k-Wave User Forum » User Favorites: golubyatnikov

k-Wave User Forum » User Favorites: golubyatnikov_tv golubyatnikov_tv Support for the k-Wave MATLAB toolbox en-US Wed, 13 May 2026 02:15:12 +0000 http://bbpress.org/?v=1.0.2 <![CDATA[Search]]> q http://www.k-wave.org/forum/search.php Bradley Treeby on "Simulate a diagnostic ultrasound brain image" http://www.k-wave.org/forum/topic/simulate-a-diagnostic-ultrasound-brain-image#post-5480 Fri, 29 Apr 2016 10:42:48 +0000 Bradley Treeby 5480@http://www.k-wave.org/forum/ Hi Marlene, I'm not sure what you mean by "neural ultrasound probe"? I would suggest looking at the manufacturers specifications or in the literature to find the relevant properties. Brad. marlenep on "Simulate a diagnostic ultrasound brain image" http://www.k-wave.org/forum/topic/simulate-a-diagnostic-ultrasound-brain-image#post-5470 Fri, 15 Apr 2016 15:55:06 +0000 marlenep 5470@http://www.k-wave.org/forum/ Thank you for your suggestions, now the simulation is working but I need to adjust the transducer input signal in order to obtain more realistic results. Can you suggest me how to create a proper pulse as similar as possible to the one generated from neural ultrasound probes? I think I should use a frequency around 6MHZ but I don't know how to set the duration and the other parameters. Marlene Bradley Treeby on "Simulate a diagnostic ultrasound brain image" http://www.k-wave.org/forum/topic/simulate-a-diagnostic-ultrasound-brain-image#post-5420 Wed, 16 Mar 2016 16:20:02 +0000 Bradley Treeby 5420@http://www.k-wave.org/forum/ Hi Marlene, Did you also change the strength of the scattering in each region, i.e., the size of the variation introduced by adding the random scattering map to the ambient properties? In general, the features (speckle) that you see in the simulated B-mode image will depend on the number (or in this case strength) of the scatterers. Brad. marlenep on "Simulate a diagnostic ultrasound brain image" http://www.k-wave.org/forum/topic/simulate-a-diagnostic-ultrasound-brain-image#post-5393 Sun, 21 Feb 2016 13:49:44 +0000 marlenep 5393@http://www.k-wave.org/forum/ Thank you Brad for the help. I obtain something very similar to your examples of the B mode ultrasound with the ball volumes but I don't understand why if for example I try to invert the properties (speed of sound and density) of the propagation medium with the properties I have inside the ball volumes I obtain more or less the same image . I think I should obtain an image with inverted colors but it doesn't happen even if the scattering phantom that appears has inverted colors. Is it normal ? Bradley Treeby on "Simulate a diagnostic ultrasound brain image" http://www.k-wave.org/forum/topic/simulate-a-diagnostic-ultrasound-brain-image#post-5384 Tue, 16 Feb 2016 02:02:37 +0000 Bradley Treeby 5384@http://www.k-wave.org/forum/ Hi Marlene, (1) To create a realistic looking ultrasound image in k-Wave, you need to define some number of sub-wavelength scatterers. These will produce the speckle pattern characteristic of ultrasound images. There are many ways you could do this. In this k-Wave example, the approach is very simplistic - Gaussian random noise is used to perturb the mean sound speed and density values. It's certainly possible to use other distributions, or point scatterers (see e.g., <a href="http://scitation.aip.org/content/aip/proceeding/aipcp/10.1063/1.4934438">this paper</a>). The limits of 1400 and 1600 are used to restrict the values to physically relevant values in soft tissue. (2) You should use the values from the layer in which the ultrasound transducer is defined. (3) The time gain compensation is a simple exponential compensation of the form exp(-alpha * f * d), where alpha is the attenuation coefficient in dB/(MHz cm). Assuming y = 1, then alpha in this formula would be medium.alpha_coeff. Hope that helps, Brad. marlenep on "Simulate a diagnostic ultrasound brain image" http://www.k-wave.org/forum/topic/simulate-a-diagnostic-ultrasound-brain-image#post-5361 Sat, 16 Jan 2016 14:53:00 +0000 marlenep 5361@http://www.k-wave.org/forum/ Hi Brad, first of all thank you very much for the detailed answers . I finally managed,with your help, to see something meaningful in my simulation . I'm sorry to bother you