-
Notifications
You must be signed in to change notification settings - Fork 0
/
harmonisation.html
334 lines (301 loc) · 21.6 KB
/
harmonisation.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
<!DOCTYPE HTML>
<html lang="en">
<head>
<meta http-equiv="Content-Type" content="text/html; charset=UTF-8">
<title>Nicola K Dinsdale</title>
<meta name="author" content="Nicola K Dinsdale">
<meta name="viewport" content="width=device-width, initial-scale=1">
<link rel="stylesheet" type="text/css" href="stylesheet.css">
</head>
<body>
<table
style="width:100%;max-width:800px;border:0px;border-spacing:0px;border-collapse:separate;margin-right:auto;margin-left:auto;">
<tbody>
<tr style="padding:0px">
<td style="padding:0px">
<table
style="width:100%;border:0px;border-spacing:0px;border-collapse:separate;margin-right:auto;margin-left:auto;">
<tbody>
<tr style="padding:0px">
<td style="padding:2.5%;width:63%;vertical-align:middle">
<p style="text-align:center">
<name>Harmonisation and Domain Adaptation</name>
</p>
<p> Different MRI scanners produce images with different characteristics which leads
to an increase in non-biological noise when the images are pooled. This is known
as the harmonisation problem. We propose to model harmonisation as a domain
adaptation problem, and have proposed a number of methods to enable models to be
trained with data from a range of scanners.
</p>
<p style="text-align:center">
<a href="index.html">Home</a>  / 
<a
href="harmonisation.html">Harmonisation</a>
 / 
<a href="segmentation.html">Segmentation</a>  / 
<a href="translation.html">Translation</a>  / 
<a href="privacy.html">Privacy</a>  / 
<a href="explainable.html">Explainable AI</a>
</p>
</td>
<td style="padding:2.5%;width:40%;max-width:40%">
<a href="images/mygif.gif"><img style="width:150%;max-width:150%"
alt="profile photo" src="images/mygif.gif" class="hoverZoomLink"></a>
</td>
</tr>
</tbody>
</table>
<table
style="width:100%;border:0px;border-spacing:0px;border-collapse:separate;margin-right:auto;margin-left:auto;">
<tbody>
<tr>
<td style="padding:20px;width:100%;vertical-align:middle">
<p style="text-align:center">
<heading>Papers</heading>
</td>
</tr>
</tbody>
</table>
<table
style="width:100%;border:0px;border-spacing:0px;border-collapse:separate;margin-right:auto;margin-left:auto;">
<tbody>
<tr>
<td style="padding:20px;width:25%;vertical-align:middle">
<img src='images/unifed.png' width="160" height="150">
</td>
<td style="padding:20px;width:75%;vertical-align:middle">
<a href="https://www.biorxiv.org/content/10.1101/2024.02.05.578912v1.full.pdf">
<papertitle>UniFed: A unified deep learning framework for
segmentation of partially labelled, distributed
neuroimaging data </papertitle>
</a>
<br>
<strong> Nicola K Dinsdale</strong>, <a
href="https://www.ndcn.ox.ac.uk/team/mark-jenkinson">Mark Jenkinson</a>, <a
href="https://www.pmb.ox.ac.uk/person/dr-ana-namburete">Ana IL Namburete</a>
<br>
<em>bioRxiv</em>, 2024
<br>
<a href="https://nkdinsdale.github.io/unifed_proj/">Project Page</a> / <a
href="https://www.biorxiv.org/content/10.1101/2024.02.05.578912v1.full.pdf">Paper</a> / <a
href="https://github.com/nkdinsdale/UniFed">Code</a>
<p></p>
<p> We, therefore, propose UniFed, a unified federated harmoni-
sation framework, which enables three key processes to be completed: 1) the training of a federated partially labelled
harmonisation network, 2) the selection of the most appropriate pretrained model for a new unseen
site, and 3) the incorporation of a new site into the harmonised federation. </p>
</td>
</tr>
<tr>
<td style="padding:20px;width:25%;vertical-align:middle">
<img src='images/bhatt.png' width="160" height="150">
</td>
<td style="padding:20px;width:75%;vertical-align:middle">
<a href="https://arxiv.org/pdf/2303.15965.pdf">
<papertitle>SFHarmony: Source Free Domain Adaptation for Distributed
Neuroimaging Analysis </papertitle>
</a>
<br>
<strong> Nicola K Dinsdale </strong>, <a
href="https://www.ndcn.ox.ac.uk/team/mark-jenkinson">Mark Jenkinson</a>, <a
href="https://www.pmb.ox.ac.uk/person/dr-ana-namburete">Ana IL Namburete</a>
