Diff of /talk/Untitled0.ipynb [000000] .. [171cba]

Switch to unified view

a b/talk/Untitled0.ipynb
1
{
2
 "metadata": {
3
  "celltoolbar": "Slideshow",
4
  "name": "",
5
  "signature": "sha256:a1c50636bceeed02c8c18724931a58c60be26150f5402e0d284e06e7c7347b67"
6
 },
7
 "nbformat": 3,
8
 "nbformat_minor": 0,
9
 "worksheets": [
10
  {
11
   "cells": [
12
    {
13
     "cell_type": "heading",
14
     "level": 1,
15
     "metadata": {
16
      "slideshow": {
17
       "slide_type": "slide"
18
      }
19
     },
20
     "source": [
21
      "Robust Extraction of Quantitative Information from Histology Images"
22
     ]
23
    },
24
    {
25
     "cell_type": "markdown",
26
     "metadata": {},
27
     "source": [
28
      "**Quentin Caudron**"
29
     ]
30
    },
31
    {
32
     "cell_type": "markdown",
33
     "metadata": {
34
      "slideshow": {
35
       "slide_type": "slide"
36
      }
37
     },
38
     "source": [
39
      "<img src=\"figures/graphics/soay.jpg\" />"
40
     ]
41
    },
42
    {
43
     "cell_type": "markdown",
44
     "metadata": {
45
      "slideshow": {
46
       "slide_type": "subslide"
47
      }
48
     },
49
     "source": [
50
      "<img src=\"figures/population3.png\" width=1200px/>"
51
     ]
52
    },
53
    {
54
     "cell_type": "markdown",
55
     "metadata": {
56
      "slideshow": {
57
       "slide_type": "subslide"
58
      }
59
     },
60
     "source": [
61
      "<img src=\"figures/graphics/lit1.jpg\" />"
62
     ]
63
    },
64
    {
65
     "cell_type": "markdown",
66
     "metadata": {
67
      "slideshow": {
68
       "slide_type": "fragment"
69
      }
70
     },
71
     "source": [
72
      "<img src=\"figures/graphics/lit2.jpg\" />"
73
     ]
74
    },
75
    {
76
     "cell_type": "markdown",
77
     "metadata": {
78
      "slideshow": {
79
       "slide_type": "fragment"
80
      }
81
     },
82
     "source": [
83
      "<img src=\"figures/graphics/lit4.jpg\" />"
84
     ]
85
    },
86
    {
87
     "cell_type": "heading",
88
     "level": 2,
89
     "metadata": {
90
      "slideshow": {
91
       "slide_type": "slide"
92
      }
93
     },
94
     "source": [
95
      "Outline"
96
     ]
97
    },
98
    {
99
     "cell_type": "markdown",
100
     "metadata": {},
101
     "source": [
102
      "- Methods and data collection\n",
103
      "- Image processing\n",
104
      "- Extracted measures\n",
105
      "- Preliminary analysis\n",
106
      "- Future directions"
107
     ]
108
    },
109
    {
110
     "cell_type": "heading",
111
     "level": 2,
112
     "metadata": {
113
      "slideshow": {
114
       "slide_type": "slide"
115
      }
116
     },
117
     "source": [
118
      "Data"
119
     ]
120
    },
121
    {
122
     "cell_type": "markdown",
123
     "metadata": {},
124
     "source": [
125
      "**In the field, winter of 2011 - 2012 :**\n",
126
      "    \n",
127
      "- Daily study area monitoring for deaths\n",
128
      "- 143 liver samples collected within a day of death"
129
     ]
130
    },
131
    {
132
     "cell_type": "markdown",
133
     "metadata": {
134
      "slideshow": {
135
       "slide_type": "fragment"
136
      }
137
     },
138
     "source": [
139
      "<br />**In the lab :**\n",
140
      "\n",
141
      "- Sectioning after paraffin treatment\n",
142
      "- H&E staining of about 1000 slides"
143
     ]
144
    },
145
    {
146
     "cell_type": "markdown",
147
     "metadata": {
148
      "slideshow": {
149
       "slide_type": "fragment"
150
      }
151
     },
152
     "source": [
153
      "<br />**Analysis :**\n",
154
      "\n",
155
      "- Pathology standard : semi-quantitative scoring\n",
156
      "- Image processing"
157
     ]
158
    },
159
    {
160
     "cell_type": "heading",
161
     "level": 3,
162
     "metadata": {
163
      "slideshow": {
164
       "slide_type": "subslide"
165
      }
166
     },
167
     "source": [
168
      "The Field"
169
     ]
170
    },
171
    {
172
     "cell_type": "markdown",
173
     "metadata": {},
174
     "source": [
175
      "Sweat-and-blood-collected in cold, cold Scotland."
