http://www.k-wave.org/forum/topic/point-target-simulation Support for the k-Wave MATLAB toolbox en-US Tue, 12 May 2026 23:38:53 +0000 http://bbpress.org/?v=1.0.2 q http://www.k-wave.org/forum/search.php http://www.k-wave.org/forum/topic/point-target-simulation#post-473 Sat, 21 Apr 2012 09:03:00 +0000 Bradley Treeby 473@http://www.k-wave.org/forum/ <p>Hi Jun,</p> <p>The simulation time will depend on how big your grid size is, and how many scan lines you wish to simulate. The grid size is set by the overall size of the domain along with the maximum frequency that you are interested in modelling (<code>dx</code> should be at least half the minimum wavelength or smaller). If the problem is small enough to fit on a GPU, then that will be fastest, otherwise, there's not a lot else you can do besides settings <code>&#39;DataCast&#39;</code> to <code>&#39;single&#39;</code> (this is already done in the example).</p> <p>What operating system are you using? If you are on 64-bit linux, there is some possibility of getting you an optimised C++ version (which you can still run through MATLAB) to try out. It will eventually be released with k-Wave, but we are still debugging at this stage. Depending on your system, it's about an order of magnitude faster than the MATLAB code.</p> <p>If you look at line 239 of the example, you'll see that the results of each simulation are passed to <code>sensor_data</code>. This contains the raw RF data for each element of the transducer before it is summed to give the scan line. Note that the raw data is not stored permanently, and <code>sensor_data</code> is over-written each scan line. It would be fairly easy to change this though. </p> <p>Brad. </p> http://www.k-wave.org/forum/topic/point-target-simulation#post-470 Fri, 20 Apr 2012 09:42:36 +0000 acegi70 470@http://www.k-wave.org/forum/ <p>Hi Brad,</p> <p>I am finally looking into some of the examples and trying to learn how k-wave works and came up with a couple of quick questions.</p> <p>1. I was looking at the "Simulating B-mode Ultrasound Images" example and I'd like to modify this example to simulate a B-mode image of a single wire target immersed in water on a CPU. Would this still take days to run? What would be the most efficient, fastest way you would suggest me to do this?</p> <p>2. I believe if you run the "Simulating B-mode Ultrasound Images" example, after hours and days of simulation it will generate the beamformed RF data just like one stored in example_us_bmode_scan_lines.mat file that consist of 96 scanlines. Is it possible to gain access to the channel RF data that, when summed, lead to each scanline in example_us_bmode_scan_lines.mat?</p> <p>Thanks,<br /> Jun </p>