Data for Dual-Probe Transcranial Full-waveform Inversion: a Brain Phantom Feasibility Study

This dataset contains ultrasound computed tomography (USCT) data for the article "Dual-Probe Transcranial Full-waveform Inversion: a Brain Phantom Feasibility Study" (Robins et al., 2023). This study investigated the application of full-waveform inversion (FWI) for performing transcranial brain imaging by applying FWI to reconstruct a brain-mimicking phantom. Imaging datasets were acquired using a custom dual-probe acquisition system which consisted of a large water tank (0.7 x 0.6 x 0.6 m^3) and a pair of P4-1 cardiac probes rotated about an imaging target to form a ring array (see Experimental_Imaging_Setup.png). All datasets were acquired using the same imaging sequence which involved a ring array of 16 probe positions (see Dual_Probe_Ring_Array.png). Estimated sets of transducer element positions (elementList.txt) and source signals (signal_filtered.mat, seen plotted in Source_Signal_Filtered.png) have been optimised for running FWI to image the phantom. All transducer positions and source signals were calibrated using a watershot dataset, which was acquired by performing the scan without the phantom being present.

Paper DOI: https://doi.org/10.1016/j.ultrasmedbio.2023.06.001

Imaging sequence:
o Imaging was performed by a pair of P4-1 cardiac probes (ATL, 96 elements per array) .
o The ring array consisted of 3072 element positions (1536 over 16 Probe 01 positions, 1536 over 16 Probe 02 positions).
o 384 source elements.
o Sampling frequency (Fs) = 22.727 MHz.
o [Number of time samples (Nt), No. of Traces] = [3328, 442368].
o Source-receiver pairings for all 442368 traces is given by elementRelas.txt.

These imaging experiments include:
(1) 01_Experimental_USCT_Brain.mat:
o USCT data from imaging the 2.5D PVA brain phantom.

(2) 02_Synthetic_USCT_Brain.mat:
o USCT data from synthetic imaging the 2.5D numerical brain phantom.
o The true numerical brain phantom for this imaging problem is given in TrueVp.mat.

(3) 03_Experimental_USCT_WaterShot.mat:
o USCT data from imaging just water.

Additional files include:
elementList.txt: Calibrated transducer element positions, [Element ID, x,y,z - positions]
elementRelas.txt: The source-receiver transducer element IDs for all traces [Trace ID, Source ID, Receiver ID]
signal_filtered.mat: 384 source wavelets for all source elements (filtered by *)
TrueVp.mat: True velocity model of the numerical brain phantom (dx = 0.186 mm, dimensions = [957, 958])

* To filter USCT data 'dataset' to match the source signals, apply the following matlab solution (or equivalent):
Fs = 22.727e06;
N = 100;
F01 = 0.4e06;
F02 = 2.0e06;
[Hd a] = fir1(N,[F01, F02]/(Fs/2));
dataset_filtered = filtfilt(Hd,a,dataset);