again but I still have some doubts relative to the example " b -mode ultrasound image simulation" : 1 ) I do not understand this part of the code in which you defined a random distribution of scatterers for the region inside the ball . The value 75 is the new standard deviation ? Why c0 is bound between 1400 and 1600 ? % define a random distribution of scatterers for the highly scattering % region scattering_map = randn([Nx_tot, Ny_tot, Nz_tot]); scattering_c0 = c0 + 25 + 75*scattering_map; %???????? è media più deviazione standard di 75? scattering_c0(scattering_c0 > 1600) = 1600; scattering_c0(scattering_c0 < 1400) = 1400; scattering_rho0 = scattering_c0/1.5; 2) In my simulation I have several regions that simulate the different layers of the brain , each with its acoustic properties . In this case it is fair to attribute the properties of the outermost layer , (the closest to the translator ) to medium parameter c0 and rho ? 3)The absorption parameters correspond to modelling power law absorption of the form alpha = alpha_0*f^y, where alpha_0= medium.alpha_coeff y =medium.alpha_power tgc_alpha is supposed to be the alpha of the formula? thank you very much Bradley Treeby on "Simulate a diagnostic ultrasound brain image" http://www.k-wave.org/forum/topic/simulate-a-diagnostic-ultrasound-brain-image#post-5357 Wed, 13 Jan 2016 16:35:44 +0000 Bradley Treeby 5357@http://www.k-wave.org/forum/ Hi Marlene, (1) For defining the source, you should use the value of the sound speed and density at the position of the ultrasound transducer. This step is necessary as in the example the input signal is defined in units of pressure, whereas the input must defined in units of velocity. (2) For the post-processing and visualisation, did you try using the code included in the example? (3) For time gain compensation, the method used is very simple, and assumes a single value for attenuation. More advances methods are certainly possible (I'll leave it to you to search the literature), but these are not included in the toolbox. Brad. marlenep on "Simulate a diagnostic ultrasound brain image" http://www.k-wave.org/forum/topic/simulate-a-diagnostic-ultrasound-brain-image#post-5345 Sat, 12 Dec 2015 14:43:06 +0000 marlenep 5345@http://www.k-wave.org/forum/ Hi everybody! I would like to ask you some questions about the simualtion of ultrasound B-mode images starting from 'Simulating B-mode Ultrasound Images Example". I have tried to reproduce it using only one ball . In my example the ball demarcates a region containing Cerebral Spinal Fluid so I define the typical density and speed of sound inside that region equal to --> density:1007 and speed of sound:1504.5; Outside the ball I imagine having the properties of brain tissue so: speed of sound:1546.3 and Density :1046. as regards the remaining part of the code I tried to follow the example provided. My questions are : 1 ) I am not sure of having correctly selected all the parameters , in particular, I have a matrix that defines me the speed of sound in the medium and one that defines the density but when I create the input signal is required the use of a single value. In my example I used the maximum speed of sound value and the minimum density value inside the matrices but I'm not sure if it is correct . this line : input_signal = (source_strength./(1546.3*1007)).*input_signal; 2 ) I do not know how to properly plot the resulting image . I tried to download the "scan_lines matrix " provided at the start of the example to figure out how to get the final image of the three balls but I do not get anything recognizable , i suppose something is wrong in my postprocessing. 