Ultrasound Plane Wave Raw Data 75 Angles - Breast Phantom and Calibration Phantom Dataset

Summary

This dataset is a collection of raw ultrasound plane wave data from a breast mimicking phantom and a calibration phantom. The breast phantom data contains samples of hyperechoic lesions, hypoechoic lesions and no lesions. Whereas the calibration phantom contains samples to experimentally validate resolution and contrast. This dataset was acquired to do experimental validation of physics based deep learning for image registration, about which a paper is a current work in progress.

Description

Sample information

One the one hand this dataset contain 220 samples of in-vitro breast phantom ultrasound plane wave data (identified by CIRS073_RUMC). This data consists of three lesion types: hyperechoic lesions, hypoechoic lesions, and no lesions. On the other hand this dataset contains data from a calibration phantom (identified by CIRS040GSE). This data consists of: 5 samples of hypoechoic cysts, 5 samples from wire targets of 100 micrometer, 5 samples with -6dB and -3dB lesions in the field of view, and 5 samples with +3dB and +6dB in the field of view. All recordings were obtained twice: once in a low attenuating area 0.7 dB/cm/mHz,  and once in a high attenuating area 0.95 dB/cm/mHz . The calibration phantom used is: Multi-Purpose Multi-Tissue Ultrasound Phantom Model 040GSE (CIRS, Norfolk, USA).

Acquisition Information

Data was acquired using a Verasonics Vantage T256 R256 system with an L12-5 50 mm linear array ultrasound transducer operating at a center frequency of 7.8 MHz. Plane wave data has been acquired consisting of 75 steering angles with an angle range of -16 degrees to +16 degrees. Verasonics scripts used for the acquisition are provided along with the data.

Image Reconstruction

The GitHub repository contains code to subsample any amount of angles from the 75 acquired angles and to do image reconstruction with the f-k migration algorithm on this dataset. This processing code is available in Python.

Involved Parties

This dataset was produced by the Computational Imaging group at Centrum Wiskunde & Informatica (CI-CWI) in Amsterdam, The Netherlands in collaboration with the Medical Ultrasound Imaging Center in Radboud UMC Nijmegen, The Netherlands.

Contact Details

r [dot] schoop [at] nki [dot] nl