WO2008013514A1 - Procédé et appareil pour améliorer la détection et le suivi d'objets mobiles à l'aide d'ultrasons - Google Patents

Procédé et appareil pour améliorer la détection et le suivi d'objets mobiles à l'aide d'ultrasons Download PDF

Info

Publication number
WO2008013514A1
WO2008013514A1 PCT/US2006/017654 US2006017654W WO2008013514A1 WO 2008013514 A1 WO2008013514 A1 WO 2008013514A1 US 2006017654 W US2006017654 W US 2006017654W WO 2008013514 A1 WO2008013514 A1 WO 2008013514A1
Authority
WO
WIPO (PCT)
Prior art keywords
echo
envelope
lines
signal
track
Prior art date
Application number
PCT/US2006/017654
Other languages
English (en)
Inventor
John Edward Lynch
David Mark Blaker
David J. Colatosti
Richard Burton Mack
Original Assignee
Luna Innovations Inc.
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by Luna Innovations Inc. filed Critical Luna Innovations Inc.
Priority to PCT/US2006/017654 priority Critical patent/WO2008013514A1/fr
Priority to EP06851391A priority patent/EP2020921A1/fr
Publication of WO2008013514A1 publication Critical patent/WO2008013514A1/fr

Links

Classifications

    • G—PHYSICS
    • G01—MEASURING; TESTING
    • G01S—RADIO DIRECTION-FINDING; RADIO NAVIGATION; DETERMINING DISTANCE OR VELOCITY BY USE OF RADIO WAVES; LOCATING OR PRESENCE-DETECTING BY USE OF THE REFLECTION OR RERADIATION OF RADIO WAVES; ANALOGOUS ARRANGEMENTS USING OTHER WAVES
    • G01S15/00—Systems using the reflection or reradiation of acoustic waves, e.g. sonar systems
    • G01S15/66—Sonar tracking systems
    • A—HUMAN NECESSITIES
    • A61—MEDICAL OR VETERINARY SCIENCE; HYGIENE
    • A61B—DIAGNOSIS; SURGERY; IDENTIFICATION
    • A61B8/00—Diagnosis using ultrasonic, sonic or infrasonic waves
    • A61B8/06—Measuring blood flow
    • G—PHYSICS
    • G01—MEASURING; TESTING
    • G01S—RADIO DIRECTION-FINDING; RADIO NAVIGATION; DETERMINING DISTANCE OR VELOCITY BY USE OF RADIO WAVES; LOCATING OR PRESENCE-DETECTING BY USE OF THE REFLECTION OR RERADIATION OF RADIO WAVES; ANALOGOUS ARRANGEMENTS USING OTHER WAVES
    • G01S15/00—Systems using the reflection or reradiation of acoustic waves, e.g. sonar systems
    • G01S15/02—Systems using the reflection or reradiation of acoustic waves, e.g. sonar systems using reflection of acoustic waves
    • G01S15/50—Systems of measurement, based on relative movement of the target
    • G01S15/52—Discriminating between fixed and moving objects or between objects moving at different speeds
    • G—PHYSICS
    • G01—MEASURING; TESTING
    • G01S—RADIO DIRECTION-FINDING; RADIO NAVIGATION; DETERMINING DISTANCE OR VELOCITY BY USE OF RADIO WAVES; LOCATING OR PRESENCE-DETECTING BY USE OF THE REFLECTION OR RERADIATION OF RADIO WAVES; ANALOGOUS ARRANGEMENTS USING OTHER WAVES
    • G01S7/00—Details of systems according to groups G01S13/00, G01S15/00, G01S17/00
    • G01S7/52—Details of systems according to groups G01S13/00, G01S15/00, G01S17/00 of systems according to group G01S15/00
    • G01S7/539—Details of systems according to groups G01S13/00, G01S15/00, G01S17/00 of systems according to group G01S15/00 using analysis of echo signal for target characterisation; Target signature; Target cross-section

