/
s_fe_probabilistic_vs_deterministic_96.m
273 lines (247 loc) · 11.6 KB
/
s_fe_probabilistic_vs_deterministic_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
function fe = s_fe_probabilistic_vs_deterministic_96()
%
% This function llustrates how to:
% - initialize a LIFE structure from a candidate connectome
% - Generate an optimized connectome from a cadidate connectome using the
% LIFE strustrue
%
% fe = s_fe_fit()
%
% INPUTS: none
% OUTPUTS: fe structure the optimized life structure
%
% Copyright Franco Pestilli (2013) Vistasoft Stanford University.
% Get the base directory for the data
datapath = '/marcovaldo/frk/2t1/predator/';
subjects = {...
'HT_96dirs_b2000_1p5iso', ... 'MP_96dirs_b2000_1p5iso', ...
'KW_96dirs_b2000_1p5iso', ...
'KK_96dirs_b2000_1p5iso', ...'JW_96dirs_b2000_1p5iso', ...
'FP_96dirs_b2000_1p5iso', ...
};
if notDefined('saveDir'), savedir = fullfile('/marcovaldo/frk/Dropbox','pestilli_etal_revision',mfilename);end
if notDefined('trackingType'), trackingType = 'lmax10';end
display.evidence = 1;
display.distributions = 1;
for isbj = 1:length(subjects)
saveDir = fullfile(savedir,subjects{isbj});
%% Load two pre-culled connectomes
feProbFileName = sprintf('*%s*prob*recomputed.mat',trackingType);
feDetFileName = '*tensor*recomputed.mat';
connectomesPath = fullfile(datapath,subjects{isbj},'connectomes');
feFileToLoad = dir(fullfile(connectomesPath,feProbFileName));
fname = feFileToLoad(1).name(1:end-4);
feProbFileToLoad = fullfile(connectomesPath,fname);
feFileToLoad = dir(fullfile(connectomesPath,feDetFileName));
fname = feFileToLoad.name(1:end-4);
feDetFileToLoad = fullfile(connectomesPath,fname);
% Extract the RMSE, R_rmse and the coordinates of the white matter
fprintf('[%s] Loading: \n%s\n ======================================== \n\n',mfilename,feProbFileToLoad)
load(feProbFileToLoad);
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
p.rmse = feGetRep(fe,'vox rmse');
p.rrmse = feGetRep(fe,'vox rmse ratio');
p.coords = feGet( fe,'roi coords');
clear fe
% Extract the RMSE, R_rmse and the coordinates of the white matter
fprintf('[%s] Loading: \n%s\n ======================================== \n\n',mfilename,feDetFileToLoad)
load(feDetFileToLoad);
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
d.rmse = feGetRep(fe, 'vox rmse');
d.rrmse = feGetRep(fe, 'vox rmse ratio');
d.coords = feGet( fe, 'roi coords');
clear fe
%% Find the common coordinates between the two connectomes
%
% There are more coordinates in the Prob conectome, because the tracking
% fills up more White-matter.
%
% So, first we find the indices in the probabilistic connectome of the
% coordinate in the deterministic conenctome.
%
% But there are some of the coordinates in the Deterministic conectome that
% are NOT in the Probabilistic connectome.
%
% So, second we find the indices in the Deterministic connectome of the
% subset of coordinates in the Probabilistic connectome found in the
% previous step.
% First we find the coordinates in the Probabilistic conectome that are
% also in the Deterministic connectome.
prob.coordsIdx = ismember(p.coords,d.coords,'rows');
% Second we find the coordinates in the Deterministic connectome that are
% also in the Probabilistic connectome.
prob.coords = p.coords(prob.coordsIdx,:);
det.coordsIdx = ismember(d.coords,prob.coords,'rows');
det.coords = d.coords(det.coordsIdx,:);
% What we really need is detCoordsIdx and probCoordsIdx. These allow us to
% find the common voxel indices in rmse and rrmse, etc.
