QUPS is intended to be an abstract, lightweight, readable tool for simulation and processing of pulse-echo ultrasound data to support prototyping. It provides a flexible, high-level representation of transducers, scattering media, and accelerated implementations of common signal processing functions. Currently, QUPS can use multiple simulation tools as a backend including k-Wave, MUST, and FieldII and can read/write to a USTB format.
The easiest way to get started is to open and run example.mlx or example_.m and interact with the simulation and beamforming examples. There are plenty of comments highlighting the different methods supported by the interface. You will need to separately download the simulator packages that you wish to use. Don't forget to add them to your path!
QUPS targets MATLAB R2020b and later on linux. While it may work for older versions of MATLAB, you may get strange errors that don't appear in later versions. QUPS does minimal error checking for compatibility in order to maintain flexibility.
If you have trouble, please submit an issue.
QUPS is (partially) internally documented following MATLAB conventions. This means you can use doc on any class and help on any class or method with help classname/methodname or help classname.methodname.
QUPS utilizes flexible abstract objects and definitions in order to preserve code re-usability. The following base classes are used to provide the majority of the functionality:
| Class | Main Properties | Main Methods |
|---|---|---|
Transducer |
numel | positions() |
Sequence |
numPulse | delays(), apodization() |
Scan |
size | getImagingGrid() |
ChannelData |
data, t0, fs | sample() |
Target |
c0, rho0 | getPropertyMap() |
All of these classes provide an overloaded plot or imagesc method for easy visualization.
A synthesis class UltrasoundSystem is provided to combine the Transducer, Scan and Sequence classes, simulate ChannelData from a Target, or beamform ChannelData into a b-mode image, which is just a regular MATLAB array ;) - you can use Scan/imagesc to display it.
QUPS objects, classes, and functions adhere to some conventions to make prototyping easier.
| Measure | Unit |
|---|---|
| Position | Meters |
| Time | Seconds |
| Angle | Degrees |
| Property | Standard |
|---|---|
| Position | {x,y,z} in dimension 1 |
| rf-data | time x receive x transmit |
| Sequence Type | Definition |
|---|---|
| Full-Synthetic Aperture (FSA) | Peak is centered on the firing element |
| Plane Wave (PW) | Wavefront centered on the origin (0,0,0) |
| Virtual Source (VS) | Peak is centered on the focus |
Note that this transmit sequence definition is likely to result in data with varying start times across multiple transmits and a negative start time i.e. before time 0. This is explicitly supported! Simply create a definition of t0 of size 1 x 1 x M where M is the number of transmits that aligns the data.
Utilities are provided to assist in writing readable, broadcasting code. The sub utility allows you to conveniently slice one dimension while the utility swapdim as well as the built-in shiftdim and permute functions are useful for placing data in broadcasting dimensions.
Dimensions are used to implicitly broadcast operations, allowing you to limit memory and computational demand for simple workflows. For example, the apodization argument for the DAS method takes in an argument of size I1 x I2 x I3 x N x M where {I1,I2,I3} is the size of the scan, N is the number of receivers, and M is the number of transmits. This can be a huge array, easily over 100GB! However, in many cases we may only need 2 or 3 of these dimensions.
For example, to evaluate a multi-monostatic configuration given a set of FSA data, we simply create an array of size 1 x 1 x 1 x N x M and apply identity matrix weights.
apod = shiftdim( (1:N)' == (1:M), -3); % we can leave this as a binary type!