/
s_ms_directions_96.m
352 lines (310 loc) · 15.8 KB
/
s_ms_directions_96.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
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
335
336
337
338
339
340
341
342
343
344
345
346
347
348
349
350
351
352
function s_ms_directions_96(trackingType)
%
% Load FE structeres obtained by preprocessing connectomesconstrained within a
% region of interest and within the cortex and makes some basic plot of
% statistics in the connectomes
%
% See also:
% Get the base directory for the data
datapath = '/marcovaldo/frk/2t1/predator/';
subjects = {...
'KK_96dirs_b2000_1p5iso', ... 'JW_96dirs_b2000_1p5iso', ...
'HT_96dirs_b2000_1p5iso', ...
'KW_96dirs_b2000_1p5iso', ...
'FP_96dirs_b2000_1p5iso', ...
'MP_96dirs_b2000_1p5iso', ...
};
if notDefined('saveDir'), savedir = fullfile('/marcovaldo/frk/Dropbox','pestilli_etal_revision',mfilename);end
if notDefined('trackingType'), trackingType = 'lmax10';end
if notDefined('numDirs'), numDirs = [96,64,32,16,8];end
addpath(genpath('/marcovaldo/frk/git/boot_dwi'))
doFD = 1;
figVisible = 'off';
probIndex = 1; %for 2000bval, Deterministic index 2: for 200 bval
for isbj = 1:length(subjects)
% High-resolution Anatomy
saveDir = fullfile(savedir,subjects{isbj});
% File to load
connectomesPath = fullfile(datapath,subjects{isbj},'connectomes');
feFileToLoad = dir(fullfile(connectomesPath,sprintf('*%s*recomputed.mat',trackingType)));
fname = feFileToLoad(probIndex).name(1:end-4);
feFileToLoad = fullfile(connectomesPath,fname);
fprintf('[%s] Loading: \n%s\n ======================================== \n\n',mfilename,feFileToLoad)
load(feFileToLoad);
fprintf('[%s] Extracting info: \n%s\n ======================================== \n\n',mfilename,feFileToLoad)
for iNumDirs = 1:length(numDirs)
xform = feGet(fe,'xform img 2 acpc');
mapsize = feGet(fe, 'map size');
nBvecs = feGet(fe,'nbvecs');
nBvals = fe.life.imagedim(4) - nBvecs;
nVoxels= feGet(fe,'nvoxels');
if isempty(fe.rep)
if ~isempty(strfind(fe.path.dwifilerep,'home'))
fe.path.dwifilerep = fullfile('/marcovaldo/',fe.path.dwifilerep(strfind(fe.path.dwifilerep,'home')+length('home'):end));
end
fe = feConnectomeSetDwi(fe,fe.path.dwifilerep,true);
end
rmseM = feGetRep(fe, 'vox rmse');
rmseD = feGetRep(fe, 'vox rmse data');
rmseR = feGetRep(fe, 'vox rmse ratio');
if isempty(fe.fg)
fe.path.savedir = fullfile('/marcovaldo/',fe.path.savedir(strfind(fe.path.savedir,'home')+length('home'):end));
fiberPath = fullfile(fileparts(fe.path.savedir),'fibers');
fibers = dir(fullfile(fiberPath,sprintf('*%s*.pdb',trackingType)));
fe = feSet(fe,'fg from acpc',fgRead(fullfile(fiberPath,fibers.name)));
end
w = feGet(fe,'fiber weights');
% Compute the total number of fibers retained at each direction number.
