-
Notifications
You must be signed in to change notification settings - Fork 0
/
simulate_PCa_experiments.m
160 lines (129 loc) · 6.56 KB
/
simulate_PCa_experiments.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
clear all
quick_test = 0;
% default experiment values
exp.R1P = 1/30; exp.R1L =1/25; exp.kPL = 0.02; exp.std_noise = 0.004; exp.Tarrival = 4; exp.Tbolus = 8;
%exp.std_noise = 0; %noiseless test
R1P_est = 1/30; R1L_est = 1/25; kPL_est = .02;
R1P_lb = 1/40; R1P_ub = 1/15;
R1L_lb = 1/35; R1L_ub = 1/15; % +- 10 s
% good for gammainput:
Rinj_est = 1; Tarrival_est = 4; A_est = 4; Tbolus_est = 8; B_est = Tbolus_est/4; % A - shape, B - scale (~ Tbolus/4 for A=4...)
Tarrival_lb = 0; Tarrival_ub = 12;
B_lb = 6 /4; B_ub = 10 /4;
% % for boxcar input
% Tarrival_est = 1.25; Tbolus_est = 16; Rinj_est = 0.06; % boxcar, flips = 2
% Tarrival_est = 2; Tbolus_est = 17; Rinj_est = 0.06; % boxcar, flips = 3
params_all = {'kPL', 'R1L', 'R1P', 'Rinj', 'Tarrival', 'A','B'};
% Options: Fix R1L, Fix Tarrive, Tbolus
% realistically, need to fit kPL and Rinj, but for simulation purposes
% maybe too much
fit_params = {{'kPL','R1L', 'Rinj', 'Tarrival', 'B'}, ... % 1 - fit everything
{'kPL', 'Rinj', 'Tarrival','B'}, ... % 2 - fix R1L, fit bolus
{'kPL', 'Rinj', 'B'}, ... % 3- fix R1L, fit bolus duration - hard
{'kPL', 'Rinj', 'Tarrival'}, ... % 4- fix R1L, fit bolus arrival - looks easier
{'kPL','R1L', 'Rinj'}, ... % 5- fix entire bolus, but fit amplitude (Rinj)
{'kPL', 'Rinj'}, ... % 6 -fix bolus except rate, fix R1L
{'kPL'}}; % 7- fix R1L and ALL bolus characteristics
fit_desc = {'fitall', ...
'fixR1L',...
'fixR1L+Tarrival', ...
'fixR1L+Tbolus', ...
'fixbolus', ...
'fixR1L+bolus', ...
'fixR1L+bolus+Rinj'};
% % fix Rinj since fitting is just modulating overall SNR
% Fitting Rinj looks to work a little better
% fit_params = {{'kPL','R1L', 'Tarrival', 'Tbolus', 'B'}, ... % fit everything
% {'kPL', 'Tarrival', 'Tbolus','B'}, ... % fix R1L, fit bolus
% {'kPL', 'Tbolus', 'B'}, ... % fix R1L, fit bolus duration
% {'kPL', 'Tarrival'}, ... % fix R1L, fit bolus arrival
% {'kPL','R1L'}, ... % fix entire bolus, but fit amplitude (Rinj), maybe add Tarrival for timing?
% {'kPL'}}; % fix R1L and bolus characteristics
% less apparent variance when fitting Rinj, some more bias though,
% fixed Rinj more T1 variations
% Models: inputless, boxcar input, gamma variate input [not sure how to put in Tbolus though B = Tbolus/2 A=1 ?? FWHM of gamma function?]
fitting_model = {@fit_kPL, @fit_kPL_withinput, @fit_kPL_withgammainput};
% these correspond to {input-less fitting, fitting with boxcar input, fitting with gamma-variate input shape}
% test multiple fitting options
% Ifitting_mode
for Ifit_params = [1:6] %1:length(fit_params) % choose which parameter combinations (fit_params) to fit
for Ifitting_model = [1,3] %1:length(fitting_model) % choose which fitting models (fitting_model) to apply
if Ifit_params>2 && Ifitting_model ==1
% input-less fitting (fit_kPL) doesn't include bolus parameters
continue
end
% we'll assume best case of a good initial guess, then perturb in further
% simulations
clear params_est params_fixed acq fitting
for Iparams = 1:length(params_all)
if find(strcmp(fit_params{Ifit_params}, params_all{Iparams}))
eval(['params_est.(params_all{Iparams}) =' params_all{Iparams} '_est;'])
if exist([params_all{Iparams} '_lb'])
eval(['params_est.([params_all{Iparams} ''_lb'']) =' params_all{Iparams} '_lb;'])
end
if exist([params_all{Iparams} '_ub'])
eval(['params_est.([params_all{Iparams} ''_ub'']) =' params_all{Iparams} '_ub;'])
end
else
eval(['params_fixed.(params_all{Iparams}) =' params_all{Iparams} '_est;'])
end
end
for flip_scheme = [4]% [1:4] % choose which flip angle scheme (flip_scheme) to apply
clear flips_all flips
switch flip_scheme
case 1
% 2D dynamic 10/10 flips
flip_desc = 'const10';
Tacq = 90; TR = 5; N = Tacq/TR;
Npe = 8; Nall = N * Npe;
flips_all(1:2,1:Nall) = repmat([10*pi/180; 10*pi/180], [1 Nall]);
% for same Mz usage
flips(1:2,1:N) = repmat(acos(cos([10*pi/180; 10*pi/180]).^Npe), [1 N]);
case 2
% 2D dynamic 10/20 flips
flip_desc = 'mband10-20';
Tacq = 90; TR = 5; N = Tacq/TR;
Npe = 8; Nall = N * Npe;
flips_all(1:2,1:Nall) = repmat([10*pi/180; 20*pi/180], [1 Nall]);
% for same Mz usage
flips(1:2,1:N) = repmat(acos(cos([10*pi/180; 20*pi/180]).^Npe), [1 N]);
case 3
% 3D dynamic, initial
flip_desc = 'vfa-tramp';
N = 18;
load tramp_vfa_n144_flips.mat
flips_all = flips;
[Sscale, Mzscale] = flips_scaling_factors(flips_all, N);
flips = acos(Mzscale);
TR = 2;
%k12 = 0.05;
%flips(1:2,1:N) = [vfa_const_amp(N, pi/2, exp(-TR * ( k12))); ...
% vfa_const_amp(N, pi/2, exp(-TR * ( - k12)))];
case 4
flip_desc = 'vfa-clinical';
N = 21;
load clinical_flips_20160928_42s.mat
flips_all = flips;
[Sscale, Mzscale] = flips_scaling_factors(flips_all, N);
flips = acos(Mzscale);
TR = 2;
end
acq.TR = TR; acq.N = N; acq.flips = flips;
fitting.fit_fcn = fitting_model{Ifitting_model};
fitting.params_est = params_est; fitting.params_fixed = params_fixed;
if quick_test
fitting.NMC = 40; %fast testing option
end
exp_desc = [func2str(fitting_model{Ifitting_model}) '_' fit_desc{Ifit_params} '_' flip_desc];
tic
[results(Ifit_params, Ifitting_model, flip_scheme) hdata hsim] = HP_montecarlo_evaluation( acq, fitting, exp );
toc
set(hsim,'Name', exp_desc);
if ~quick_test
figure(hsim)
print([ exp_desc], '-dpdf')
end
end
end
end
return