k-Wave

1. federico.rossano
   

   Member
   Joined: Dec '21
   Posts: 7

   Hi,

   I would like to ask if there is a strategy to calculate Isppa and Ispta values without integrating over the pulse duration the square of p(t). I have problems due to RAM memory with computations of both big domains and time vectors, so I am unable to compute sensor.p to achieve the I(t) through p^2/2*rho*c.

   I've already considered to start recording my signal in the steady state, but it still requires a lot of memory. Maybe I can calculate Isppa/Ispta with p_max or p_max_all?
   And what is the difference between the latter two?

   Thank you in advance for the help!
   Federico

   Posted 5 months ago #

2. bencox
   

   Developer
   Joined: Mar '10
   Posts: 284

   Hi Federco,

   There are various things you could try.

   kspaceFirstOrder3D will directly output average intensity I_avg as well as the time-varying intensity I from which you could compute the average. You can ask it to do that by setting sensor.record = {'I_avg'} or sensor.record = {'I}.

   Beware there may be a transient at the start of a signal which may have a higher peak than the steady state, so you may want to consider setting sensor.record_start_index to avoid a possible transient.

   If your wave is close enough to a single frequency plane wave in a steady state then you could also find the pressure amplitude and use the formula you gave above for the time-averaged intensity. You could use p_max if you just look at the steady state part of the propagation by avoiding the transient. (p_max is the maximum pressure at the defined sensor points, not everywhere.)

   If you are recording on very many points, eg. at all points in the domain, and that means the memory require is still very high, could you reduce the region over which you are recording? Or break it into subregions and run one simulation per subregion?

   Best wishes
   Ben

   Posted 5 months ago #

3. federico.rossano
   

   Member
   Joined: Dec '21
   Posts: 7

   Hi Ben,

   Thank you for your answer and for the amazing work.

   I would just only clarify the point about p_max. In the k-Wave User Manual page 38, I see:
   [...] The exceptions are the averaged quantities (‘p_max’, ‘p_rms’, ‘u_max, ‘p_rms, ‘I_avg’) and the quantities recorded over all the grid points within the domain (‘p_max_all’, ‘p_min_all’,‘p_final’, ‘u_max_all’, ‘u_min_all’, ‘u_final’). The averaged quantities return the maximum, average, or root-mean-squared values at each sensor point for the complete simulation and are indexed as (sensor_point_index).

   For "averaged quantities" you mean averaged on time? As for p_max, does k-Wave take the maximum pressure in each voxel for each time step and then the result is the average of these maximum pressures recorded over the time vector? Or it just takes the maximum pressure recorded at any time during the simulation?

   Thank you so much again,
   Besti wishes,
   Federico

   Posted 4 months ago #

RSS feed for this topic

Reply

You must log in to post.