m.optimized.nfibers(iNumDirs,isbj) = sum(w > 0);
m.optimized.ndirs(iNumDirs,isbj) = numDirs(iNumDirs);
numFibers_total(iNumDirs,isbj) = length(w);
numFibers_good(iNumDirs,isbj) = sum(w > 0);
fgOpt = fgExtract(feGet(fe,'fibers acpc'),w > 0,'keep');
m.nfibers.y(iNumDirs,isbj) = numFibers_good(iNumDirs,isbj);
m.rmse.dataMean(iNumDirs,isbj) = mean(rmseD);
m.rmse.modelMean(iNumDirs,isbj) = mean(rmseM);
m.rmse.dataMedian(iNumDirs,isbj) = median(rmseD);
m.rmse.modelMedian(iNumDirs,isbj) = median(rmseM);
m.rrmse.mean(iNumDirs,isbj) = mean(rmseR);
m.rrmse.median(iNumDirs,isbj) = median(rmseR);
%theseIndices = randsample(1:nBvecs,numDirs(iNumDirs));
if numDirs(iNumDirs) < size(fe.life.bvecs,1)
[~, theseIndices] = bd_subsample(fe.life.bvecs',numDirs(iNumDirs));
else
theseIndices = 1:numDirs(iNumDirs);
end
% Update the FE structure fields that depend on the bvecs
fe.life.bvecs = fe.life.bvecs(theseIndices,:);
%if ~all(all(fe.life.bvecs==sample_bvecs));keyboard,end
fe.life.bvals = fe.life.bvals(theseIndices);
fe.life.bvecsindices = fe.life.bvecsindices(theseIndices);
fe.life.diffusion_signal_img = fe.life.diffusion_signal_img(:,theseIndices);
fe.life.imagedim(end) = numDirs(iNumDirs) + nBvals;
fe.rep.bvecs = fe.rep.bvecs(theseIndices,:);
fe.rep.bvals = fe.rep.bvals(theseIndices);
fe.rep.bvecsindices = fe.rep.bvecsindices(theseIndices);
fe.rep.diffusion_signal_img = fe.rep.diffusion_signal_img(:,theseIndices);
fe.rep.diffusion_S0_img = fe.rep.diffusion_S0_img(theseIndices);
fe.rep.imagedim(end) = numDirs(iNumDirs) + nBvals;
directionsIndices = false(nBvecs,1);
directionsIndices(theseIndices) = true;
allIndices = repmat(directionsIndices,nVoxels,1);
fe.life.Mfiber = fe.life.Mfiber(allIndices,:);
fe.life.dSig = fe.life.dSig(allIndices);
fe = feSet(fe,'fit',feFitModel(fe.life.Mfiber,fe.life.dSig','bbnnls'));
if doFD
fprintf('[%s] Computing fiber density: \n%s\n ======================================== \n\n',mfilename,feFileToLoad)
% Get the fiber density
% fd = feGet(fe,'fiber density');
fdImg = dtiComputeFiberDensityNoGUI(feGet(fe,'fibers acpc'), xform, mapsize);
fdOImg = dtiComputeFiberDensityNoGUI(fgOpt, xform, mapsize);
end
fprintf('[%s] Making histrograms: \n%s\n ======================================== \n\n',mfilename,feFileToLoad)
% Histogram plots
fdImg(fdImg==0) = nan;
fdOImg(fdOImg==0) = nan;
m.density.candidate_mean(iNumDirs,isbj) = nanmean( fdImg(:));
m.density.candidate_median(iNumDirs,isbj) = nanmedian( fdImg(:));
m.density.optimal_mean(iNumDirs,isbj) = nanmean( fdOImg(:));
m.density.candidate_median(iNumDirs,isbj) = nanmedian(fdOImg(:));
x = [1 2.^[1 2 3 4 5 6 7 8 9 10]];
x = [2:2:512];
[yFD(iNumDirs,isbj,:),xFD(iNumDirs,isbj,:)] = hist(fdImg(:),x);
yFD(iNumDirs,isbj,:) = 100*yFD(iNumDirs,isbj,:)./sum(yFD(iNumDirs,isbj,:));
[yoFD(iNumDirs,isbj,:),xoFD(iNumDirs,isbj,:)]= hist(fdOImg(:),x);
yoFD(iNumDirs,isbj,:) = 100*yoFD(iNumDirs,isbj,:)./sum(yoFD(iNumDirs,isbj,:));
x = 0:5:100;
[yRMSE(iNumDirs,isbj,:),xRMSE(iNumDirs,isbj,:)] = hist(rmseD(:),x);
yRMSE(iNumDirs,isbj,:) = 100*yRMSE(iNumDirs,isbj,:)./sum(yRMSE(iNumDirs,isbj,:));
[yoRMSE(iNumDirs,isbj,:),xoRMSE(iNumDirs,isbj,:)]= hist(rmseM(:),x);
yoRMSE(iNumDirs,isbj,:) = 100*yoRMSE(iNumDirs,isbj,:)./sum(yoRMSE(iNumDirs,isbj,:));
x = logspace(-.3,.3,32);
[yRrmse(iNumDirs,isbj,:),xRrmse(iNumDirs,isbj,:)] = hist(rmseR(:),x);
yRrmse(iNumDirs,isbj,:) = 100*(yRrmse(iNumDirs,isbj,:)./sum(yRrmse(iNumDirs,isbj,:)));
clear fdOImg fdImg
end
end
m.nfibers.x = numDirs;
% Average histograms
saveDir = fullfile(savedir,'average_96_1p5mm');
% Save the results to file, it takes along time to load all these FE strctures...
