3.21 void-detectability ratio VDR number characterizing the visibility of an image area corresponding to a void of defined diameter surrounded by tissue-mimicking material TMM in the
General phantom specification
The phantom enables the execution of test procedures by offering anechoic targets positioned at specific locations within tissue-mimicking material Image analysis is performed using digitized data obtained during phantom scans Additionally, the manufacturer will supply an instruction manual containing guidance on proper usage and maintenance.
TMM specifications
The following parameters of the TMM shall lie within the specified limits:
Speed of sound: (1 540 ± 10) m s –1 at 3 MHz
The specific attenuation coefficient ranges from (0.7 + 0.2 / - 0.05) dB cm –1 MHz –1 within the frequency range of 1 MHz to 15 MHz When creating a phantom using layered materials, as detailed in Annex A, the mean value of the specific attenuation coefficient should be utilized.
Backscatter coefficient: (3 × 10 –4 cm –1 sr –1 ) ± 10 dB at 3 MHz; with a “frequency to the n”
The backscatter coefficient of the phantom must be reported as a function of frequency within the range of 1 MHz to 15 MHz, specifically for values of \( f(n) \) where \( 2 < n < 4 \) Additionally, the phantom should contain a sufficient number density of scatterers to ensure that the echo-amplitude distribution adheres to Rayleigh statistics.
The required scatterer number density varies based on the frequency and focusing characteristics of the transducer and ultrasound system being evaluated Generally, a density of approximately 10 scatterers per cubic millimeter is adequate for most transducers functioning at frequencies up to 15 MHz.
Phantoms built to TMM specifications can utilize materials such as open pore sponges or polyurethane foam soaked in saline These materials feature uniformly distributed microscopic inhomogeneities, which help achieve the required attenuation level.
Anechoic targets
Anechoic targets must have a backscatter contrast of at least -60 dB compared to the background TMM material A suitable filling material for these targets is degassed saline solution, which should be adjusted in concentration to achieve a sound speed of (1,540 ± 10) m/s The relationship between sound speeds in saltwater, saline concentration, and temperature is illustrated in Figure A.2.3.
Anechoic targets must be strategically placed at various depths within the phantom volume, ensuring that targets of the same diameter have their centers aligned in a coplanar manner This arrangement allows for at least six targets to be visible at different lateral positions from the transducer at each depth Additionally, the targets should be positioned laterally to ensure visibility from multiple locations within the scanning plane The targets, as detailed in Annex A, are cylindrical in shape with faces that are parallel to the scanning surface.
A variety of anechoic targets will be available at different depths and lateral locations, with two sizes of voids present for each frequency region The dimensions of these voids will be chosen based on the realistic azimuthal and elevational beam width, as well as the frequency of the transducer.
– Voids of 4 mm and 2,5 mm diameter are satisfactory for transducers operating in the
– Voids of 3 mm and 1,5 mm diameter are satisfactory for transducers operating in the
– Voids of 2,5 mm and 1 mm diameter are satisfactory for transducers operating in the
Artifactual signals in anechoic voids are primarily caused by the side-lobes of the ultrasonic beam To effectively detect these artifactual signals, which may arise from side-lobes or grating-lobes, it is essential to achieve an echo-amplitude difference of better than -60 dB when compared to the surrounding tissue-mimicking material.
Phantom enclosure
The enclosure is designed to safeguard its contents from degradation, particularly fluid evaporation, over time during both use and storage To achieve this, the walls of the enclosure will be constructed from materials that effectively prevent any deterioration of the contents.
Scanning surface
The scanning surface must ensure complete acoustic contact between the transducer's active surface and the phantom If a window material, like a foil or membrane, is used to prevent the TMM from drying out or to protect its contents from transducer damage, detailed specifications of the membrane's properties—such as material composition, thickness, density, and specific attenuation coefficient—must be provided Alternatively, information on transmission losses related to frequency should be included.
Dimensions
The phantom's dimensions must be appropriate for evaluating transducers by ensuring that the volume of diagnostic relevance (VDR) covers at least two-thirds of the imaged field typically used by the transducer For instance, an ultrasound system with a 24 cm imaging depth necessitates a phantom with a minimum depth of 16 cm This dimension is generally at least four times larger than the transducer's transmit-receive aperture, allowing for the detection of any degradation due to side lobes or inadequate lateral resolution within this range.
Phantom stability
The manufacturer shall state the duration of stability and indicate criterion of use.
Digitized image data
The technical specification outlines test and analysis methods for digitized image data obtained from the evaluated ultrasound system and its transducer It is essential that grey level values for every spatial location in the image are accessible Typically, the image data is organized in a matrix of approximately 300 × 300 pixels, with a minimum amplitude resolution of 8 bits (256 levels), which is directly proportional to the grey-scale.
