Safety and Accuracy

This section refers to the use of Stradwin for 3D ultrasound.

Stradwin is a powerful tool. It is based on the Stradx system that was first released in 1997 and was shown in 2003 to be the highest definition freehand three-dimensional ultrasound system in the literature. Like all powerful tools, its safety and performance is dependent on correct usage. This page lists some of the issues you need to be aware of in order to ensure that the software produces results of sufficient accuracy for your application and is used as part of a safe system.

Safety

Here is a list of some of the safety issues that you should consider before using Stradwin.

• The computer that you run Stradwin on must be safe and must not interfere inappropriately with any other equipment in the environment.
• Any position sensing device you use must be safe and must not interfere inappropriately with any other equipment in the environment.
• The ultrasound machine that you use must be safe and must not interfere inappropriately with any other equipment in the environment. Although ultrasound is generally considered safer than some other imaging modalities, it does involve risks. Before using any medical imaging equipment you should be appropriately qualified so that you do so in an ethical and safe way in the light of the risks involved.
• Any electrical connections between the equipment you use should be made by an appropriately qualified person who is able to ensure they are safe.
• It is dangerous to try and deduce information from ultrasound images unless you have appropriate qualifications and experience. Furthermore, the interpretation of some of the derived images produced by Stradwin requires particular specialist knowledge over and above the knowledge required for the interpretation of conventional B-scan images. For example, ultrasound artifacts appear in a different form in some derived images.

Configuration

Here is a list of some of the things that need to be considered when setting up Stradwin.

• The fixed part of the position sensor should be permanently and rigidly mounted so it cannot be moved either deliberately or by accident.
• Isocentre calibration should be performed as described in the documentation.
• Pointer calibration should be performed as described in the documentation.
• Position sensor mounts should be rigidly attached to all the probes that are to be used with the system. These mounts must be rigid and stable. They must be tamper-proof to avoid users moving them. You will also wish to ensure that they do not void any warranties or maintenance contracts associated with the ultrasound machine.
• Make sure that the mobile part of the position sensor is attached correctly to the mount on the probe. If the users are to be allowed to move the sensor onto different probes, they need to be taught to ensure that the sensor is mounted with the correct orientation (i.e. the same orientation as when the probe calibration was performed). If the sensor is mounted wrongly the calibration will be invalid.
• All the probes to be used should be calibrated at all the required depth settings as described in the documentation. The resulting template files should be saved somewhere accessible to the intended users.
• All probe calibrations should be tested for consistency with the pointer calibration, as described in the documentation for the 'Probe Calib' task page.
• Users should be carefully trained never to invert the ultrasound image, flip it left-to-right or touch the pan and zoom controls. They should always load the appropriate template file for the probe and depth setting in use.
• Phantom experiments should be performed to check the performance of the complete system.
• Regularly check that the calibrations of the isocentre, pointer and all the probes at the various depth settings are still valid.

User issues

Here is a list of some of the things that need to be considered when using Stradwin.

• Do not use the system unless it has been properly set up and configured by an appropriately qualified person.
• Do not use the system unless you have been appropriately trained.
• If the patient moves during the scan or between the scan and the radiotherapy treatment, this will seriously compromise the usefulness of the data. One way to minimise breathing artifacts is not to mention breathing to the patient and let them just breath gently and normally. If a patient is trying consciously to breathe in a particular way, this tends to accentuate rather than reduce the motion artifacts.
• Always load the correct template corresponding to the probe and depth setting you are using. If a template does not exist for a particular combination of probe and depth setting, you should not use this combination for 3D scans until a template has been prepared.
• Never adjust the way the position sensor is mounted on the probe. If the mounting is changed, the probe calibration will be invalid.
• Never change the depth setting without loading the appropriate template file.
• Never use the pan or zoom controls on the ultrasound machine.
• Never flip the B-scan image up-down or left-right.
• Periodically check the consistency of the probe calibrations with the pointer, as described in the documentation for the 'Probe Calib' task page. If the calibrations are inconsistent do not use the system.

Accuracy

Here is a list of some of the sources of error you should consider when assessing the accuracy of any measurements made using Stradwin.

• Errors caused by the movement of the scanned material.
• Differences between the speed of ultrasound in the object scanned and the speed assumed by Stradwin and the ultrasound machine you use.
• The size of the resolution cell (or spatial sensitivity function) of the overall imaging system.
• Distortion of the tissue due to pressure from the probe during the scan.
• Errors generated by the algorithms in Stradwin which attempt to correct for probe pressure distortion.
• Reverberation, shadowing and other ultrasound artifacts.
• The inherent accuracy of the position sensor you use.
• Local disturbances in the performance of the position sensor, for example optical or magnetic disturbances.
• Distortion introduced in the B-scan when it is transferred as an analogue video signal.
• Errors in the time-stamping of the image and position streams.
• Temporal calibration errors - errors in the constants that determine the relative latencies of the position sensor and image data streams.
• Spatial calibration of the transformation between the mobile part of the position sensor and the location of the B-scan in three-dimensional space.
• Interpolation errors.
• Segmentation errors.
• Errors in the way landmarks are defined either on the data or using the pointer.
• Interpretive errors.
• Errors caused by delays, persistence or image averaging inside the ultrasound scanner.
• Errors caused by the approximate temperature compensation used in the calibration of high curvature probes.