176
     ]
177
    },
178
    {
179
     "cell_type": "markdown",
180
     "metadata": {
181
      "slideshow": {
182
       "slide_type": "fragment"
183
      }
184
     },
185
     "source": [
186
      "Eight physical measurements :\n",
187
      "- Age at death\n",
188
      "- Weight\n",
189
      "- Sex\n",
190
      "- Limb length\n",
191
      "- Environmental \"stress\""
192
     ]
193
    },
194
    {
195
     "cell_type": "heading",
196
     "level": 3,
197
     "metadata": {
198
      "slideshow": {
199
       "slide_type": "subslide"
200
      }
201
     },
202
     "source": [
203
      "Clinical Pathology"
204
     ]
205
    },
206
    {
207
     "cell_type": "markdown",
208
     "metadata": {
209
      "slideshow": {
210
       "slide_type": "-"
211
      }
212
     },
213
     "source": [
214
      "Operator-driven visual analysis of 98 slides under microscopy."
215
     ]
216
    },
217
    {
218
     "cell_type": "markdown",
219
     "metadata": {
220
      "slideshow": {
221
       "slide_type": "fragment"
222
      }
223
     },
224
     "source": [
225
      "Eleven discrete and continuous measures :\n",
226
      "\n",
227
      "- Inflammation\n",
228
      "- Necrosis\n",
229
      "- Apoptosis\n",
230
      "- Hyperplasia\n",
231
      "- Fibrosis\n",
232
      "- Hepatitis"
233
     ]
234
    },
235
    {
236
     "cell_type": "heading",
237
     "level": 3,
238
     "metadata": {
239
      "slideshow": {
240
       "slide_type": "subslide"
241
      }
242
     },
243
     "source": [
244
      "Image Processing"
245
     ]
246
    },
247
    {
248
     "cell_type": "markdown",
249
     "metadata": {},
250
     "source": [
251
      "Automated analysis of 4430 images of slides representing 143 sheep."
252
     ]
253
    },
254
    {
255
     "cell_type": "markdown",
256
     "metadata": {
257
      "slideshow": {
258
       "slide_type": "fragment"
259
      }
260
     },
261
     "source": [
262
      "Seven structural and textural measures with varying levels of biological interpretation :\n",
263
      "\n",
264
      "- Inflammation\n",
265
      "- Hyperplasia / tissue density\n",
266
      "- Best-guess proxies for \"generic degeneration\""
267
     ]
268
    },
269
    {
270
     "cell_type": "heading",
271
     "level": 2,
272
     "metadata": {
273
      "slideshow": {
274
       "slide_type": "slide"
275
      }
276
     },
277
     "source": [
278
      "Image Processing"
279
     ]
280
    },
281
    {
282
     "cell_type": "markdown",
283
     "metadata": {
284
      "slideshow": {
285
       "slide_type": "subslide"
286
      }
287
     },
288
     "source": [
289
      "<img src=\"figures/graphics/sheep.jpg\"></img>"
290
     ]
291
    },
292
    {
293
     "cell_type": "markdown",
294
     "metadata": {
295
      "slideshow": {
296
       "slide_type": "subslide"
297
      }
298
     },
299
     "source": [
300
      "<img src=\"figures/graphics/processed.jpg\"></img>"
301
     ]
302
    },
303
    {
304
     "cell_type": "heading",
305
     "level": 3,
306
     "metadata": {
307
      "slideshow": {
308
       "slide_type": "subslide"
309
      }
310
     },
311
     "source": [
312
      "The Challenge"
313
     ]
314
    },
315
    {
316
     "cell_type": "markdown",
317
     "metadata": {
318
      "slideshow": {
319
       "slide_type": "-"
320
      }
321
     },
322