3 ) the value of the attenuation coefficient alpha should be different between the two regions (inside and outside the ball ) or it should be a scalar , for example a mean value? I put my code here : Thank you very much for any help and suggestion. clear all clc close all % simulation settings DATA_CAST = 'single'; % create the computational grid Nx = 100; Ny = 20; Nz=40; % number of grid points in the y (column) direction dx = 0.1e-3; % grid point spacing in the x direction [m] dy = 0.1e-3; % grid point spacing in the y direction [m] dz=0.1e-3; kgrid = makeGrid(Nx, dx, Ny,dy,Nz,dz); object=makeBall(Nx, 100, Nz, Nx/2, Ny/2, Nz/2, 10, true); medium.sound_speed = 1546.3*ones(Nx,100,Nz); % [m/s] medium.density = 1046*ones(Nx,100,Nz); % [kg/m^3] medium.sound_speed(object==1)=1504.5; %inside the ball medium.density(object==1) =1007; % define properties of the input signal source_strength = 1e6; % [Pa] tone_burst_freq = 0.5e6; % [Hz] tone_burst_cycles = 5; % create the time array % kgrid.t_array = makeTime(kgrid, medium.sound_speed, [], t_end); [kgrid.t_array, dt] = makeTime(kgrid, medium.sound_speed); % create the input signal using toneBurst input_signal = toneBurst(1/kgrid.dt, tone_burst_freq, tone_burst_cycles); % scale the source magnitude by the source_strength divided by the % impedance (the source is assigned to the particle velocity) input_signal = (source_strength./(1546.3*1007)).*input_signal; plot(input_signal); % append input signal used to drive the transducer transducer.input_signal = input_signal; % physical properties of the transducer transducer.number_elements =19; % total number of transducer elements transducer.element_width = 1; % width of each element [grid points] transducer.element_length = 8; % length of each element [grid points] transducer.element_spacing = 0; % spacing (kerf width) between the elements [grid points] transducer.radius = inf; % radius of curvature of the transducer [m] % calculate the width of the transducer in grid points transducer_width = transducer.number_elements*transducer.element_width ... + (transducer.number_elements - 1)*transducer.element_spacing; % use this to position the transducer in the middle of the computational grid transducer.position = round([1, Ny/2 - transducer_width/2, Nz/2 - transducer.element_length/2]); % properties used to derive the beamforming delays % properties used to derive the beamforming delays transducer.sound_speed = 1504.5; % sound speed [m/s]%WHICH VALUE OF THE MATRIX SHOULD I PUT? transducer.focus_distance = Nx/4; % focus distance [m] transducer.elevation_focus_distance = 19e-3;% focus distance in the elevation plane [m] transducer.steering_angle = 0; % steering angle [degrees] transducer.transmit_apodization = 'Rectangular'; transducer.receive_apodization = 'Rectangular'; % define the transducer elements that are currently active transducer.active_elements = zeros(transducer.number_elements, 1); transducer.active_elements(1:19) = 1; % % append input signal used to drive the transducer transducer.input_signal = input_signal; % create the transducer using the defined settings transducer = makeTransducer(kgrid, transducer); transducer.plot; transducer.properties; t=medium.sound_speed; u= medium.density; number_scan_lines=79; medium_position=[1, Ny/2 - transducer_width/2, Nz/2 - transducer.element_length/2]; % run the simulation for scan_line_index = 1:number_scan_lines % update the command line status disp(''); disp(['Computing scan line ' num2str(scan_line_index) ' of ' num2str(number_scan_lines)]); % load the current section of the medium medium.sound_speed = t(:, medium_position(2):medium_position(2)+ Ny-1 , :); medium.density= u(:, medium_position(2):medium_position(2) + Ny-1 , :); % run the simulation sensor_data = kspaceFirstOrder3D(kgrid, medium, transducer, transducer)%,input_args{:}); % extract the scan line from the sensor data scan_lines(scan_line_index, :) = transducer.scan_line(sensor_data); % update medium position medium_position = medium_position + [0 transducer.element_width 0]; end c0=1540 ; %WHICH VALUE SHOULD I PUT? % create radius variable assuming that t0 corresponds to the middle of the % input signal t0 = length(input_signal)*kgrid.dt/2; r = c0*( (1:length(kgrid.t_array))*kgrid.dt/2 - t0); % [m] % create time gain compensation function based on attenuation value, % transmit frequency, and round trip distance tgc_alpha = 0.6; % [dB/(MHz cm)] %IT's THE ATTENUATION OUTSIDE THE BALL, HOW CAN I TAKE IN ACCOUNT ALSO THE %ALPHA OF THE INSIDE MEDIUM? tgc = exp(2*tgc_alpha*tone_burst_freq/1e6*r*100); % apply the time gain compensation to each of the scan lines scan_lines = bsxfun(@times, tgc, scan_lines); % filter the scan lines using both the transmit frequency and the second harmonic scan_lines_fund = gaussianFilter(scan_lines, 1/kgrid.dt, tone_burst_freq, 100, true); scan_lines_harm = gaussianFilter(scan_lines, 1/kgrid.dt, 