Definitions

  • the technical field relates to detecting moving objects using ultrasound technology.
  • One non-limiting application is to detect emboli in the bloodstream.
  • Embolic particles carried by the bloodstream can causes strokes and other circulatory disorders.
  • emboli may occur when clots form in the blood, air enters into the bloodstream, or tissue fragments break loose or become dislodged.
  • the blood carries the emboli into increasingly smaller arteries until they become lodged and obstruct the flow of blood.
  • the amount of damage that results depends on the size of the emboli, the point in which it lodges in the blood flow, the amount of blood leaking around the emboli, and how blood is supplied by collateral paths around the obstruction.
  • the resulting functional deficit depends in part on the composition of the emboli.
  • the number of embolic events can be counted by monitoring the number of reflected echoes that exceed a predetermined threshold.
  • the '051 patent also describes a method to characterize emboli by composition and size so that an embolus may be classified for example as a gas or a fat particle based on detailed analysis of the echo signal for each embolus.
  • reflected signals from the moving object need to be processed to eliminate reflections from stationary objects that are of less interest.
  • these stationary objects include the blood vessel walls and surrounding tissue.
  • the reflections from surrounding tissue are generally stronger than those from the flowing blood and from the emboli.
  • the strong reflections from stationary objects may be reduced using a moving object indicator (MOI).
  • MOI temporarily stores one line of echo data and subtracts it from a subsequent line of echo data. Differencing two lines of echo data substantially cancels the stationary object signals leaving the signal reflected from the moving objects, e.g., from the blood flow and the emboli contained therein.
  • the noise performance of an ultrasonic moving object indicator is a significant issue.
  • One way of improving noise performance is to average multiple lines in such a way that the signal-to-noise ratio is improved. In that case, differences are determined between the averages.
  • the signal-to-noise ratio improves by a factor of the square root of the number of lines averaged when the noise is incoherent and the reflected signal is coherent.
  • Averaging multiple lines results in a waveform that responds slowly to changes. The averaged waveform does not change significantly even when a moving object, e.g., an embolus, passes through the ultrasound beam.
  • the differencing however, produces a large value when the moving object is present in the ultrasound beam.
  • the averaging "filter” still leaves significant background noise artifacts.
  • the inventors realized that the performance of an ultrasonic moving object indicator could be significantly improved. They also recognized that traditional signal-to-noise ratio enhancements are very computationally intensive and thus may often require a specialized or expensive hardware and/or software. In addition, detection of embolic signals based on exceeding a predetermined threshold is not coherent from pulse to pulse, as the amplitude of the return signal often varies with time. Incoherent detection reduces the accuracy of subsequent line tracking and object classification. The echo signal enhancements described below overcome all of these problems.
  • An ultrasonic pulse echo apparatus detects a moving object that is moving with respect to its stationary object.
  • An ultrasonic transducer transmits a series of ultrasound pulses in a direction that intersects a path of a moving object.
  • An ultrasound receiver receives a series of lines of echoes from objects in the field of view of the ultrasonic signal. Each echo line corresponds to one of the ultrasonic pulses.
  • a signal processor processes the echo lines from the moving object.
  • the echo lines are time shifted by different amounts and combined at different time shifts to produce different composite lines.
  • the composite line having an optimal signal-to-noise ratio is selected.
  • the signal processor determines an amplitude of each echo line at a particular range from the ultrasound transducer for each of the different time shifts.
  • a median value or an average value of the amplitudes is determined for each time shift.
  • the time shift where the amplitudes have a minimum variance from the median value or average value is selected.
  • the echo envelope lines at the selected time shift are combined to produce the combined echo signal with the optimal signal-to-noise ratio for the particular range.
  • the median value of the echo amplitudes is determined at each time shift.
  • the time shift where the echo amplitudes have a maximum of median value Is then selected.
  • the echo lines at the selected time shift are combined to produce the combined echo signal with the optimal signal-to-Doise ratio for the particular range.
  • each echo envelope signal is then sampled at a reduced rate. More specifically, each echo envelope signal is sampled at a reduced rate relative to higher-frequency sampling in the received echo lines required for accurate echo representation. The reduced rate still ensures that at least three sample points are contained within an envelope formed from a one-cycle ultrasonic echo.
  • Line tracking is then performed to associate the detected peak in the echo line with detected peaks and success of echo lines corresponding to the moving object.
  • a track is determined corresponding to the detected peaks in successive echo lines corresponding to the moving object.
  • the signal processor determines how close a recently-detected peak is to that track, and updates the track by adding that detected peak if it is determined to be sufficiently close.
  • the accuracy of this association process has been improved through a mathematical estimate of the future velocity and position of the moving object based on its current track.