m.density.candidatey = squeeze(mean(yFD,2));
m.density.candidateSte= squeeze(std(yFD,[],2)./sqrt(size(yFD,2)));
m.density.optimaly = squeeze(mean(yoFD,2));
m.density.optimalSte = squeeze(std(yoFD,[],2)./sqrt(size(yoFD,2)));
m.density.x = squeeze(xFD(:,isbj,:));
m.density.units = {'x=Fascicles per voxel','y=percent voxels'};
m.density.yFD=yFD;
m.density.yoFD=yoFD;
% rmse data vs. model
m.rmse.data = squeeze(mean(yRMSE,2));
m.rmse.dataSte = squeeze(std(yRMSE,[],2)./sqrt(size(yRMSE,2)));
m.rmse.model = squeeze(mean(yoRMSE,2));
m.rmse.modelSte= squeeze(std(yoRMSE,[],2)./sqrt(size(yoRMSE,2)));
m.rmse.x = squeeze(xRMSE(:,isbj,:));
m.rmse.units = {'x=rmse (raw scanner units)','y=percent voxels'};
m.rmse.yRMSE=yRMSE;
m.rmse.yoRMSE=yoRMSE;
% rmse data vs. model
m.rrmse.y = squeeze(mean(yRrmse,2));
m.rrmse.ste = squeeze(std(yRrmse,[],2)./sqrt(size(yRrmse,2)));
m.rrmse.x = squeeze(xRrmse(:,isbj,:));
m.rrmse.units = {'x=Rrmse (a.u.)','y=percent voxels'};
m.rrmse.yRrmse=yRrmse;
mkdir(saveDir)
save(fullfile(saveDir,'mean_histograms_recomputed_steps10.mat'),'m','numFibers_total','numFibers_good')
% Histogram plots
figName = sprintf('FibDensHistCandVSOpt_NDirs96__%i_%i_%i_%i_%s_RECOM',numDirs, fname);
colors = {[.9 .3 .3],[.9 .45 .35],[.9 .55 .5],[.9 .6 .6],[.9 .8 .8]};
fh = figure('name',figName,'visible',figVisible,'color','w');
for iNDirs = 1:size(m.density.x,1)
semilogx(m.density.x(iNDirs, :),m.density.candidatey(iNDirs, :),'k-','linewidth',2);
hold on
semilogx([m.density.x(iNDirs, :);m.density.x(iNDirs, :)], [m.density.candidatey(iNDirs, :)-m.density.candidateSte(iNDirs, :); ...
m.density.candidatey(iNDirs, :)+m.density.candidateSte(iNDirs, :)],'k-');
semilogx(m.density.x(iNDirs, :),m.density.optimaly(iNDirs, :),'r-','linewidth',2, 'color',colors{iNDirs})
semilogx([m.density.x(iNDirs, :);m.density.x(iNDirs, :)],[m.density.optimaly(iNDirs, :)-m.density.optimalSte(iNDirs, :); ...
m.density.optimaly(iNDirs, :)+m.density.optimalSte(iNDirs, :)],'-','linewidth',2,'color',colors{iNDirs})
ylabel('Percent voxels','FontSize',16,'FontAngle','oblique')
xlabel('Fascicles per voxel','FontSize',16,'FontAngle','oblique')
legend(gca,{'Candidated','Optimized'},'box','off')
set(gca,'fontsize',16, ...
'ylim', [0 30], ...
'ytick',[0 15 30], ...
'xlim', [0.5 2^10],'xtick',[0 m.density.x(iNDirs, :)],...
