/
s_fe_weights_150.m
197 lines (168 loc) · 8.05 KB
/
s_fe_weights_150.m
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
function s_fe_weights_150(bval,tractographyType)
%
% This function:
% - Load a series of results from precomputed connectomes
% - Extract the weights and fiber lenths.
% - Make a plot of the distribution of weights and fiber lengths.
%
% s_fe_weights(bval,tractographyType)
%
% Copyright Franco Pestilli (2014) Stanford University
% Get the base directory for the data
datapath = '/marcovaldo/frk/2t2/predator/';
subjects = {'FP_150dirs_b1000_2000_4000_2iso'};
if notDefined('tractographyType'), tractographyType = 'lmax10'; end
if notDefined('bval'), bval = []; end
if notDefined('saveDir'), saveDir = fullfile('/marcovaldo/frk/Dropbox','pestilli_etal_revision',mfilename);end
% These are default parameters for plotting
nBins = 11; % first bin is zero weight second is
bins_w = linspace(-10,-1,nBins);
proportionDeleted = nan(size(subjects,2),2);
w = cell(1,length(subjects));
len = w;
for isbj = 1:length(subjects)
% Directory where to load the results
resultsPath = fullfile(datapath,subjects{isbj},'results');
savedir = fullfile(saveDir,subjects{isbj});
if ~isempty(bval)
resFiles = dir(fullfile(resultsPath,sprintf('*%s*%s*.mat',num2str(bval),tractographyType)));
else
resFiles = dir(fullfile(resultsPath,sprintf('*%s*.mat',tractographyType)));
end
% We build one modelper fiber group, whole brain fiber group
for iRes = 1:length(resFiles)
% The final connectome and dat astructure will be saved with this name:
[~,resFileName] = fileparts(resFiles(iRes).name);
% Buil a full-file of the fibers and the RESULTS structure to load
resFileName2Load = fullfile(resultsPath,[resFileName,'.mat']);
% Initialize the Connectome
fprintf('[%s] Loading results: \n%s\n',mfilename,resFileName2Load)
load(resFileName2Load);
% Reorganize the weights and fiber-lengths in structures divided by
% subjects
w{isbj} = results.weights(results.weights > 0);
olen{isbj} = results.optimized.length;
clen{isbj} = results.candidate.length;
% Distribution of length
[f, ol(isbj,:), cl(isbj,:), lbins] = plotLengthHist(tractographyType,olen{isbj},clen{isbj});
feSavefig(f,'verbose','yes','figName',[resFileName, '_optimized_fiberlength_hist'],'figDir',savedir,'figType','eps');
close(f)
% Proportion of deleted fibers at each fiber length
ratioLength = ol./cl;
drl = mean(ratioLength,1);
f = plotLengthRatioHist(tractographyType,drl,lbins);
feSavefig(f,'verbose','yes','figName',[resFileName, '_length_ratio_hist'],'figDir',savedir,'figType','eps');
close(f)
% Distribution of weights
[f, ow(isbj,:), wbins] = plotWeightsHist(tractographyType,w{isbj});
feSavefig(f,'verbose','yes','figName',[resFileName, '_optimized_weight_hist'],'figDir',savedir,'figType','eps');
close(f)
% 2d histogram
%[f, h] = plotLengthWeightScatter(tractographyType,w{isbj},len{isbj});
%feSavefig(f,'verbose','yes','figName',[resFileName, '_weight_vs_fiberlength_2d'],'figDir',savedir,'figType','eps');
%set(gca, 'visible', 'off')
%feSavefig(f,'verbose','yes','figName',[resFileName, '_weight_vs_fiberlength_2d'],'figDir',savedir,'figType','jpg');
%close(f)
% Just reorganize the vectors
%len{isbj} = len{isbj}';
%w{isbj} = w{isbj}';
% Save the proportion of zero-weight and non-zero-weight fibers
proportionDeleted(isbj,:) = [results.candidate.n - results.optimized.n results.optimized.n]./results.candidate.n;
[f, h] = plotFiberPie(resFileName,proportionDeleted(isbj,:));
feSavefig(f,'verbose','yes','figName',[resFileName, '_pie'],'figDir',savedir,'figType','eps');
close(f)
end
end
% mow = mean(ow,1);
% owerr = [mow;mow]+[-std(ow)./sqrt(size(ow,1));std(ow)./sqrt(size(ow,1))];
% f = plotMeanWeightsHist(tractographyType,mow,owerr,wbins);close(f)
%
% mol = mean(ol,1);
% olerr = [mol;mol]+[-std(ol)./sqrt(size(ow,1));std(ol)./sqrt(size(ow,1))];
% mcl = mean(cl,1);
% clerr = [mcl;mcl]+[-std(cl)./sqrt(size(ow,1));std(cl)./sqrt(size(ow,1))];
% f = plotMeanLengthHist(tractographyType,mol,olerr,mcl,clerr,lbins);close(f)
%
% [f, h] = plotProportionDeleted(tractographyType,proportionDeleted);
% feSavefig(f,'verbose','yes','figName',[tractographyType, '_deleted_fibers_across_subjects'],'figDir','~/2t1/predator/average_figures','figType','eps');
% close(f)
% [f, h] = plotLengthWeightScatter(['_ACROSS_FIVE_SUBJECTS_',tractographyType],cell2mat(w)',cell2mat(len)');
% feSavefig(f,'verbose','yes','figName',[tractographyType, '_weight_vs_fiberlength_2d_across_subjects'],'figDir','~/2t1/predator/average_figures','figType','eps');
% set(gca, 'visible', 'off')
% feSavefig(f,'verbose','yes','figName',[tractographyType, '_weight_vs_fiberlength_2d_across_subjects'],'figDir','~/2t1/predator/average_figures','figType','jpg');
% close(f)
%
end % End main function
function f = plotLengthRatioHist(tractographyType,drl,bins)
fontSiz = 15;
% Make a 2D histogram (scatter plot) of the percent deleted and kept fibers
f = figure('name',sprintf('length_ratio_hist_%s',tractographyType),'color','w');
sh = plot(bins,(drl),'-','color','k','markeredgecolor','w','markerfacecolor','k','markersize',12);
ylabel('Proportion deleted fascicles','fontsize',fontSiz)
xlabel('Fascicle length','fontsize',fontSiz)
set(gca, 'ylim', [0.125/8 .6], ...
