forked from ME-ICA/tedana-ohbm-2020
-
Notifications
You must be signed in to change notification settings - Fork 0
/
OHBM2020_1083_tedana_poster_narration.srt
167 lines (126 loc) · 5.24 KB
/
OHBM2020_1083_tedana_poster_narration.srt
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
1
00:00:00,000 --> 00:00:05,000
Welcome to poster 1083 “tedana multi-echo software and communal resources.
2
00:00:05,100 --> 00:00:11,000
I’m Dan Handwerker and I’m representing the growing tedana community. Tedana has two main parts.
3
00:00:11,100 --> 00:00:17,000
First, it’s an open source software that is designed to process and denoise multi-echo fMRI data.
4
00:00:17,100 --> 00:00:23,000
Second, it’s a community and resources for people interested in multi-echo fMRI whether or not they use tedana.
5
00:00:23,100 --> 00:00:28,000
The software is based on a multi-echo ICA approach that was initially developed by Prantik Kundu.
6
00:00:28,100 --> 00:00:34,000
We have attempted to make that method more modular and understandable so members of our community, perhaps including you,
7
00:00:34,100 --> 00:00:38,500
can better validate existing methods and develop better multi-echo fMRI tools.
8
00:00:38,600 --> 00:00:47,000
A key addition in the past year has been an interactive reports page where users can examine each component that was accepted or rejected to better understand why.
9
00:00:47,100 --> 00:00:53,000
The poster includes a static version of this report and here I am going to demonstrate the dynamic version
10
00:00:53,100 --> 00:00:58,000
What you see up here is a pie chart of variance explained of each ICA component.
11
00:00:58,100 --> 00:01:03,000
For each of these components you can see as I hover over it, it tells us some information about the component
12
00:00:58,100 --> 00:01:10,000
We can see the same data as a scatter plot of kappa, which is a rough T2* weighting
13
00:01:10,100 --> 00:01:18,000
versus rho, which is a rough S0 measure. Higher kappa and lower rho is typically considered signal that we want to keep.
14
00:01:18,100 --> 00:01:23,000
and these are again sorted by rho values and kappa values.
15
00:01:23,100 --> 00:01:27,000
What's dynamic about this report is that not only do you see this, I can click on one of these and you can see
16
00:01:27,100 --> 00:01:31,000
this is a high variance rejected component. It's the linear drift of our signal.
17
00:01:31,100 --> 00:01:36,000
This is the time series, the spatial map, and the frequency. So we can click on other rejected components
18
00:01:36,100 --> 00:01:41,000
to get a sense of what's being rejected, and also click on accepted components to see what's being accepted
19
00:01:41,100 --> 00:01:46,000
for example this is our highest kappa accepted component, and this looks like the response
20
00:01:46,100 --> 00:01:52,000
from a task design with high visual cortex activity, and this is the frequency of that signal
21
00:01:52,100 --> 00:02:00,000
so this is a very nice improvement because, when we get these results sometimes it's a little bit opaque of what's happening
22
00:02:00,100 --> 00:02:04,000
so this is one more way of understanding what's happening, and better understand both
23
00:02:04,100 --> 00:02:08,000
your own data, and, if you're interested, better try to figure out what
24
00:02:08,100 --> 00:02:11,000
is or isn't working to both help yourself and us improve the method.
25
00:02:11,100 --> 00:02:18,000
We do recognize that not everyone who is working with multi-echo fMRI can or wants to use tedana as a software.
26
00:02:18,100 --> 00:02:24,000
In response to discussions at the last OHBM, we’ve greatly expanded the documentation on the tedana website
27
00:02:24,100 --> 00:02:27,000
to include resources that may benefit the wider community.
28
00:02:27,100 --> 00:02:33,000
In our documenation, you can see that we have the tedana pipeline, so we have all the stuff that's software specific
29
00:02:33,100 --> 00:02:38,000
We have the pipeline. We have the API, which goes through every function.
30
00:02:38,100 --> 00:02:41,000
But we've also added a lot of text on things like "What is multi-echo fMRI"
31
00:02:41,100 --> 00:02:45,000
basic physics of it with links to things where you can read more
32
00:02:45,100 --> 00:02:50,000
"Why would you want to use multi-echo", Considerations for when you're going to use it.
33
00:02:50,100 --> 00:02:56,000
that might influence your decision. We've added resources of the different vendors
34
00:02:56,100 --> 00:03:03,000
that includes sequences you can get for each vendor. How to think about what parameters to use.
35
00:03:03,100 --> 00:03:06,000
"what parameters have been used in publications"
36
00:03:06,100 --> 00:03:13,000
and finally, we have this section on good starting point articles, and videos, and lessons
37
00:03:13,100 --> 00:03:18,000
if you want to learn more about multi-echo fMRI. Other people's software that uses multi-echo fMRI
38
00:03:18,100 --> 00:03:22,000
and open data sets we're starting to collect and
39
00:03:22,100 --> 00:03:26,000
and we've been trying to maintain a list of all publications that use multi-echo fMRI
40
00:03:26,100 --> 00:03:32,000
If you use multi-echo fMRI, whether or not you use tedana software, we welcome contributions.
41
00:03:32,100 --> 00:03:38,000
Particularly, if you have published a study with multi-echo fMRI, make sure we have it on our list of all multi-echo fMRI publications.
42
00:03:38,100 --> 00:03:44,000
Thank you for taking the time to visit our poster and we welcome questions and contributions.