'box','off','tickdir','out','ticklength',[0.025 0])
saveFig(fh,fullfile(saveDir, figName),1)
end
figName = sprintf('RMSE_mean_HistDataVSOpt_NDirs96_%i_%i_%i_%i_%s_RECOMP',numDirs, fname);
fh = figure('name',figName,'visible',figVisible,'color','w');
for iNDirs = 1:size(m.rmse.x,1)
plot(m.rmse.x(iNDirs, :),m.rmse.data(iNDirs, :),'k-','linewidth',2,'color',[.15 .15 .15].*iNDirs)
hold on
plot(m.rmse.x(iNDirs, :),m.rmse.model(iNDirs, :),'r-','linewidth',2, 'color',colors{iNDirs})
plot([m.rmse.x(iNDirs, :);m.rmse.x(iNDirs, :)], ...
[m.rmse.data(iNDirs, :)-m.rmse.dataSte(iNDirs, :);m.rmse.data(iNDirs, :)+m.rmse.dataSte(iNDirs, :)],'k-','linewidth',2,'color',[.1 .1 .1].*iNDirs);
plot([m.rmse.x(iNDirs, :);m.rmse.x(iNDirs, :)],[m.rmse.model(iNDirs, :)-m.rmse.modelSte(iNDirs, :);m.rmse.model(iNDirs, :)+m.rmse.modelSte(iNDirs, :)],'r-','linewidth',2, 'color',colors{iNDirs})
ylabel('Percent voxels','FontSize',16,'FontAngle','oblique')
xlabel('RMSE (raw scanner units)','FontSize',16,'FontAngle','oblique')
legend(gca,{'Data','Model'},'box','off')
set(gca,'fontsize',16, ...
'ylim', [0 30], ...
'ytick',[0 15 30], ...
'xlim', [0 100],'xtick',[0 50 100],...
'box','off','tickdir','out','ticklength',[0.025 0])
saveFig(fh,fullfile(saveDir, figName),1)
end
for iNDirs = 1:size(m.rrmse.x,1)
figName = sprintf('RRMSE_ratio_HistDataVSOpt_NDirs%i_%s_RECOMP',numDirs(iNDirs), fname);
fh = figure('name',figName,'visible',figVisible,'color','w');
xpatch = [.5 m.rrmse.x(iNDirs, m.rrmse.x(iNDirs, :)<=1) 1];
ypatch = m.rrmse.y(iNDirs, m.rrmse.x(iNDirs, :)<=1);
proportionModelBetter = sum(ypatch);
ypatch = [0 ypatch ypatch(end)];
patch([xpatch,fliplr(xpatch)],[ypatch,zeros(size(ypatch)),],[.8 .8 .8])
hold on
plot([1 1],[0 14],'k--')
plot(m.rrmse.x(iNDirs, :),m.rrmse.y(iNDirs, :),'k-','linewidth',2)
plot([m.rrmse.x(iNDirs, :);m.rrmse.x(iNDirs, :)],[m.rrmse.y(iNDirs, :)-m.rrmse.ste(iNDirs, :);m.rrmse.y(iNDirs, :)+m.rrmse.ste(iNDirs, :)],'r-','linewidth',2)
ylabel('Percent voxels','FontSize',16,'FontAngle','oblique')
xlabel('R_{rmse}','FontSize',16,'FontAngle','oblique');
title(sprintf('R < 1 in %2.0f%% of voxels',round(proportionModelBetter)),'FontSize',16,'FontAngle','oblique')
set(gca,'fontsize',16, ...
'ylim', [0 14], ...
'ytick',[0 7 14], 'xscale', 'log', ...
'xlim', [.5 2],'xtick',[.5 1 2],...
'box','off','tickdir','out','ticklength',[0.025 0])
saveFig(fh,fullfile(saveDir, figName),1)
end
figName = sprintf('NFibers_hist_NDirs%i_%i_%i_%i_%i_%s_RECOMP',numDirs, fname);
fh = figure('name',figName,'visible',figVisible,'color','w');
y = m.nfibers.y(:,1:length(numDirs))./500000;
x = m.nfibers.x;
semilogx(x,mean(y,1),'ko-')
hold on
semilogx([x; x], ...
[mean(y,1);mean(y,1)] + [std(y,[],1)./sqrt(size(y,1));-std(y,[],1)./sqrt(size(y,1))],'k-')
ylabel('Proportion supported fascicles','FontSize',16,'FontAngle','oblique')
xlabel('Number of diffusion directions','FontSize',16,'FontAngle','oblique')
set(gca,'fontsize',16, ...
'xlim', [6 100],'xtick',fliplr(x),...
'ylim', [.25 .75],'ytick',[.25 .5 .75],...
