><path fill="currentColor" d="M432,320H400a16,16,0,0,0-16,16V448H64V128H208a16,16,0,0,0,16-16V80a16,16,0,0,0-16-16H48A48,48,0,0,0,0,112V464a48,48,0,0,0,48,48H400a48,48,0,0,0,48-48V336A16,16,0,0,0,432,320ZM488,0h-128c-21.37,0-32.05,25.91-17,41l35.73,35.73L135,320.37a24,24,0,0,0,0,34L157.67,377a24,24,0,0,0,34,0L435.28,133.32,471,169c15,15,41,4.5,41-17V24A24,24,0,0,0,488,0Z"></path></svg> </a>
</td>
<td>
<ul>
<li><b>NetMets metrics: Geometric False Negative rate (FNR),
Geometric False Positive rate (FPR)</b> computed by
<a href="https://github.com/Neubias-WG5/neubiaswg5-utilities/blob/master/neubiaswg5/metrics/netmets_obj.py" target="_blank">NetMets Python code.</a></li>
<li>UNMATCHED_VOXEL_RATE (UVR), computed by
<a href="https://github.com/Neubias-WG5/neubiaswg5-utilities" target="_blank">custom</a> Python code</li>
</ul>
<ul>
<li>GATING_DIST (UVR): Maximum distance between skeleton voxels in reference
and prediction skeletons to be considered as matched (default = 5 pix)</li>
<li>Sigma (NetMets): tolerance in centerline position (default: 5 pix).</li>
</ul>
</td>
<td>
<ul>
<li><a href="https://imagej.net/SNT">ImageJ / SNT</a></li>
<li><a href="https://alleninstitute.org/what-we-do/brain-science/research/products-tools/vaa3d/">Vaa3D</a></li>
<li><a href="https://www.google.com/search?client=firefox-b-d&q=neutube">Neutube (2D)</a></li>
<li><a href="https://www.mbfbioscience.com/neurolucida">Neurolucida</a></li>
</ul>
</td>
</tr>
<tr>
<th class="theme-bg-light" id="LooTrc">Filament Networks Tracing</th>
<td>Estimate the medial axis of one or several connected filament network(s)</td>
<td>LooTrc</td>
<td>
<figure class="figure docs-figure py-3">
<a href="assets/images/class-08-filament_tracing_network.png" data-title="2D/3D skeleton binary masks with skeleton pixels > 0, background = 0" data-toggle="lightbox"><img class="figure-img img-fluid shadow rounded" src="assets/images/class-08-filament_tracing_network_small.png" alt="Skeleton binary masks" style="width: 600px;"></a></figure></br>
Skeleton binary masks
</td>
<td style="white-space: nowrap;">
<a class="theme-link" href="assets/images/08-Filament-tracing-networks.zip" target="_blank">sample <svg class="svg-inline--fa fa-external-link-alt fa-w-16" aria-hidden="true" focusable="false" data-prefix="fas" data-icon="external-link-alt" role="img" xmlns="http://www.w3.org/2000/svg" viewBox="0 0 512 512" data-fa-i2svg=""><path fill="currentColor" d="M432,320H400a16,16,0,0,0-16,16V448H64V128H208a16,16,0,0,0,16-16V80a16,16,0,0,0-16-16H48A48,48,0,0,0,0,112V464a48,48,0,0,0,48,48H400a48,48,0,0,0,48-48V336A16,16,0,0,0,432,320ZM488,0h-128c-21.37,0-32.05,25.91-17,41l35.73,35.73L135,320.37a24,24,0,0,0,0,34L157.67,377a24,24,0,0,0,34,0L435.28,133.32,471,169c15,15,41,4.5,41-17V24A24,24,0,0,0,488,0Z"></path></svg> </a></br>
</td>
<td>
<ul>
<li><b>NetMets metrics: Geometric False Negative rate (FNR),
Geometric False Positive rate (FPR)</b> computed by
<a href="https://github.com/Neubias-WG5/neubiaswg5-utilities/blob/master/neubiaswg5/metrics/netmets_obj.py" target="_blank">NetMets Python code.</a></li>
<li>UNMATCHED_VOXEL_RATE (UVR), computed by
<a href="https://github.com/Neubias-WG5/neubiaswg5-utilities" target="_blank">custom</a> Python code</li>
</ul>
<ul>
<li>GATING_DIST (UVR): Maximum distance between skeleton voxels in reference
and prediction skeletons to be considered as matched (default = 5 pix)</li>
<li>Sigma (NetMets): tolerance in centerline position (default: 5 pix)</li>