     "source": [
323
      "**Information extraction must be**\n",
324
      "- automagical - no operator input\n",
325
      "- reasonably quick - restricted computing time\n",
326
      "- robust - invariant to slicing, staining, field-related variation \n",
327
      "- unbiased - same algorithms for everyone"
328
     ]
329
    },
330
    {
331
     "cell_type": "markdown",
332
     "metadata": {
333
      "slideshow": {
334
       "slide_type": "subslide"
335
      }
336
     },
337
     "source": [
338
      "![image](figures/graphics/robust3.jpg)"
339
     ]
340
    },
341
    {
342
     "cell_type": "markdown",
343
     "metadata": {
344
      "slideshow": {
345
       "slide_type": "subslide"
346
      }
347
     },
348
     "source": [
349
      "![image](figures/graphics/robust4.jpg)"
350
     ]
351
    },
352
    {
353
     "cell_type": "markdown",
354
     "metadata": {
355
      "slideshow": {
356
       "slide_type": "subslide"
357
      }
358
     },
359
     "source": [
360
      "![image](figures/graphics/robust1.jpg)"
361
     ]
362
    },
363
    {
364
     "cell_type": "markdown",
365
     "metadata": {
366
      "slideshow": {
367
       "slide_type": "subslide"
368
      }
369
     },
370
     "source": [
371
      "![image](figures/graphics/robust2.jpg)"
372
     ]
373
    },
374
    {
375
     "cell_type": "heading",
376
     "level": 2,
377
     "metadata": {
378
      "slideshow": {
379
       "slide_type": "slide"
380
      }
381
     },
382
     "source": [
383
      "Structural and Textural Measures"
384
     ]
385
    },
386
    {
387
     "cell_type": "markdown",
388
     "metadata": {
389
      "slideshow": {
390
       "slide_type": "subslide"
391
      }
392
     },
393
     "source": [
394
      "- characteristic **scale** of sinusoid widths\n",
395
      "- **directional** amplitude of preferred sinusoid alignment\n",
396
      "- **tissue to sinusoid** ratio\n",
397
      "- **count** of inflammatory foci per image\n",
398
      "- **mean size** of inflammatory foci per image\n",
399
      "- information **entropy** of sinusoid distribution\n",
400
      "- **lacunarity** ( clustering ) of sinusoids"
401
     ]
402
    },
403
    {
404
     "cell_type": "markdown",
405
     "metadata": {
406
      "slideshow": {
407
       "slide_type": "subslide"
408
      }
409
     },
410
     "source": [
411
      "<img src=\"figures/graphics/gif.gif\"></img>"
412
     ]
413
    },
414
    {
415
     "cell_type": "markdown",
416
     "metadata": {
417
      "slideshow": {
418
       "slide_type": "subslide"
419
      }
420
     },
421
     "source": [
422
      "<img src=\"figures/graphics/intra3.png\" width=100%/>"
423
     ]
424
    },
425
    {
426
     "cell_type": "markdown",
427
     "metadata": {
428
      "slideshow": {
429
       "slide_type": "subslide"
430
      }
431
     },
432
     "source": [
433
      "<img src=\"figures/graphics/inter3.png\"/>"
434
     ]
435
    },
436
    {
437
     "cell_type": "heading",
438
     "level": 2,
439
     "metadata": {
440
      "slideshow": {
441
       "slide_type": "slide"
442
      }
443
     },
444
     "source": [
445
      "Exploratory Analysis"
446
     ]
447
    },
448
    {
449
     "cell_type": "markdown",
450
     "metadata": {},
451
     "source": [
452
      "by individual"
453
     ]
454
    },
455
    {
456
     "cell_type": "markdown",