Digitized image data can be captured using a video frame grabber, which digitizes images from output connectors typically used for analog monitors or recording devices To prevent or reduce signal distortions, the video signal digitization must occur under precisely defined conditions, with particular focus on key parameters.
– The input dynamic range of the video-frame grabber shall be adjusted according to the maximum signal amplitude of the video output
The digitizing amplitude resolution, determined by the pixel byte size, must exceed the grey-scale resolution of the ultrasound image video output A minimum of 8 bits, allowing for 256 grey levels, is essential for optimal performance.
Ensuring the linearity of the conversion function from ultrasound image signals to TV is crucial Additionally, the spatial resolution, determined by the voxel size, must exceed the original video line density of the image for optimal quality.
To ensure effective data acquisition when the imaged field is in motion, the video-capture frame rate of the video-frame grabber must be sufficiently high to match the input data rates It is important to distinguish between the scanning frame rate and the output video frame rate.
– A cable matched for input/output impedance has to be used to avoid reflections in the line A cable length of 1 m to 2 m is generally not critical
The digitized image data from the diagnostic ultrasound system must accurately reflect what is displayed on the monitor Additionally, this data should not be altered through post-processing between the data processing stage and the monitor output signal prior to analysis, as outlined in this technical specification.
Data also can be acquired using DICOM-images (Digital Imaging and Communications in
Most ultrasound system manufacturers utilize a method for in-house quality-control testing and image-processing development, which involves the use of images in various formats from the ultrasound system Additionally, this method can often be extended for use by clinical personnel through resources such as file-transfer-protocol (ftp).
Many diagnostic ultrasound systems offer image files on removable media like USB drives, magneto-optical disks, zip disks, or CD-ROMs, making them suitable sources for digital image data.
In addition, many imaging centres use commercially available Picture Archiving and
Communication Systems (PACS) for viewing and storing ultrasound-image data
PACS system manufacturers typically offer the capability to acquire images in uncompressed formats, like TIFF (Tagged Image File Format) or DICOM, for workstations with the appropriate access rights to the image data.
When image data is obtained from within the ultrasound system rather than from the video output, it is essential to verify that the digitized amplitude accurately reflects the grey levels displayed on the screen.
Until DICOM offers a standard for 3D-images the best procedure is to use a VGA or DVI converter to digitize the video output signal of the ultrasound system
7 Principle of measurement using the 3D anechoic void phantom
General
The VDR measurement setup includes a phantom, a transducer, and an ultrasound system for testing, along with a method to capture digitized image data from scan planes that cover the volume of the phantom containing voids.
To obtain accurate 3D data, utilize the mechanical transducer positioner outlined in Annex A and capture digitized image data from closely spaced scan planes The spacing between these scan planes should match the voxel separation, but it must be less than one-fourth of the diameter of the void from which the VDR will be calculated.
Alternative methods for 3D data acquisition involve utilizing specialized transducers equipped with mechanical translation capabilities or manually moving the transducer while capturing image data in a cine loop within the ultrasound system However, the manual approach may compromise the consistency of B-plane spacing.
A uniformity measurement of transducers is essential It should be done prior to VDR measurement One possible method is described in Annex D.
Analysis
The following is an acceptable method of analysis:
The VDR is calculated for anechoic targets, specifically voids, defined by their diameter, depth, and lateral position in planes that are parallel to the accessible C-planes, as outlined in Annex A For this technical specification, it is essential that the results are reported specifically for C-planes that include voids.
Within images reconstructed from the acquired 3D data a region of interest (ROI) is defined
The region's data will consist of cubic voxels, ensuring equal dimensions in all three directions To accomplish this, it is essential to obtain information regarding the image scale The 3D-ROI data is organized in a matrix format, facilitating the viewing and processing of the 3D dataset.
In the analysis of each C-plane within the 3D-ROI, the mean (\$à_1\$) and standard deviation (\$\sigma_1\$) values for the TMM are computed Various methods have been proposed to exclude void regions from these calculations, with some leveraging knowledge of void distribution while others rely on local grey-scale comparisons among voxel amplitudes within the C-plane to determine voxel inclusion Detailed descriptions of these approaches for excluding void regions from TMM mean and standard deviation calculations can be found in Annex A.
The à 1 and σ 1 values are used to calculate VDR i for each individual voxel according to the expression
The VDR i values are stored in a matrix similar to the matrix containing the 3D-ROI grey-scale information-set to allow direct viewing and further processing
The evaluation of VDR i values is conducted for each C-plane using various methods One approach involves deriving statistical data directly from the entire C-plane, while another method calculates VDR data for each void individually before consolidating this information into a single dataset for the plane Statistical analysis can focus on either the mean or maximum VDR values.
