k-Wave

1. hhamtaii
   

   Member
   Joined: Dec '12
   Posts: 4

   hi Brad,
   the "beamforming_delays" method of "kWaveTransducer" class, is expected to calculate the appropriate delay values, but it does not count for sub_array windows other than centered window. suppose that you have an array of total 16 elements with a sliding active sub-array of 8 elements. delay values for all of these sub-arrays, are mistakenly calculated the same as the centered (middle) 8 elemnts.
   I propose the following solution:
   instead of the present line:

   % create indexing variable
   element_index = -(obj.number_active_elements - 1)/2:(obj.number_active_elements - 1)/2;

   one can replace the following two lines:

   % form a vector of active element indices
   ActiveIdc = find(obj.active_elements == 1);

   % create indexing variable
   element_index = -(obj.number_elements + 1)/2 + ActiveIdc;

   and it works fine.
   and ... I should thank you for your useful toolbox.

   Posted 12 years ago #

2. Bradley Treeby
   

   Developer
   Joined: Mar '10
   Posts: 1,643

   You make a good point, although I believe this is how the delays are normally calculated for linear arrays. Perhaps the most general way would be to add an additional property to the kWaveTransducer class that allows the centre element to be selected. This way, the beamforming delays could be computed using either the centre of the active elements (as currently), or the centre of the complete array as you suggest, or anywhere else. We will take a look before the next release.

   Apologies for the delay in replying!

   Brad.

   Posted 12 years ago #

3. hhamtaii
   

   Member
   Joined: Dec '12
   Posts: 4

   Hi there,
   thanks for your response. that's a very good idea to add a property that allows the center element to be selected. however I think as long as one concerns phased array imaging methods based on multiple sets of active elements, every single scan-line of the final enhanced image would be the sum of corresponding scan-lines of many poor images (each according to a proper sub set of elements). to do so, delays for every set of active elements must have been calculated according to the center of the complete array. otherwise, summing poor images up to form an enhanced image would be meaningless.

   imaging using linear or convex probes, on the other hand, may be achieved without that summing up process because there we can form the final image by attaching partial images together. so no need to keep a constant coordinate i.e. the center of complete elements.

   To the best of my knowledge, beside linear imaging, the only delay values, that are calculated according to the center of active elements (instead of complete elements), are those who concern "defocusing", which is not a matter of k-wave toolbox.

   hope to be useful
   hhamtaii

   Posted 12 years ago #

RSS feed for this topic

Reply

You must log in to post.