coreyjkelly Support for the k-Wave MATLAB toolbox en-US Wed, 13 May 2026 15:38:18 +0000 http://bbpress.org/?v=1.0.2 q http://www.k-wave.org/forum/search.php http://www.k-wave.org/forum/topic/define-sensor-by-cartesian-coordinates#post-6775 Sun, 17 Mar 2019 21:54:06 +0000 Bradley Treeby 6775@http://www.k-wave.org/forum/ <p>Hi coreyjkelly,</p> <p>We've done this in an ad-hoc way in the past by calculating the equivalent binary sensor mask that contains the neighbouring points needed for interpolating the pressure values to the Cartesian points, and then post-processing. I won't say that it was entirely straightforward, but is certainly one approach.</p> <p>Another approach is to use nearest-neighbour interpolation (e.g., convert the Cartesian sensor mask to a binary mask using <code>cart2grid</code>). This approach obviously comes with the caveat you are recording the pressure values at a slightly different point to that intended.</p> <p>I'll give some thought to how this might be done more efficiently in the medium term. We have a new <code>kWaveArray</code> class under development (supported by both MATLAB and C++ codes), so it could possibly form part of this.</p> <p>Brad. </p> http://www.k-wave.org/forum/topic/define-sensor-by-cartesian-coordinates#post-6746 Tue, 05 Feb 2019 04:26:31 +0000 coreyjkelly 6746@http://www.k-wave.org/forum/ <p>Thanks so much for providing these tools to the community!</p> <p>I understand that with the compiled binaries, defining a sensor by an NxM matrix of coordinates hasn't been implemented. </p> <p>I won't go so far as to call this a "feature request" - rather, I'm wondering if there's a big hurdle to implementing this if I were to try it myself. I figure someone else must have considered it! </p> <p>I'm trying to model a large, curved transducer, so scaling up my simulation until each sensor is in a separate voxel would be very expensive computationally. </p>