prob.rmse = p.rmse( prob.coordsIdx);
prob.rrmse = p.rrmse(prob.coordsIdx);
det.rmse = d.rmse( det.coordsIdx);
det.rrmse = d.rrmse(det.coordsIdx);
%% Strength of evidence
se = feComputeEvidence(prob.rmse,det.rmse);
if display.distributions
% Raw RMSE distirbutions
fig(1).name = sprintf('rmse_distributions_%s',mfilename);
fig(1).fh = figure('name',fig(1).name,'color','w');
set(fig(1).fh,'Units','normalized','Position',[0.007 0.55 0.28 0.36]);
plot(se.lesion.xhist,se.lesion.hist,'-','color', [.95 .45 .1],'linewidth',2); hold on
plot(se.nolesion.xhist,se.nolesion.hist,'-','linewidth',2, 'color', [.1 .45 .95])
plot([se.nolesion.rmse.mean,se.nolesion.rmse.mean], [0,0.2],'-','color',[.1 .45 .95] )
plot([se.lesion.rmse.mean,se.lesion.rmse.mean], [0,0.2], '-', 'color',[.95 .45 .1])
title(sprintf('mean RMSE\nno-lesion %2.3f | lesion %2.2f', ...
se.nolesion.rmse.mean,se.lesion.rmse.mean),'fontsize',16)
ylabel('Probability', 'fontsize',14);xlabel('RMSE', 'fontsize',14)
legend({'Lesion','No lesion'},'fontsize',14);
set(gca,'box','off','xtick',[0 round(se.xrange(2)/2) se.xrange(2)], ...
'ytick',[0 .1 .2],'xlim',[0 se.xrange(2)],'ylim',[0 .25], ...
'tickdir', 'out', 'ticklength', [0.025 0]);
% Plot the null distribution and the empirical difference
ywo_e = se.s.lesioned_e;
y_e = se.s.unlesioned_e;
woxhis = se.s.lesioned.xbins;
xhis = se.s.unlesioned.xbins;
min_x = se.s.min_x;
max_x = se.s.max_x;
fig(2).name = sprintf('virtual_lesion_test_mean_rmse_hist_%s_%s',mfilename,'prob_det');
fig(2).fh = figure('name',fig(2).name,'color','w');
set(fig(2).fh,'Units','normalized','Position',[0.007 0.55 0.28 0.36]);
patch([xhis,xhis],y_e(:),[.1 .45 .95],'FaceColor',[.1 .45 .95],'EdgeColor',[.1 .45 .95]); % Distribution as the +/- 2SD
hold on
patch([woxhis,woxhis],ywo_e(:),[.95 .45 .1],'FaceColor',[.95 .45 .1],'EdgeColor',[.95 .45 .1]); % Distribution as the +/- 2SD
set(gca,'tickdir','out', ...
'box','off', ...
'ylim',[0 0.6], ...
'xlim',[min_x,max_x], ...
'ytick',[0 0.25 0.5], ...
'xtick',round(linspace(min_x,max_x,4)), ...
'fontsize',16)
ylabel('Probability','fontsize',16)
xlabel('rmse','fontsize',16')
title(sprintf('Strength of connection evidence %2.3f',(se.s.mean)), ...
'FontSize',16)
saveFig(fig(1).fh,fullfile(saveDir, fig(1).name),1)
end
% RMSE scatter-density plot of Probabilistic and Deterministic
figNameRmse = sprintf('prob_vs_det_rmse_common_voxels_ma_%s',fname);
fh = scatterPlotRMSE(det,prob,figNameRmse);
saveFig(fh,fullfile(saveDir, figNameRmse),1)
saveFig(fh,fullfile(saveDir, figNameRmse),0)
if display.evidence
% Binned RMSE distributions
fig(3).name = sprintf('Size_of_effect_of_the_lesion_%s',mfilename);
fig(3).fh = figure('name',fig(3).name,'color','w');
set(fig(3).fh,'Units','normalized','Position',[0.007 0.55 0.28 0.36]);
subplot(1,4,1)
plot(1,se.s.mean,'-o','color', [.95 .45 .1],'linewidth',2); hold on
plot([1,1], [se.s.mean,se.s.mean] + [-se.s.std,se.s.std], '-','color',[.95 .45 .1] )
ylabel('S (s.d.)', 'fontsize',14);
set(gca,'box','off','xlim',[0 2], 'ylim',[0 ceil(se.s.mean + se.s.std)], ...