  • the accumulated echoes from each track are used to create a representative echo from the moving object. This representative echo is then used to classify the moving object based on waveform characteristics of the echo.
  • FIGURE 1 is a function block diagram illustrating one non-limiting example of an ultrasonic detection apparatus
  • FIGURE 2 illustrates an ultrasonic pulse being reflected as RF echoes by stationary and moving objects
  • FIGURE 3 illustrates the ultrasonic transducer positioned at an angle with respect to the direction of movement of the moving objects and related parameters
  • FIGURE 4 is a flow chart diagram illustrating non-limiting example steps for detecting a moving object using the ultrasonic detection apparatus shown in Figure 1;
  • FIGURE 5 includes an example to illustrate stationary object cancellation
  • FIGURE 6 illustrates an example of determining a rectified envelope of the remaining moving object echo signals and re-sampling the rectified envelope at a reduced sampling rate
  • FIGURE 7 illustrates the effect of a delay shifting filter
  • FIGURES 8A and 8B illustrate delay shifting examples
  • FIGURE 9 illustrates the effect of a waveform smoothing filter
  • FIGURE 10 illustrates a peak detection example
  • FIGURE 11 is a flow chart diagram illustrating non-limiting example steps for implementing line tracking
  • FIGURE 12 illustrates line tracking
  • FIGURE 13 represents conceptually the accumulation of echoes from a line track and the resulting echo waveform used for signal classification.
  • FIG. 1 shows a non-limiting example embodiment of a moving object detection system which is indicated by the numeral 10.
  • the moving object detection apparatus 10 is sometimes described in the context of an emboli detection application. Of course, this technology may be used with other applications.
  • Several other example applications include mechanically scanning structures, parts, or other apparatus for defects, scanning any type of fluid for particles, free hand ultrasound applications, or preferentially enhancing signals from moving fluid through a stationary background.
  • the moving object detection system 10 includes an ultrasonic processing apparatus 12 that controls an ultrasound transducer 14 positioned so that as moving objects 18 and 20 pass by the ultrasound transducer 14, ultrasonic pulses impinge on the moving objects resulting in reflected echoes that are detected by the ultrasound transducer 14.
  • There are stationary objects 16, shown as a tube or vessel with close and far walls, are also insonified by the ultrasonic pulses and also produce reflected echoes which are detected by the ultrasound transducer 14.
  • the stationary objects correspond to blood vessel walls or walls of other blood transport conduit, and the moving objects correspond to emboli.
  • the term "depth" corresponds to the perpendicular direction away from the ultrasound transducer 14 towards the objects.
  • the ultrasonic processing apparatus includes a data processor 22 coupled to memory 24 and to an ultrasonic pulser/receiver 26.
  • the ultrasound transducer 14 transmits ultrasound pulses into the body and receives echoes or reflections from within the body.
  • the transducer 14 may be a PZT composite having a quarter wave impedance matching layer to increase the coupling of sound from the transducer 14 into the object objects.
  • the ultrasonic pulser 26 also preferably (but not necessarily) applies fast-rise time step pulses to the transducer 14 which is converted by the transducer 14 into ultrasound signals that reflect off the objects being scanned.
  • One non-limiting example drive pulse has a voltage over 100 volts and a rise time on the order of 15 nanoseconds.
  • Reflections or echoes from the acoustic impedance changes in the body return to the transducer 14 which converts the reflected acoustic energy into corresponding electronic echo signals.
  • the transducer 20 preferably has a broad bandwidth so that, among other things, it can preserve the polarity of the reflected signals.
  • a plurality of ultrasound transducers may be arranged in an array and operated sequentially to produce adjacent beams that collectively cover larger areas.
  • FIG. 2 is a simplified conceptual drawing that illustrates the pulse from the ultrasonic pulser 26 energizing the ultrasonic transducer 14 to generate an ultrasonic quasi-sinusoidal looking wave that impinges on the stationary walls 16 and the moving objects 18 and 20.
  • each moving and stationary object produces one or more RF echoes.
  • RF echo 1 corresponds to the stationary front wall 16;
  • RF echo 2 corresponds to the moving object 18;
  • RF echo 3 corresponds to the moving object 20; and
  • RF echo 4 corresponds to the stationary back wall 16. All of the echoes that result from one ultrasonic pulse or "ping" are combined together as a single “line”. Multiple lines are then processed as described below.
  • the ultrasound beam is preferably angled with respect to the moving object direction, e.g., the blood flow direction for emboli, so that the effective range from the transducer face to the moving object changes as the moving object passes through the sound beam.
  • This change in the range with respect to time produces a moving object indicator (MOI) shift.
  • Figure 3 illustrates that angle ⁇ , which may have any suitable value. One non-limiting example value is 30 degrees. Angling the beam makes the moving object shift much larger than that of the surrounding stationary objects, so that the moving object indicator can more easily separate and cancel the echoes from the stationary objects.
  • an angled beam reduces the strength of echoes from the stationary walls so that dynamic range requirements for detection and signal processing are reduced.
  • the width of the transducer is designated as "b", the width of the tube or blood vessel is “a”, and the range “n” is in a direction perpendicular to the surface of the transducer towards the tube or vessel.
