Closed MargaretDuff closed 1 month ago
I think https://github.com/SyneRBI/SIRF-Exercises/blob/7a97fdea85cd24dce6c9fd3169be194c76d447ee/notebooks/PET/Solution_Snippets/DIY_OSEM_01_MLEM.py#L1-L9
should be
def MLEM(acquired_data, acq_model, initial_image, num_iterations): estimated_image = initial_image.clone() sensitivity = acq_model.backward(acquired_data.get_uniform_copy(1)) for i in range(num_iterations): quotient = acquired_data/acq_model.forward(estimated_image) # y / (Ax + b) quotient.fill(numpy.nan_to_num(quotient.as_array())) mult_update = acq_model.backward(quotient)/sensitivity # A^t * quotient / A^t1 mult_update.fill(numpy.nan_to_num(mult_update.as_array())) estimated_image*=mult_update return estimated_image
Closed by #237
I think https://github.com/SyneRBI/SIRF-Exercises/blob/7a97fdea85cd24dce6c9fd3169be194c76d447ee/notebooks/PET/Solution_Snippets/DIY_OSEM_01_MLEM.py#L1-L9
should be