<li>Skeleton sampling distance (NetMets): skeletons are sampled to be converted
to <a href="https://meet.google.com/linkredirect?authuser=0&dest=http%3A%2F%2Fstim.ee.uh.edu%2Fresources%2Fsoftware%2Fnetmets%2F"> OBJ models</a>. (default: 3 voxels, default Z Ratio: 1).</li>
</ul>
</td>
<td>
<ul>
<li></li>
</ul>
</td>
</tr>
<tr>
<th class="theme-bg-light" id="LndDet">Landmark Detection</th>
<td>Estimate the position of specific feature points</td>
<td>LndDet</td>
<td>
<figure class="figure docs-figure py-3">
<a href="assets/images/class-04-Landmark-detection.png" data-title="2D/3D class masks, exactly 1 landmark per non null pixel, gray level encodes landmark class (1 to N, N is the number of landmarks) > 0, background = 0" data-toggle="lightbox"><img class="figure-img img-fluid shadow rounded" src="assets/images/class-04-Landmark-detection_small2.png" alt="2D/3D class masks, exactly 1 landmark per non null pixel, gray level encodes landmark class (1 to N, N is the number of landmarks)" style="width: 600px;"></a></figure></br>
Label masks
</td>
<td style="white-space: nowrap;">
<a class="theme-link" href="assets/images/04-Landmark-detection.zip" target="_blank">sample <svg class="svg-inline--fa fa-external-link-alt fa-w-16" aria-hidden="true" focusable="false" data-prefix="fas" data-icon="external-link-alt" role="img" xmlns="http://www.w3.org/2000/svg" viewBox="0 0 512 512" data-fa-i2svg=""><path fill="currentColor" d="M432,320H400a16,16,0,0,0-16,16V448H64V128H208a16,16,0,0,0,16-16V80a16,16,0,0,0-16-16H48A48,48,0,0,0,0,112V464a48,48,0,0,0,48,48H400a48,48,0,0,0,48-48V336A16,16,0,0,0,432,320ZM488,0h-128c-21.37,0-32.05,25.91-17,41l35.73,35.73L135,320.37a24,24,0,0,0,0,34L157.67,377a24,24,0,0,0,34,0L435.28,133.32,471,169c15,15,41,4.5,41-17V24A24,24,0,0,0,488,0Z"></path></svg> </a></br>
</td>
<td>
<ul>
<li><b>Mean distance from
predicted landmarks to closest reference landmarks with same class (MRE)</b>.</li>
<li>Number of reference / predicted landmarks (NREF, NPRED)</li>
<li>
All metrics computed by
<a href="https://github.com/Neubias-WG5/neubiaswg5-utilities" target="_blank">custom</a> Python code.</li>
</ul>
</td>
<td>
<ul>
<li></li>
</ul>
</td>
</tr>
<tr>
<th class="theme-bg-light" id="PrtTrk">Particle Tracking</th>
<td>Estimate the tracks followed by particles (no division)</td>
<td>PrtTrk</td>
<td>
<figure class="figure docs-figure py-3">
<a href="assets/images/class-01-label_mask.png" data-title="2D/3D label masks, exactly 1 particle per non null pixel, gray level encodes particle track ID" data-toggle="lightbox"><img class="figure-img img-fluid shadow rounded" src="assets/images/class-01-label_mask.png" alt="2D/3D label masks, exactly 1 particle per non null pixel, gray level encodes particle track ID" style="width: 600px;"></a></figure></br>
Label masks
</td>
<td style="white-space: nowrap;">
<a class="theme-link" href="assets/images/05-Object-Tracking.zip" target="_blank">sample <svg class="svg-inline--fa fa-external-link-alt fa-w-16" aria-hidden="true" focusable="false" data-prefix="fas" data-icon="external-link-alt" role="img" xmlns="http://www.w3.org/2000/svg" viewBox="0 0 512 512" data-fa-i2svg=""><path fill="currentColor" d="M432,320H400a16,16,0,0,0-16,16V448H64V128H208a16,16,0,0,0,16-16V80a16,16,0,0,0-16-16H48A48,48,0,0,0,0,112V464a48,48,0,0,0,48,48H400a48,48,0,0,0,48-48V336A16,16,0,0,0,432,320ZM488,0h-128c-21.37,0-32.05,25.91-17,41l35.73,35.73L135,320.37a24,24,0,0,0,0,34L157.67,377a24,24,0,0,0,34,0L435.28,133.32,471,169c15,15,41,4.5,41-17V24A24,24,0,0,0,488,0Z"></path></svg> </a></br>