457
     "metadata": {
458
      "slideshow": {
459
       "slide_type": "subslide"
460
      }
461
     },
462
     "source": [
463
      "<img src=\"figures/regressions/BDHyperplasia/lm-0.png\" />"
464
     ]
465
    },
466
    {
467
     "cell_type": "markdown",
468
     "metadata": {
469
      "slideshow": {
470
       "slide_type": "subslide"
471
      }
472
     },
473
     "source": [
474
      "<img src=\"figures/regressions/PortalInflammation/lm-0.png\" />"
475
     ]
476
    },
477
    {
478
     "cell_type": "markdown",
479
     "metadata": {
480
      "slideshow": {
481
       "slide_type": "subslide"
482
      }
483
     },
484
     "source": [
485
      "<img src=\"figures/regressions/PortalInflammation/lm-1.png\" />"
486
     ]
487
    },
488
    {
489
     "cell_type": "heading",
490
     "level": 2,
491
     "metadata": {
492
      "slideshow": {
493
       "slide_type": "slide"
494
      }
495
     },
496
     "source": [
497
      "Exploratory Analysis"
498
     ]
499
    },
500
    {
501
     "cell_type": "markdown",
502
     "metadata": {},
503
     "source": [
504
      "controlled for age / cohort"
505
     ]
506
    },
507
    {
508
     "cell_type": "markdown",
509
     "metadata": {
510
      "slideshow": {
511
       "slide_type": "subslide"
512
      }
513
     },
514
     "source": [
515
      "<img src=\"figures/regressions/PortalInflammation/mm_0.png\" />"
516
     ]
517
    },
518
    {
519
     "cell_type": "markdown",
520
     "metadata": {
521
      "slideshow": {
522
       "slide_type": "subslide"
523
      }
524
     },
525
     "source": [
526
      "<img src=\"figures/regressions/BDHyperplasia/mm_0.png\" />"
527
     ]
528
    },
529
    {
530
     "cell_type": "markdown",
531
     "metadata": {
532
      "slideshow": {
533
       "slide_type": "subslide"
534
      }
535
     },
536
     "source": [
537
      "<img src=\"figures/regressions/BDHyperplasia/mm_1.png\" />"
538
     ]
539
    },
540
    {
541
     "cell_type": "markdown",
542
     "metadata": {
543
      "slideshow": {
544
       "slide_type": "subslide"
545
      }
546
     },
547
     "source": [
548
      "<img src=\"figures/regressions/TawfikTotal/mm_0.png\" />"
549
     ]
550
    },
551
    {
552
     "cell_type": "markdown",
553
     "metadata": {
554
      "slideshow": {
555
       "slide_type": "subslide"
556
      }
557
     },
558
     "source": [
559
      "<img src=\"figures/regressions/Fibrosis/mm_0.png\" />"
560
     ]
561
    },
562
    {
563
     "cell_type": "markdown",
564
     "metadata": {
565
      "slideshow": {
566
       "slide_type": "subslide"
567
      }
568
     },
569
     "source": [
570
      "<img src=\"figures/regressions/Hindleg/mm_0.png\" />"
571
     ]
572
    },
573
    {
574
     "cell_type": "markdown",
575
     "metadata": {
576
      "slideshow": {
577
       "slide_type": "subslide"
578
      }
579
     },
580
     "source": [
581
      "<img src=\"figures/regressions/Weight/mm_0.png\" />"
582
     ]
583
    },
584
    {
585
     "cell_type": "heading",
586
     "level": 2,
587
     "metadata": {
588
      "slideshow": {
589
       "slide_type": "slide"
590
      }
591
     },
592
     "source": [
593
      "Further analysis"
594
     ]
595
    },
596
    {
597
     "cell_type": "markdown",
598
     "metadata": {},
599
     "source": [
600
      "Age or cohort effect ?"