'ytick',[0.125/8 0.125/4 0.125/2 0.125 0.25 0.5], ...
'xlim',[bins(1) bins(end)],...
'xtick',[bins(1) 8 32 128 bins(end)],...
'yscale','log', ...
'xscale','log', ...
'tickdir','out','box','off', ...
'fontsize',fontSiz,'visible','on')
end
function [f, oy,bins] = plotWeightsHist(tractographyType,ow)
fontSiz = 15;
% Make a 2D histogram (scatter plot) of the percent deleted and kept fibers
f = figure('name',sprintf('weights_hist_%s',tractographyType),'color','w');
bins = linspace(-6,0,22);
[oy,x] = hist(log10(ow),bins);
sh = plot(bins,oy,'-','color','k','markeredgecolor','w','markerfacecolor','k','markersize',12);
ylabel('Number of fascicles','fontsize',fontSiz)
xlabel('log_{10}(Fiber weight)','fontsize',fontSiz)
set(gca,'xlim',[-6 0],...
'xtick',[-6 -3 0],...
'ylim', [0 24000], ...
'ytick',[0 12000 24000], ...
'tickdir','out','box','off', ...
'fontsize',fontSiz,'visible','on')
end
function [f, oy, cy, bins] = plotLengthHist(tractographyType,olen,clen)
fontSiz = 15;
% Make a 2D histogram (scatter plot) of the percent deleted and kept fibers
f = figure('name',sprintf('fiberlength_hist_%s',tractographyType),'color','w');
bins = [2:2:512];
[oy,x] = hist(olen,bins);
[cy,x] = hist(clen,bins);
patch([x,fliplr(x)],[cy,zeros(size(cy))],[.6 .6 .6]);
hold on
patch([x,fliplr(x)],[oy,zeros(size(oy))],[.4 .4 .4]);
ylabel('Number of fascicles','fontsize',fontSiz)
xlabel('Fascicle length (mm)','fontsize',fontSiz)
set(gca,'xscale','log', ...
'xlim',[2 256],...
'xtick',[2 4 8 16 32 64 128 256],...
'ylim', [0 26000], ...
'ytick',[0 13000 26000], ...
'tickdir','out','box','off', ...
'fontsize',fontSiz,'visible','on')
end
function [f, h] = plotFiberPie(resFileName,proportionDeleted)
% Make a pie graphshwing the percent of values below and above zero
fontSiz = 15;
f = figure('name',sprintf('pie_of_deleted_fibers_%s',resFileName),'color','w');
h = pie( proportionDeleted,[1 0]);
colormap gray
patchObjs = findobj(h,'Type','patch');
textObjs = findobj(h,'Type','text');
oldStr = get(textObjs,{'String'});
val = get(textObjs,{'Extent'});
oldExt = cat(1,val{:});
newStr = {sprintf('%s deleted',oldStr{1});sprintf('%s kept',oldStr{2})};
set(textObjs,{'String'},newStr)
end
% load run01_fliprot_aligned_trilin_csd_lmax10_run01_fliprot_aligned_trilin_brainmask_run01_fliprot_aligned_trilin_wm_prob-800000-fiberStatsResults.mat
% w = results.weights(results.weights>0);
% [f,x_vals] = ecdf(log10(w));
% F = plot(x_vals,f);
% hold on
% G = plot(x_vals,normcdf(x_vals,mean(log10(w)),std(log10(w))),'r-');
% [H, pValue, KSstatistic, criticalValue] = kstest(f)