</td>
<td>
<ul>
<li><b>Full normalized pairing score beta (FNPSB)</b></li>
<li>Normalized pairing score alpha (NPSA)</li>
<li>Number of reference tracks (NRT)</li>
<li>Number of candidate tracks (NCT)</li>
<li>Jaccard Similarity Tracks (JST)</li>
<li>Number of paired tracks (NPT)</li>
<li>Number of missed tracks (NMT)</li>
<li>Number of spurious tracks (NST)</li>
<li>Number of reference detections (NRD)</li>
<li>Number of candidate detections (NCD)</li>
<li>Jaccard similarity detections (JSD)</li>
<li>Number of paired detections (NPD)</li>
<li>Number of missed detections (NMD)</li>
<li>Number of spurious detections (NSD)</li>
<p>
All metric computed by
<a href="http://bioimageanalysis.org/track/" target="_blank">Particle Tracking Challenge</a>
Java code (archived <a href="https://github.com/Neubias-WG5/neubiaswg5-utilities/blob/master/bin/TrackingPerformance.jar" target="_blank">here</a>).
</p>
</ul>
<ul>
<li>GATING_DIST (default = 5, maximum distance between particle detections
in reference / prediction tracks to be considered as matching)</li>
</ul>
</td>
<td>
<ul>
<li><a href="https://imagej.nih.gov/ij/plugins/track/track.html">ImageJ Manual Tracking</a></li>
<li><a href="https://imagej.net/TrackMate">ImageJ TrackMate</a></li>
</ul>
</td>
</tr>
<tr>
<th class="theme-bg-light" id="ObjTrk">Object Tracking</th>
<td>Estimate object tracks and segmentation masks (with possible divisions)</td>
<td>ObjTrk</td>
<td>
<figure class="figure docs-figure py-3">
<a href="assets/images/class-01-label_mask.png" data-title="2D/3D label masks, gray level encodes object ID + division text file
(see Cell Tracking Challenge format)" data-toggle="lightbox"><img class="figure-img img-fluid shadow rounded" src="assets/images/class-01-label_mask.png" alt="2D/3D label masks, gray level encodes object ID + division text file
(see Cell Tracking Challenge format)" style="width: 600px;"></a></figure></br>
Label masks + <a href="http://celltrackingchallenge.net/datasets/">Division text file</a>
</td>
<td style="white-space: nowrap;">
<a class="theme-link" href="assets/images/06-Object_tracking_with_division.zip" target="_blank">sample <svg class="svg-inline--fa fa-external-link-alt fa-w-16" aria-hidden="true" focusable="false" data-prefix="fas" data-icon="external-link-alt" role="img" xmlns="http://www.w3.org/2000/svg" viewBox="0 0 512 512" data-fa-i2svg=""><path fill="currentColor" d="M432,320H400a16,16,0,0,0-16,16V448H64V128H208a16,16,0,0,0,16-16V80a16,16,0,0,0-16-16H48A48,48,0,0,0,0,112V464a48,48,0,0,0,48,48H400a48,48,0,0,0,48-48V336A16,16,0,0,0,432,320ZM488,0h-128c-21.37,0-32.05,25.91-17,41l35.73,35.73L135,320.37a24,24,0,0,0,0,34L157.67,377a24,24,0,0,0,34,0L435.28,133.32,471,169c15,15,41,4.5,41-17V24A24,24,0,0,0,488,0Z"></path></svg> </a></br>
</td>
<td>
<ul>
<li><b>Segmentation measure (SEG)</b>, implementation archived <a href="https://github.com/Neubias-WG5/neubiaswg5-utilities/blob/master/bin/SEGMeasure" target="_blank">here</a></li>
<li><b>Tracking measure (TRA)</b>, implementation archived <a href="https://github.com/Neubias-WG5/neubiaswg5-utilities/blob/master/bin/TRAMeasure" target="_blank">here</a></li>
<p>
All computed from
<a href="http://celltrackingchallenge.net/evaluation-methodology/" target="_blank">Cell Tracking Challenge</a> metric command-line executables.