2*tone_burst_freq, 30, true); % envelope detection scan_lines_fund = envelopeDetection(scan_lines_fund); scan_lines_harm = envelopeDetection(scan_lines_harm); % normalised log compression compression_ratio = 3; scan_lines_fund = logCompression(scan_lines_fund, compression_ratio, true); scan_lines_harm = logCompression(scan_lines_harm, compression_ratio, true); % upsample the image using linear interpolation scale_factor = 2; scan_lines_fund = interp2(1:kgrid.Nt, (1:number_scan_lines).', scan_lines_fund, 1:kgrid.Nt, (1:1/scale_factor:number_scan_lines).'); scan_lines_harm = interp2(1:kgrid.Nt, (1:number_scan_lines).', scan_lines_harm, 1:kgrid.Nt, (1:1/scale_factor:number_scan_lines).'); bencox on "What is the cause of the calculation error without using filterTimeSeries" http://www.k-wave.org/forum/topic/what-is-the-cause-of-the-calculation-error-without-using-filtertimeseries#post-5080 Sun, 10 May 2015 20:32:04 +0000 bencox 5080@http://www.k-wave.org/forum/ Hi Tim, An even simpler example will show this error: if you send the toneburst from a point source to a sensor some distance away you will see it. It is an instance where you can see the effect of the underlying numerical scheme that k-Wave uses. You are seeing the effect of the bandlimited interpolant inherent in the k-space pseudospectral scheme. See Fig. 2 in <a href="http://www.k-wave.org/papers/2011-Treeby-JASA.pdf">this paper</a> and the accompanying text for an explanation. There are two ways to ameliorate it: ramp up the signal (as you have done using filterTimeSeries) or define a larger source region with the source amplitudes weighted so that it is tapered towards the edges, eg. use a Gaussian as a source rather than a point. Hope that helps. Ben golubyatnikov_tv on "What is the cause of the calculation error without using filterTimeSeries" http://www.k-wave.org/forum/topic/what-is-the-cause-of-the-calculation-error-without-using-filtertimeseries#post-5075 Fri, 08 May 2015 22:30:34 +0000 golubyatnikov_tv 5075@http://www.k-wave.org/forum/ Hello, my name is Tim. Sorry for my English. I try to simulate the acoustic wave propogation with the signal frequency 4kHz. When I use filterTimeSeries function, I get good signal on the receiver. But without using this function I see high-frequency calculation error. Here is example image: <img src="http://imageshack.com/a/img903/8317/hZqxFI.png" /> While the simulation was running, I saw: "... dt: 4.5045us, t_end: 5ms, time steps: 1111 input grid size: 200 by 200 grid points (1 by 1m) maximum supported frequency: 33.3kHz expanding computational grid... computational grid size: 240 by 240 grid points ". The maximum supported frequency of grid (33.3kHz) is much larger than the signal frequency (4kHz). I can't understand what is the cause of this calculation error? I have not found the explanations in the documentation and on the forum. Here is simple code to reproduce this situation: <pre><code>% area settings gridSize = [200, 200]; cellSize = [0.005,0.005]; % [m] medium.sound_speed = 333; % [m/s] ts = [50, 100]; % tranceiver positions amplitude = 10; % [Pa] freq = 4000; emitTime = 0.001; % [s] modelingTime = 0.005; % [s] source.p_mask = zeros(gridSize); source.p_mask(ts(1),ts(2)) = 1; sensor.mask = source.p_mask; circleReflectorRadius = 12; circleReflectorPosition = [150,100]; kgrid = makeGrid(gridSize(1), cellSize(1), gridSize(2), cellSize(2)); [kgrid.t_array, dt] = makeTime(kgrid, medium.sound_speed, [], modelingTime); source.p = amplitude * toneBurst(1/dt, freq, ceil(freq*emitTime), 'Envelope', 'Rectangular', 'SignalLength', kgrid.Nt); %source.p = filterTimeSeries(kgrid, medium, source.p, 'ZeroPhase', true); disk = makeDisc(gridSize(1), gridSize(2), circleReflectorPosition(1), circleReflectorPosition(2), circleReflectorRadius); medium.density = ones(gridSize) + disk*100; % define the acoustic parameters to record sensor.record = {'p'}; %% prepare display parameters densityMask = medium.density - 1; dm = source.p_mask + sensor.mask + densityMask; dm(dm>0) = 1; simulationParameters = {'DisplayMask', dm, 'PMLInside', false, 'PlotPML', false}; % run the simulation sensor_data = kspaceFirstOrder2D(kgrid, medium, source, sensor, simulationParameters{:}); plot(kgrid.t_array, sensor_data.p);</code></pre> Thank you.