<br>
<em>ICCV</em>, 2023
<br>
<a href="https://nkdinsdale.github.io/sfharmony_proj/">Project Page</a> / <a
href="https://arxiv.org/pdf/2303.15965.pdf">Paper</a> / <a
href="https://github.com/nkdinsdale/SFHarmony">Code</a>
<p></p>
<p> We propose an Unsupervised Source-Free Domain Adaptation (SFDA) method,
SFHarmony, and demonstrate the approach for classification, regression and
segmentation tasks.</p>
</td>
</tr>
<tr>
<td style="padding:20px;width:25%;vertical-align:middle">
<img src='images/federated.png' width="160" height="150">
</td>
<td style="padding:20px;width:75%;vertical-align:middle">
<a href="https://arxiv.org/abs/2205.15970">
<papertitle>FedHarmony: Unlearning Scanner Bias with Distributed Data
</papertitle>
</a>
<br>
<strong> Nicola K Dinsdale </strong>, <a
href="https://www.ndcn.ox.ac.uk/team/mark-jenkinson">Mark Jenkinson</a>, <a
href="https://www.pmb.ox.ac.uk/person/dr-ana-namburete">Ana IL Namburete</a>
<br>
<em>MICCAI 2022</em>, 2022 <span style="color:red;">[Early Acceptance]</span>
<br>
<a href="https://nkdinsdale.github.io/fedharmony_proj/">Project Page</a> / <a
href="https://arxiv.org/abs/2205.15970">Paper</a> / <a
href="https://github.com/nkdinsdale/FedHarmony">Code</a>
<p></p>
<p> Adapted my harmonisation framework to work in a distributed setting, through
modelling site features as Gaussian distributions.</p>
</td>
</tr>
<tr>
<td style="padding:20px;width:25%;vertical-align:middle">
<img src='images/wmh.png' width="160" height="150">
</td>
<td style="padding:20px;width:75%;vertical-align:middle">
<a href="https://ieeexplore.ieee.org/abstract/document/9761539">
<papertitle>Omni-Supervised Domain Adversarial Training for White Matter
Hyperintensity Segmentation in the UK Biobank</papertitle>
</a> <br>
<a href="https://www.ndcn.ox.ac.uk/team/vaanathi-sundaresan">Vaanathi
Sundaresan,</a> <strong> Nicola K Dinsdale</strong>, Ludovica Griffanti, <a
href="https://www.ndcn.ox.ac.uk/team/mark-jenkinson">Mark Jenkinson</a>,
<br>
<em>ISBI 2022 </em> <span style="color:red;">[Oral Presentation]</span>
<br>
<a href="https://nkdinsdale.github.io/isbi_proj/">Project Page</a> / <a
href="https://ieeexplore.ieee.org/abstract/document/9761539">Paper</a> / <a
href="https://github.com/v-sundaresan/omnisup_agepred_semidann">Code</a>
<p></p>
<p> Exploring the use of omni-supervised learning for white matter hyperintensity
segmentation, using age prediction as the selection criteria. Leads to
significant increase in sensitivity especially for small lesions. </p>
</td>
</tr>
<tr>
<td style="padding:20px;width:25%;vertical-align:middle">
<img src='images/harm.png' width="160" height="150">
</td>
<td style="padding:20px;width:75%;vertical-align:middle">
<a
href="https://www.sciencedirect.com/science/article/pii/S1053811920311745?via%3Dihub">
<papertitle>Deep learning-based unlearning of dataset bias for MRI harmonisation
and confound removal</papertitle>
</a>
<br>
<strong> Nicola K Dinsdale </strong>, <a
href="https://www.ndcn.ox.ac.uk/team/mark-jenkinson">Mark Jenkinson</a>, <a
href="https://www.pmb.ox.ac.uk/person/dr-ana-namburete">Ana IL Namburete</a>
<br>
<em>Neuroimage</em>, 2021
<br>
<a href="https://nkdinsdale.github.io/harmproj/">Project Page</a> / <a
href="https://www.sciencedirect.com/science/article/pii/S1053811920311745?via%3Dihub">Paper</a>
/ <a href="https://github.com/nkdinsdale/Unlearning_for_MRI_harmonisation">Code</a>
<p></p>
<p> IDP harmonisation using iterative unlearning framework, applied across tasks and
architectures.</p>
<p> Parts of this work were presented at <a
href="https://link.springer.com/chapter/10.1007/978-3-030-52791-4_2">MIUA
2020</a> and <a
href="https://www.springerprofessional.de/en/unlearning-scanner-bias-for-mri-harmonisation/18442854">MICCAI
2020</a> <span style="color:red;">[Early Acceptance]</span>. </p>
</td>
</tr>
<tr>
<td style="padding:20px;width:25%;vertical-align:middle">
<img src='images/wmh.png' width="160" height="150">
</td>
<td style="padding:20px;width:75%;vertical-align:middle">
<a href="https://doi.org/10.1016/j.media.2021.102215">