'box','off','tickdir','out','ticklength',[0.025 0])
saveFig(fh,fullfile(saveDir, figName),1)
end % Main function
%---------------------------------%
function saveMapSagital(fh,figName,saveDir,M,m,SD,maxfd,map)
% This helper function saves two figures for each map and eps with onlythe
% axis and a jpg with only the brain slice.
% The two can then be combined in illustrator.
%
% First we save only the slice as jpeg.
set(gca,'fontsize',16,'ytick',[-80 -40 0 40 80], ...
'ztick',[-40 0 40 80], ...
'xlim',[-80 80],'ylim',[-110 100],'zlim',[-60 80],'tickdir','out','ticklength',[0.025 0])
axis off
saveFig(fh,fullfile(saveDir,figName),'tiff')
saveFig(fh,fullfile(saveDir,figName),'png')
% Then we save the slice with the axis as
% eps. This will only generate the axis
% that can be then combined in illustrator.
axis on
grid off
title(sprintf('mean %2.2f | median %2.2f | SD %2.2f', ...
M,m,SD),'fontsize',16,'FontAngle','oblique')
zlabel('Z (mm)','fontsize',16,'FontAngle','oblique')
xlabel('X (mm)','fontsize',16,'FontAngle','oblique')
cmap = colormap(eval(sprintf('%s(255)',map)));
colorbar('ytick',linspace(0,1,5),'yticklabel', ...
{1, num2str(ceil(maxfd/8)), num2str(ceil(maxfd/4)), ...
num2str(ceil(maxfd/2)), num2str(ceil(maxfd))}, ...
'tickdir','out','ticklength',[0.025 0],'fontsize',16)
saveFig(fh,fullfile(saveDir,figName),1)
end
%---------------------------------%
function saveMapCoronal(fh,figName,saveDir,M,m,SD,maxfd,map)
% This helper function saves two figures for each map and eps with onlythe
% axis and a jpg with only the brain slice.
% The two can then be combined in illustrator.
%
% First we save only the slice as jpeg.
set(gca,'fontsize',16,'ztick',[-20 -10 0 10 20], ...
'xtick',[0 10 20 30 40 50], ...
'xlim',[-5 70],'zlim',[-30 40],'tickdir','out','ticklength',[0.025 0])
axis off
saveFig(fh,fullfile(saveDir, figName),'tiff')
saveFig(fh,fullfile(saveDir, figName),'png')
% Then we save the slice with the axis as
% eps. This will only generate the axis
% that can be then combined in illustrator.
axis on
grid off
title(sprintf('mean %2.2f | median %2.2f | SD %2.2f', ...
M,m,SD),'fontsize',16,'FontAngle','oblique')
zlabel('Z (mm)','fontsize',16,'FontAngle','oblique')
xlabel('X (mm)','fontsize',16,'FontAngle','oblique')
cmap = colormap(eval(sprintf('%s(255)',map)));
colorbar('ytick',linspace(0,1,5),'yticklabel', ...
{1, num2str(ceil(maxfd/8)), num2str(ceil(maxfd/4)), ...
num2str(ceil(maxfd/2)), num2str(ceil(maxfd))}, ...
'tickdir','out','ticklength',[0.025 0],'fontsize',16)
saveFig(fh,fullfile(saveDir, figName),1)
end
%-------------------------------%
function saveFig(h,figName,eps)
if ~exist( fileparts(figName), 'dir'), mkdir(fileparts(figName));end
fprintf('[%s] saving figure... \n%s\n',mfilename,figName);
switch eps
case {0,'jpeg'}
eval(sprintf('print(%s, ''-djpeg90'', ''-opengl'', ''%s'')', num2str(h),[figName,'.jpg']));
case {1,'eps'}
eval(sprintf('print(%s, ''-cmyk'', ''-painters'',''-depsc2'',''-tiff'',''-r500'' , ''-noui'', ''%s'')', num2str(h),[figName,'.eps']));
case 'png'
eval(sprintf('print(%s, ''-dpng'',''-r500'', ''%s'')', num2str(h),[figName,'.png']));
case 'tiff'
eval(sprintf('print(%s, ''-dtiff'',''-r500'', ''%s'')', num2str(h),[figName,'.tif']));
case 'bmp'
eval(sprintf('print(%s, ''-dbmp256'',''-r500'', ''%s'')', num2str(h),[figName,'.bmp']));
otherwise
end
end