601
     ]
602
    },
603
    {
604
     "cell_type": "markdown",
605
     "metadata": {
606
      "slideshow": {
607
       "slide_type": "subslide"
608
      }
609
     },
610
     "source": [
611
      "<img src=\"figures/regressions/BDHyperplasia/new/mm_coefs_color_E.png\" width=85%/>"
612
     ]
613
    },
614
    {
615
     "cell_type": "markdown",
616
     "metadata": {
617
      "slideshow": {
618
       "slide_type": "subslide"
619
      }
620
     },
621
     "source": [
622
      "<img src=\"figures/regressions/BDHyperplasia/new/mm_coefs_color_CES.png\" width=85%/>"
623
     ]
624
    },
625
    {
626
     "cell_type": "markdown",
627
     "metadata": {
628
      "slideshow": {
629
       "slide_type": "subslide"
630
      }
631
     },
632
     "source": [
633
      "<img src=\"figures/regressions/BDHyperplasia/new/mm_coefs_color_RES.png\" width=85%/>"
634
     ]
635
    },
636
    {
637
     "cell_type": "heading",
638
     "level": 2,
639
     "metadata": {
640
      "slideshow": {
641
       "slide_type": "slide"
642
      }
643
     },
644
     "source": [
645
      "Conclusions"
646
     ]
647
    },
648
    {
649
     "cell_type": "markdown",
650
     "metadata": {},
651
     "source": [
652
      "- our image measures capture **relevant** and **useful** information\n",
653
      "- a number of correlations can be **explained** biologically\n",
654
      "- underlying **structure** in the data needs thought\n",
655
      "- still no **map** from image or histological measures to condition of individual"
656
     ]
657
    },
658
    {
659
     "cell_type": "heading",
660
     "level": 2,
661
     "metadata": {
662
      "slideshow": {
663
       "slide_type": "slide"
664
      }
665
     },
666
     "source": [
667
      "Future directions"
668
     ]
669
    },
670
    {
671
     "cell_type": "heading",
672
     "level": 3,
673
     "metadata": {
674
      "slideshow": {
675
       "slide_type": "subslide"
676
      }
677
     },
678
     "source": [
679
      "Further exploration of the dataset"
680
     ]
681
    },
682
    {
683
     "cell_type": "markdown",
684
     "metadata": {},
685
     "source": [
686
      "- 145 sheep ( 89 females )\n",
687
      "- 12 age classes\n",
688
      "- potential redundancy in various measures"
689
     ]
690
    },
691
    {
692
     "cell_type": "markdown",
693
     "metadata": {},
694
     "source": [
695
      "- 4460 entries across 31 variables\n",
696
      "- 3596 with full image and histological information\n",
697
      "- 1196 for which **complete** information is available"
698
     ]
699
    },
700
    {
701
     "cell_type": "heading",
702
     "level": 3,
703
     "metadata": {
704
      "slideshow": {
705
       "slide_type": "subslide"
706
      }
707
     },
708
     "source": [
709
      "More data"
710
     ]
711
    },
712
    {
713
     "cell_type": "markdown",
714
     "metadata": {},
715
     "source": [
716
      "- nutritional information\n",
717
      "- immunity data"
718
     ]
719
    },
720
    {
721
     "cell_type": "heading",
722
     "level": 3,
723
     "metadata": {
724
      "slideshow": {
725
       "slide_type": "subslide"
726
      }
727
     },
728
     "source": [
729
      "Narrow-field images"
730
     ]
731
    },
732
    {
733
     "cell_type": "markdown",
734
     "metadata": {},
735
     "source": [
736
      "- 12536 images\n",
737
      "- spatial distribution of nuclei"
738
     ]
739
    },
740
    {
741
     "cell_type": "markdown",
742
     "metadata": {
743
      "slideshow": {
744
       "slide_type": "subslide"
745
      }
746
     },
747
     "source": [
748
      "![image](figures/graphics/10.jpg)"
749
     ]
750
    },
751
    {
752
     "cell_type": "markdown",
753
     "metadata": {
754
      "slideshow": {
755
       "slide_type": "subslide"
756
      }
757
     },
758
     "source": [
759
      "![image](figures/graphics/Processed2.jpg)"
760
     ]
761
    },
762
    {
763
     "cell_type": "markdown",
764
     "metadata": {
765
      "slideshow": {
766
       "slide_type": "subslide"
767
      }
768
     },
769
     "source": [
770
      "![image](figures/graphics/Segmented.jpg)"
771
     ]
772
    },
773
    {
774
     "cell_type": "markdown",
775
     "metadata": {
776
      "slideshow": {
777
       "slide_type": "subslide"
778
      }
779
     },
780
     "source": [
781
      "<img src=\"figures/graphics/10x.png\" width=100%></src>"
782
     ]
783
    },
784
    {
785
     "cell_type": "heading",
786
     "level": 2,
787
     "metadata": {
788
      "slideshow": {
789
       "slide_type": "slide"
790
      }
791
     },
792
     "source": [
793
      "With thanks to"
794
     ]
795
    },
796
    {
797
     "cell_type": "markdown",
798
     "metadata": {},
799
     "source": [
800
      "Romain Garnier\n",
801
      "\n",
802
      "Andrea Graham\n",
803
      "\n",
804
      "Tawfik Aboellail (CSU)\n",
805
      "\n",
806
      "Bryan Grenfell\n"
807
     ]
808
    }
809
   ],
810
   "metadata": {}
811
  }
812
 ]
813
}