add DRO, small changes to simulation

Author: DKuppens

doc/create_local_copy_of_repository.md

@@ -8,4 +8,4 @@ The instructions outlined below shows a few (but far from all) ways for creating
 # Alternative approach
 - Create a github user
 - `git clone <repository name>`
-- e.g. `git clone git@github.com:DKuppens/TF2.4_IVIM_code_collection.git`
+- e.g. `git clone git@github.com:OSIPI/TF2.4_IVIM-MRI_CodeCollection.git`

utils/data_simulation/ivim_simulation.py

@@ -28,8 +28,8 @@ def simulate_ivim_signal(D, Dp, f, S0, bvalues, SNR_array, rg):
     simulated_data = simulated_data.cpu().detach().numpy()
 
     # create 2 signal arrays filled with gaussian noise
-    noise_real = rg.normal(0, 1 / SNR_array, (1, len(bvalues)))
-    noise_imag = rg.normal(0, 1 / SNR_array, (1, len(bvalues)))
+    noise_real = rg.normal(0, 1 / SNR, (1, len(bvalues)))
+    noise_imag = rg.normal(0, 1 / SNR, (1, len(bvalues)))
 
     # add Rician noise to the simulated data
     simulated_data = np.sqrt(np.power(simulated_data + noise_real, 2) + np.power(noise_imag, 2)).squeeze()

utils/digital_reference_object/howto_load_dro.md

@@ -0,0 +1,10 @@
+The digital reference object (DRO.npy) is a numpy array containing the following fields:
+- `DRO['D']`: value of diffusion coefficient used for generating signal
+- `DRO['f']`: value of perfusion fraction used for generating signal
+- `DRO['Dp']`: value of pseudo-diffusion coefficient used for generating signal
+- `DRO['S0']`: value of signal at b=0 used for generating signal
+- `DRO['bvals']`: numpy array of b-values used for generating signal
+- `DRO['signals']`: numpy array of signals generated using the above parameters
+
+The DRO can be loaded using the following code:
+data = np.load('DRO.npy', allow_pickle=True)