</p>
</ul>
</td>
<td>
<ul>
<li><a href="https://imagej.nih.gov/ij/plugins/track/track.html">ImageJ Manual Tracking</a></li>
<li><a href="https://imagej.net/TrackMate">ImageJ TrackMate</a></li>
</ul>
</td>
</tr>
</tbody>
</table>
</section><!--//doc-section-->
</div><!--//content-inner-->
</div><!--//doc-content-->
<div class="doc-sidebar col-md-3 col-12 order-0 d-none d-md-flex">
<div id="doc-nav" class="doc-nav">
<nav id="doc-menu" class="nav doc-menu flex-column sticky">
<a class="nav-link scrollto" href="#introduction-section">Introduction</a>
<a class="nav-link scrollto" href="#steps-section">Problem Class</a>
<nav class="doc-sub-menu nav flex-column">
<a class="nav-link scrollto" href="#ObjSeg">Object Segmentation</a>
<a class="nav-link scrollto" href="#SptCnt">Spot / object counting</a>
<a class="nav-link scrollto" href="#ObjDet">Spot / object detection</a>
<a class="nav-link scrollto" href="#PixCla">Pixel/Voxel Classification</a>
<a class="nav-link scrollto" href="#TreTrc">Filament Tree Tracing</a>
<a class="nav-link scrollto" href="#LooTrc">Filament Networks Tracing</a>
<a class="nav-link scrollto" href="#LndDet">Landmark Detection</a>
<a class="nav-link scrollto" href="#PrtTrk">Particle Tracking</a>
<a class="nav-link scrollto" href="#ObjTrk">Object Tracking</a>
</nav><!--//nav-->
</nav><!--//doc-menu-->
</div>
</div><!--//doc-sidebar-->
</div><!--//doc-body-->
</div><!--//container-->
</div><!--//doc-wrapper-->
</div><!--//page-wrapper-->
<footer id="footer" class="footer text-center">
<div class="container">
<!--/* This template is released under the Creative Commons Attribution 3.0 License. Please keep the attribution link below when using for your own project. Thank you for your support. :) If you'd like to use the template without the attribution, you can buy the commercial license via our website: themes.3rdwavemedia.com */-->
<small class="copyright">Designed with <i class="fas fa-heart"></i> by <a href="https://themes.3rdwavemedia.com/" target="_blank">Xiaoying Riley</a> for developers</small>
<p>
<br><br>
<a href="http://neubias.org/" target="_blank"><img src="assets/images/logo-neubias.png" height="50"></a>
<a href="https://cost.eu/COST_Actions/ca/CA15124" target="_blank"><img src="assets/images/logo-cost.png" height="50"></a>
<a href="https://cytomine.org"><img src="assets/images/logo-cytomine-org.png" height="50"></a>
<a href="http://europe.wallonie.be/" target="_blank"><img src="assets/images/logo-feder.jpg" height="50"></a>
</p>
</div><!--//container-->
</footer><!--//footer-->
<!-- Main Javascript -->
<script type="text/javascript" src="assets/plugins/jquery-3.3.1.min.js"></script>
<script type="text/javascript" src="assets/plugins/bootstrap/js/bootstrap.min.js"></script>
<script type="text/javascript" src="assets/plugins/lightbox/dist/ekko-lightbox.min.js"></script>
<script type="text/javascript" src="assets/plugins/prism/prism.js"></script>
<script type="text/javascript" src="assets/plugins/jquery-scrollTo/jquery.scrollTo.min.js"></script>
<script type="text/javascript" src="assets/plugins/stickyfill/dist/stickyfill.min.js"></script>
<script type="text/javascript" src="assets/js/main.js"></script>
</body>
</html>