quality-assessment-protocol/qap/viz/html/qap_functional_temporal.html at 26648b3cb08245338f6fad84d32774d86cdc3289 · preprocessed-connectomes-project/quality-assessment-protocol · GitHub

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
<!DOCTYPE html>
<html xmlns="http://www.w3.org/1999/xhtml">

<head>
    <title>PCP Quality Assessment Protocol</title>
	<meta http-equiv="Content-Type" content="text/html; charset=utf-8" />
</head>

<body>
<!-- HEADER -->
<div id="header">
    <h1 id="project_title">PCP Quality Assessment Protocol - QAP Functional Temporal Reports</h1>
</div>

<p>After all processing has been completed, the designated output directory for the <code>qap_functional_temporal.py</code> workflow will contain a set of pdf files that contain the relevant reports for the set of scans undergoing quality assessment. The set of output pdfs includes one pdf file per input scan, e.g.: <code>qap_functional_temporal_sub-01.pdf</code>, which contains the tSNR maps, FD plot and QC metrics for that scan. There will also be a group report pdf in that main output directory, e.g.: <code>qap_functional_temporal_group.pdf</code>, that contains summary metrics for the entire set of scans.</p>

<h2> For the individual scan reports:</h2>

<h3 id="mosaic">The tSNR Slice Mosaic</h3>

<p>The image showing the tSNR map for a particular scan, e.g.:</p>
<pre><code>tSNR volume, subject sub-01 (session_1_func_1)</code></pre>
<p>is the rendering of the temporal signal-to-noise ratio information from the scan.
The tSNR plot is similar to the mean EPI plot generated by the <code>qap_functional_spatial.py</code> workflow, in that both metrics reduce the 4-dimensional
scan to a representative 3-dimensional volume that is then split and displayed as a stack of
axial (horizontal) slices.
For this tSNR plot, the mean of each voxel's timeseries is
also computed and is then divided by the timeseries standard deviation.
Hence, the tSNR plot shows the voxels in which one would expect to have SNR good enough for statistical analyses.
Differences in tSNR are particularly important for comparing the results from region of
interest (ROI) analyses, since any observed functional differences might actually be attributable
to systematic differences in SNR across the regions being compared.
You can learn more about the utility of tSNR plots for fMRI analyses
<a href="http://practicalfmri.blogspot.com.es/2011/01/comparing-fmri-protocols.html" target="_blank" >here</a>.</p>

<h3 id="fd_plot">The framewise displacement plot</h3>

<p>The image showing the framewise displacement plot for the scan being assessed, e.g.:</p>
<pre><code>FD, subject sub-01 (session_1_func_1)</code></pre>
<p>is the framewise displacement that occurred throughout the scan.
This is a temporal motion quality assurance metric and tracks head motions over time, making
it easy to determine whether or not the data potentially suffered from significant corruption
due to motion. For instance, it is possible to detect if the participant's head was slowly sinking into the
cushions in the head coil, or whether the participant was possibly restless or agitated, which
would result in several position changes or movement spikes. The framewise displacement is a
frame-by-frame representation of the differences between the BOLD signal intensity value of
the n and n+1 timepoints, the n+1 and n+2 timepoints, and so on. The report page for framewise
displacement in the functional scan includes both this frame-by-frame plot, as well as a
histogram that can be used to visually determine what proportion of timepoints exceeded some
pre-set movement threshold (eg: 0.2 mm).</p>

<h3 id="metrics">The Temporal Metrics computed on the Functional Scan</h3>

<p>The metrics displayed in the Summary Report were computed using the <code>qap_functional_temporal.py</code> workflow and have been displayed as violin plots. Eg:</p>
<pre><code>QC measures (subject sub-01_session_1)</code></pre>
<p>The stars in these plots denote where the score for this particular scan falls in the
distribution of all scores for scans that were included as inputs to the the
functional-temporal workflow.</p>

<p>The metrics computed are as follows:</p>
<ol>
	<li>DVARS - Spatial standard deviation of the voxelwise temporal derivates </li>
	<li>GCOR - Global Correlation</li>
	<li>Mean RMSD - Mean Fractional Displacement</li>
	<li>Outliers - Mean fraction of outliers per fMRI volume</li>
	<li>Quality - Median Distance Index</li>
</ol>

<p>All metrics are described in more detail in the
<a href="http://preprocessed-connectomes-project.github.io/quality-assessment-protocol/#taxonomy-of-qa-measures" target="_blank" >Taxonomy of QA Measures section</a> of the QAP documentation. Please refer to the QAP website for descriptions of these metrics.</p>


<h2>For the group reports:</h2>

<p>The violin plots included in the group report, e.g.: <code>QC measures (session_1)</code>
are a graphical representation of the columnar values in the <code>qap_functional_temporal.csv</code>
file that was created in the main output directory for the workflow.
The scores for each metric described above were aggregated to create the distributions that were
plotted in both the individual and group reports. Hence, the violin plots in the individual scan
reports and the group reports are identical, except that the group reports do not contain any
stars denoting individual scans. These group reports are intended to provide the user a means of
visually inspecting the overall quality of the temporal data for that group of functional scans.</p>

</body>