Acquisitions were initiated by the experiment computer. When instructed by the computer, the Arduino opened the camera shutter, waited 50 ms, and then sent a TTL pulse to the waveform generator to generate a 100 000-cycle, 86 ms pulse at 1.16 MHz and a 150 000-cycle, 67 ms pulse at 2.25 MHz. The 50 ms delay was set based on an observed 75–100 ms delay between when the Arduino commanded the shutter to open and when it actually opened, and the pulse durations were set long enough to accommodate some variability in the shutter delay while ensuring that photos were captured when the pressure field was at steady state. Using a shorter delay caused the shutter to sometimes open after the pulse had already shut off. The camera settings were: image size 4000 × 6000 pixels, International Organization of Standardization (ISO) 640, shutter speed 1/800 s (1.25 ms), and f-number f/5. The photographs were saved on the computer in the RAW image format. The shutter speed corresponded to 1450 cycles for the 1.16 MHz transducer and 2813 cycles for the 2.25 MHz transducer. During the experiments, the whole measurement setup was covered by a black cloth to suppress ambient light, therefore, the iPad (Apple Inc, Cupertino, CA) provided the only illumination.
The background images displayed by the iPad were bed-of-nails patterns comprising black dots/nails on a regular grid with a white background. The size of each dot was 2 pixels × 2 pixels. The distance between consecutive dots in each direction was 16 pixels, which corresponded to a physical distance of 1.7 mm and was set based on the maximum expected displacement in the experiments. To obtain a high-resolution projected pressure map, the background image was moved in intervals of 4 pixels in the x and z dimensions as illustrated in Fig. Fig.4,4, and a total of 16 photos were acquired, 1 for each image position. A single high-resolution (0.425 × 0.425 mm2) RMS projected pressure map was calculated from the 16 photos as described below.
To measure high-resolution RMS projected pressure maps, multiple CW-BOS photos were collected over a range of grid translations in the x and z directions. The reconstructed RMS projected pressure values were then tiled into the final map.
The following acquisitions were performed to evaluate the proposed CW-BOS RMS projected pressure beam mapping method:
The acquired photos were segmented into small rectangular patches around each dot using matlab's bwmorph() function (The MathWorks, Natick, MA), then the RMS projected pressure was calculated by the neural network for each dot as described below, and those values were tiled into the final reconstructed beam map. With the camera placed a total distance of 31 cm from the iPad (Apple Inc, Cupertino, CA), each rectangular patch comprised between 42 × 42 and 46 × 46 pixels, and was upsampled to 54 × 54 pixels for reconstruction. To avoid optical color dispersion, only the green channel from the photos was used for reconstruction, which has the largest weight among red, green, and blue channels in the Rec.ITU-R BT.601–7 (Ref. 54) video standard's luminescence calculation (). The total scan time for five averages was 3–5 min and was dominated by the time taken for the data transfer from the camera to the PC. Without these delays, the total scan times were approximately 8 s (16 photos × 100 ms of FUS on-time × 5 averages).
Do you have any questions about this protocol?
Post your question to gather feedback from the community. We will also invite the authors of this article to respond.
Tips for asking effective questions
+ Description
Write a detailed description. Include all information that will help others answer your question including experimental processes, conditions, and relevant images.