Maximum values provide a practical method for assessing void visibility, as the grey-scale distribution within each void significantly decreases towards its center Observations indicate a strong correlation between the visibility of the void and this maximum value.
Obtaining mean values is challenging because it requires defining the boundaries for each void, which can be achieved by setting an amplitude limit or utilizing supplementary information regarding the position of the individual voids.
NOTE 2 This determination of the void edges is the main source of error for the mean evaluation
The mean and maximum VDR values for all C-planes are illustrated in a graph that depicts VDR as a function of depth Statistical analysis can be conducted on this data by applying a fitted curve.
A void of a given diameter and location shall be called detectable if its value of VDR exceeds
2,5 The range where the VDR-plot exceeds this value is the useful working range for the void diameters contained in this ROI [4]
The stored ROI grey-scale and VDR i 3D data sets are utilized to visually verify the automated evaluation of the transducer Additionally, VDR i visualization of non-void regions is crucial, as it highlights areas where the interference pattern may falsely suggest the presence of voids.
NOTE 3 The VDRi-values on their own do not represent relevant information, but in the context of all the other
VDRi values are essential for generating a 3D image of VDR levels within an image of a void, C-plane, or region of interest (ROI) They provide crucial information, including the mean VDR values within a single void or multiple voids, as well as the maximum VDR values observed in these images.
Description of construction of an example phantom and test results
In this annex the construction and properties of a particular phantom satisfying the technical specification are described, together with examples of tests performed with it
The VDR measurement requires specific equipment, including a phantom, a transducer positioning slider with a platform, and a PC or notebook equipped with software for image recording and analysis Additionally, hardware is necessary to connect the video output of the diagnostic ultrasound system to the positioning slider.
Figure A.1 – Example of measurement test equipment
This arrangement is suitable for testing linear array transducers as well curved arrays under certain conditions (see Annex D)
A.1.1.1 data acquisition system centralized system receiving data from one or more remote points
NOTE 1 Data may be transported In either analog or digital form and digitised thereafter if necessary
NOTE 2 In the present case data are acquired via a frame grabber or converter from the available output signals of a diagnostic ultrasound system
A.1.1.2 slice layer of attenuating and backscattering TMM
Image digitizer or converter Phantom
Sensor for transducer position feedback to the computer
A.1.1.2.1 attenuation slice layer of TMM having an attenuation coefficient greater than the average attenuation coefficient of an assembled phantom
A.1.1.2.2 void slice layer of TMM containing voids and having an attenuation coefficient less than the average attenuation coefficient of an assembled phantom
The TCC 2) is a 3D artificial anechoic cyst phantom designed with alternating slices of defined attenuation and void materials These slices are primarily oriented perpendicular to the sound propagation direction and are filled with an anechoic liquid This configuration ensures that the speed of sound, along with the mean attenuation and backscatter coefficients, closely mimics those of human soft tissue.
The example phantom is housed in a tight plastic box with external dimensions: height 22 cm × length 15 cm × width 8 cm
The phantom body is made up of alternating 5 mm thick layers of polyurethane foam, consisting of both attenuation slices and void slices The void slices, present in every second layer, feature artificial cylindrical voids that are integrated into the foam structure.
The foam and voids are saturated with degassed saline water at a concentration of 7% by weight This saline concentration is specifically calibrated to achieve a sound speed of 1,540 ± 10 m/s in the soaked foam at a temperature of 20 °C.
It was found that the backscattering level for both foams immersed in saline was the same
NOTE Polyurethane foam as recommended is available from different manufacturers worldwide The material is very stable
The packet of slices measures about 18 cm in height After filling the phantom with saline, it is essential to evacuate any trapped air bubbles The phantom is entirely sealed with a 0.25 mm polyurethane foil over the initial void slice, which serves as a coupling window with an area of 11 × 5.5 cm², exhibiting negligible attenuation.
2) Tissue Characterization Consulting, A-4850 Timelkam Austria This information is given for the convenience of users of this document and does not constitute an endorsement by IEC of the company cited
Figure A.2a) – Package of TMM slices containing alternating void slices and attenuation slices of polyurethane foam
Figure A.2b) – Holes of different diameters in the void slices allow the use of the phantom with different ultrasound frequencies (1 – 15 MHz) Figure A.2 – TMM slices
The foam of the attenuation slices has a density of 120 kg m– 3 to 130 kg m– 3 ; in the void slices it is 20 kg m– 3 to 30 kg m– 3 The foam has attenuating and back-scattering properties
The specific attenuation coefficient for the attenuation slices is approximately 0.7 dB cm\(^{-1}\) MHz\(^{-1}\), while for the void slices, it is around 0.2 dB cm\(^{-1}\) MHz\(^{-1}\) Consequently, the average specific attenuation coefficient of the phantom is about 0.45 dB cm\(^{-1}\) MHz\(^{-1}\).