<papertitle> Comparison of domain adaptation techniques for white matter
hyperintensity segmentation in brain MR images </papertitle>
</a>
<br>
<a href="https://www.ndcn.ox.ac.uk/team/vaanathi-sundaresan">Vaanathi
Sundaresan</a>, Giovanna Zamboni, <strong> Nicola K Dinsdale </strong>, Peter
Rothwell, Ludovica Griffanti, <a
href="https://www.ndcn.ox.ac.uk/team/mark-jenkinson">Mark Jenkinson</a>
<br>
<em> Medical Image Analysis </em> 2021
<br>
<a href="https://doi.org/10.1016/j.media.2021.102215">Paper</a> / <a
href="https://git.fmrib.ox.ac.uk/vaanathi/true_net_wmh_segmentation_pytorch">Code</a>
<p></p>
<p> Comparison of DA methods for white matter hyperintensity segmentation, comparing
methods including my proposed method: <a
href="https://nkdinsdale.github.io/harmproj/">paper</a>.</p>
</td>
</tr>
<tr>
<td style="padding:20px;width:25%;vertical-align:middle">
<img src='images/cardiac.png' width="160" height="150">
</td>
<td style="padding:20px;width:75%;vertical-align:middle">
<a href="https://link.springer.com/chapter/10.1007/978-3-030-68107-4_20">
<papertitle> A 2-step deep learning method with domain adaptation for
multi-centre, multi-vendor and multi-disease cardiac magnetic resonance
segmentation </papertitle>
</a>
<br>
Jorge Corral Acero, <a
href="https://www.ndcn.ox.ac.uk/team/vaanathi-sundaresan">Vaanathi
Sundaresan</a>, <strong> Nicola K Dinsdale </strong>, Vicente Grau, <a
href="https://www.ndcn.ox.ac.uk/team/mark-jenkinson">Mark Jenkinson</a>
<br>
<em> STACOM 2020 - 6th place </em>
<br>
<a href="https://link.springer.com/chapter/10.1007/978-3-030-68107-4_20">Paper</a>
<p></p>
<p> Segmentation of cardiac MRI across sites, using the method proposed in my <a
href="https://nkdinsdale.github.io/harmproj/">paper</a></p>
</td>
</tr>
<tr>
<td style="padding:20px;width:25%;vertical-align:middle">
<img src='images/harm.png' width="160" height="150">
</td>
<td style="padding:20px;width:75%;vertical-align:middle">
<a href="https://link.springer.com/chapter/10.1007/978-3-030-59713-9_36">
<papertitle> Unlearning Scanner Bias for MRI Harmonisation </papertitle>
</a>
<br>
<strong> Nicola K Dinsdale </strong>, <a
href="https://www.ndcn.ox.ac.uk/team/mark-jenkinson">Mark Jenkinson</a>, <a
href="https://www.pmb.ox.ac.uk/person/dr-ana-namburete">Ana IL Namburete</a>
<br>
<em>MICCAI</em>, 2020 <span style="color:red;">[Early Acceptance]</span>
<br>
<a href="https://link.springer.com/chapter/10.1007/978-3-030-59713-9_36">Paper</a> /
<a href="https://github.com/nkdinsdale/Unlearning_for_MRI_harmonisation">Code</a>
<p></p>
<p> MRI harmonisation using iterative unlearning framework, explored for age
prediction.</p>
</td>
</tr>
<tr>
<td style="padding:20px;width:25%;vertical-align:middle">
<img src='images/seg.png' width="160" height="150">
</td>
<td style="padding:20px;width:75%;vertical-align:middle">
<a
href="https://ora.ox.ac.uk/objects/uuid:1ffcbaa8-1f5c-4166-8e86-c3618c895e5f/download_file?safe_filename=Unlearning_scanner_bias_for_MRI_harmonisation_in_medical_image_segmentation.pdf&type_of_work=Conference+item">
<papertitle> Unlearning Scanner Bias for MRI Harmonisation in Medical Image
Segmentation </papertitle>
</a>
<br>
<strong> Nicola K Dinsdale </strong>, <a
href="https://www.ndcn.ox.ac.uk/team/mark-jenkinson">Mark Jenkinson</a>, <a
href="https://www.pmb.ox.ac.uk/person/dr-ana-namburete">Ana IL Namburete</a>
<br>
<em>MIUA</em>, 2020
<br>
<a
href="https://ora.ox.ac.uk/objects/uuid:1ffcbaa8-1f5c-4166-8e86-c3618c895e5f/download_file?safe_filename=Unlearning_scanner_bias_for_MRI_harmonisation_in_medical_image_segmentation.pdf&type_of_work=Conference+item">Paper</a>
/ <a href="https://github.com/nkdinsdale/Unlearning_for_MRI_harmonisation">Code</a>
<p></p>
<p> MRI harmonisation using iterative unlearning framework, explored for
segmentation tasks.</p>
</td>
</tr>
</tbody>
</table>
<br>
<p style="text-align:right;font-size:small;">
The template of this webpage is from <a
href="https://github.com/jonbarron/jonbarron_website">source
code</a>.
</p>
</td>
</tr>
</tbody>
</table>
</body>
</html>