Stradx main window

3D Ultrasound using Stradx 7.4g

We do not intend to release any further Stradx executables. All our new developments (eg. elastography) are being implemented in Stradwin, our next generation 3D ultrasound system which runs under both Linux and Windows. Please contact us if you have a particular need for Stradx, we will happily supply custom Linux executables.

Welcome to the documentation for Stradx version 7.4g. Stradx is a tool for the acquisition and visualisation of 3D ultrasound using a conventional 2D ultrasound machine. To record qualitative 3D ultrasound data with Stradx you need: If you want to make quantitative measurements from the 3D ultrasound data, you'll also need a Polhemus Fastrak or Patriot, Northern Digital Polaris or Ascension MiniBird, LaserBird or Flock of Birds position sensor. If you only want to visualise prerecorded data, you can run Stradx on any Linux PC.

Stradx is a research tool. It is not intended for commercial use. Any commercial use of Stradx potentially violates the intellectual property of a number of companies. Neither is Stradx intended for clinical use. You use it at your own risk. It may not work as documented and, in some cases, it may not work at all!

What can I do with Stradx?

With this version of Stradx, you can record image sequences (either greyscale or Doppler), or raw RF ultrasound signals, with matching position sensor information. The recordings can be made to memory or to disk. The position information can come from a Polhemus Fastrak or Patriot, Northern Digital Polaris or Ascension MiniBird, LaserBird or Flock of Birds position sensor.

If you do not want to make quantitative measurements from the data, you can dispense with the position sensor altogether. Stradx's proprietary image registration algorithms will estimate the relative locations of the B-scans by analysing the B-scan images themselves.

After recording, you can immediately reslice the data in any plane. The reslicing can even be performed as the data is being acquired. You can create extended field-of-view panoramic displays, similar in appearance to the Siemens Siescape facility. Non-planar reslicing, volume rendering and body-centered visualisation are also possible. You can project all these displays into a 3D window and then interactively adjust the viewpoint to obtain arbitrary views of the scanned anatomy.

You can manually or semi-automatically segment the 3D data into its constituent parts, and visualise the segmented data inside Stradx or using the Geomview 3D visualisation package. Stradx also provides two tools for calculating the volume of the segmented object, and special facilities for dealing with large structures that cannot be completely scanned with a single sweep of the ultrasound probe.

You can visualise and analyse structures moving under the influence of the cardiac pulse (eg. arteries). Stradx includes tools to `freeze frame' the anatomy at whatever phase of the cardiac cycle you like.

You can automatically calibrate the position sensing device using optimisation algorithms built into the Stradx system. Even without a specialised phantom, you can achieve average calibration accuracy. If you obtain our phantom, you will have access to one of the fastest and most accurate calibration tools currently available.

You can post-process the 3D data set to compensate for the effects of varying probe pressure and correct small errors in the position sensor readings, allowing crystal clear 3D reconstructions from even the highest resolution B-scans.

You can take the data recorded by Stradx, register it to any other data using fiducial points, and convert it into a voxel format for comparison, or for use with other visualisation packages such as 3DViewnix (available from the University of Pennsylvania). Also provided is the SelectSX utility, which you can use to edit pre-recorded Stradx data sets.

Click here for the previous release of Stradx (version 7.3), and here for details of the changes and new facilities in version 7.4.

Version 7.4g is almost identical to version 7.4 (there is a workaround for a V4L2 bug that would otherwise stop video acquisition working on recent Linux kernels).

Stradx 7.4g was written by Richard Prager, Andrew Gee and Graham Treece, who lead the Medical Imaging Group in the Machine Intelligence Laboratory at Cambridge University Engineering Department. The authors are grateful for the assistance of Petri Tuomola, Chris Wright, Nicholas Metcalfe and Susan Buchan, who all made valuable contributions to early prototypes of several Stradx modules.