'tickdir', 'out', 'ticklength', [0.025 0])
subplot(1,4,2)
plot(1,se.em.mean,'-o','color', [.95 .45 .1],'linewidth',2); hold on
ylabel('Earth mover''s distance', 'fontsize',14);
set(gca,'box','off','xlim',[0 2], 'ylim',[0 ceil(se.em.mean)], ...
'tickdir', 'out', 'ticklength', [0.025 0])
subplot(1,4,3)
plot(1,se.kl.mean,'-o','color', [.95 .45 .1],'linewidth',2); hold on
ylabel('K-L divergence (bits)', 'fontsize',14);
set(gca,'box','off','xlim',[0 2], 'ylim',[0 ceil(se.kl.mean)], ...
'tickdir', 'out', 'ticklength', [0.025 0])
subplot(1,4,4)
plot(1,se.j.mean,'-o','color', [.95 .45 .1],'linewidth',2); hold on
ylabel('Jeffrey''s divergence (bits)', 'fontsize',14);
set(gca,'box','off','xlim',[0 2], 'ylim',[0 ceil(se.j.mean)], ...
'tickdir', 'out', 'ticklength', [0.025 0])
saveFig(fig(2).fh,fullfile(saveDir, fig(2).name),1)
saveFig(fig(3).fh,fullfile(saveDir, fig(3).name),1)
end
close all
save(fullfile(saveDir,sprintf('average_results_%s.mat',trackingType)),'se')
end
end
%------------------------------------%
function fhRmseMap = scatterPlotRMSE(det,prob,figNameRmse)
fhRmseMap = mrvNewGraphWin(figNameRmse);
[ymap,x] = hist3([det.rmse;prob.rmse]',{[10:1:90], [10:1:90]});
ymap = ymap./length(prob.rmse);
sh = imagesc(flipud(log10(ymap)));
cm = colormap(flipud(hot)); view(0,90);
axis('square')
set(gca, ...
'xlim',[1 length(x{1})],...
'ylim',[1 length(x{1})], ...
'ytick',[1 (length(x{1})/2) length(x{1})], ...
'xtick',[1 (length(x{1})/2) length(x{1})], ...
'yticklabel',[x{1}(end) x{1}(round(end/2)) x{1}(1)], ...
'xticklabel',[x{1}(1) x{1}(round(end/2)) x{1}(end)], ...
'tickdir','out','ticklen',[.025 .05],'box','off', ...
'fontsize',16,'visible','on')
hold on
plot3([1 length(x{1})],[length(x{1}) 1],[max(ymap(:)) max(ymap(:))],'k-','linewidth',1)
ylabel('Deterministic_{rmse}','fontsize',12)
xlabel('Probabilistic_{rmse}','fontsize',12)
cb = colorbar;
tck = get(cb,'ytick');
set(cb,'yTick',[min(tck) mean(tck) max(tck)], ... 'yTickLabel',10.^[min(tck), mean(tck), max(tck)], ...
'tickdir','out','ticklen',[.025 .05],'box','on', ...
'fontsize',16,'visible','on')
end
%---------------------------------------%
function fh = distributionPlotRMSE(y_e,ywo_e,dprime,xhis,woxhis,figName)
h1.ylim = [0 0.6];
h1.xlim = [22,34];
h1.ytick = [0 0.3 0.6];
h1.xtick = [28 30 32 34];
h2.ylim = [0 0.4];
h2.xlim = [28,32];
h2.ytick = [0 0.2 0.4];
h2.xtick = [28 30 32];
histcolor{1} = [0 0 0];
histcolor{2} = [.95 .6 .5];
fh = mrvNewGraphWin(figName);
patch([xhis,xhis],y_e(:),histcolor{1},'FaceColor',histcolor{1},'EdgeColor',histcolor{1});
hold on
patch([woxhis,woxhis],ywo_e(:),histcolor{2},'FaceColor',histcolor{2},'EdgeColor',histcolor{2});
set(gca,'tickdir','out', ...
'box','off', ...
'ylim',[0 .6], ...
'xlim',[28 34], ...
'xtick',[28 30 32 34], ...
'ytick',[0 .3 .6], ...
'fontsize',16)
ylabel('Probability','fontsize',16)
xlabel('rmse','fontsize',16')
title(sprintf('Strength of evidence:\n mean %2.3f - std %2.3f',mean(dprime),std(dprime)), ...
'FontSize',16)
legend({'Probabilistic','Deterministic'})
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