  • the range distance from the tube wall to the external surface of the transducer is denoted as "c”.
  • the ultrasonic processing apparatus is set to acquire echo lines at a maximum acquisition depth equal to the distance from the transducer face to the back wall of the tube, which equals a/cos ⁇ + c.
  • the distance "c" from the transducer tube wall is shortest at the top of the transducer and longest at the bottom, angled away from the transducer, so that echoes from the tube wall will be spread out over multiple range values n.
  • the ultrasonic receiver 26 amplifies the small electrical echoes from the transducer 14 to a level suitable for analyzing and processing.
  • the receiver includes amplification, time gain compensation, filtering, and analog-to-digital conversion.
  • Time gain compensation increases gain with time to compensate for the acoustic attenuation experienced as the ultrasound pulse travels deeper in the depth direction shown in Figure 1, e.g., into the body.
  • Analog-to-digital conversion needs to take place at a rate high enough to preserve the characteristics of the reflected echo signals from the moving objects.
  • analog-to-digital (A-to-D) conversion rates should be 20 MHz or higher.
  • the A-to-D converter must also have sufficient accuracy to preserve amplitude and polarity information.
  • the digitized echo outputs are passed to the data processor 22 for subsequent signal processing and stored in the memory 24. Absent the processing enhancements employed by the inventors, very large numbers of samples must be stored and processed for each transmit/receive sequence.
  • the data processor 22 analyzes the electronic echo signals to detect and preferably classify each moving object based on size and composition. If desired, the results of the moving object detection and classification may be displayed or used to produce audible tones, alarms, pre-recorded voice messages, or other signals.
  • FIG. 4 illustrates a flowchart labeled "moving object” that outlines non-limiting, example signal processing procedures that may be performed on digitized echo signals.
  • the ultrasonic transducer is periodically pulsed to acquire RF echoes from both stationary and moving objects for each pulse. Each set of received echoes per pulse is referred to as a line.
  • the echoes from stationary objects in each line are determined and removed (step S2).
  • a rectified envelope is then determined for each line, and the rectified envelope is re-sampled at a lower sampling rate to reduce the amount of data needed to be processed (step S3).
  • the samples for each line envelope are then processed using a delay shift filtering algorithm to reduce noise and to enhance the coherent combination of multiple envelope lines (step S4).
  • step S 5 a further filtering operation is performed to "smooth" the combined signal to remove false peaks (step S 5). Then, in step S 6, one or more peaks (depending on a number of moving objects) are detected in the combination signal, with each peak detected corresponding to a moving object. The detected peaks in successive lines that corresponds to the same moving object are associated in a process called “line tracking" (step S7).
  • This line tracking allows prediction of the future position (range) and velocity of the moving object which allows subsequent analyzing and processing to be focused on a narrower region in terms of position and velocity by excluding all other positions and velocities.
  • the last step describes classifying the "signature" or characteristics of the RF wave form associated with each detected peak (step S8).
  • classification of the moving object can be very important.
  • the polarity or the phase of the echo may be used to classify an embolus as either gaseous or solid.
  • Figure 5 lathis simplified example, three lines of echoes are added and divided by three to obtain a line average. Then the line average is subtracted from each individual line. The figure shows the result of subtracting the line average from line 2. It is apparent that the large echo waves corresponding to the stationary objects have been removed. But the moving objects do not constructively combine in the line average so they remain in the difference waveform.
  • a high-pass filter may be used at each range point "n" taken across multiple lines to remove the large echoes from stationary or slow-moving objects and enhance echoes from faster moving objects.
  • An SOC filter H(f) is a moving window average that is subtracted from each line. The moving window average has a sine frequency response defined by the Fourier transform of a rectangular pulse:
  • H(f) sin(M ⁇ /fp)/(M ⁇ /f P )
  • f is the frequency of the ultrasonic pulse
  • M is the number of points in the moving average.
  • An example ultrasonic pulse repetition frequency f p is 1 kHz.
  • H(f) 1- sin(M ⁇ /fp)/(M ⁇ /f P )
  • FIG. 6 shows an example where two echo waveforms are rectified and an envelope determined.
  • the Hubert transform is implemented by applying a finite impulse response (FIR) filter on the RF line signal data.
  • h_9o(n) -h +9 o(n)
  • the resulting signal is stored in memory and then phase-shifted +90° to produce a second signal, which is in-phase with the original signal but has the DC bias filtered out.
  • the magnitude of this in-phase signal combined with the -90° phase- shifted signal is then used to form the envelope.
  • a second advantageous signal processing step relates to reducing the amount of data that has to be processed.
  • the inventors determined that considerable computational resources could be conserved without any reduction in processing accuracy by re-sampling the rectified waveform at a much reduced rate.
  • Each echo envelope signal is sampled at a reduced rate relative to higher- frequency sampling in the received echo lines required for accurate echo representation. For instance, a 4 MHz echo should be sampled at 48 MHz to fully represent the echo waveform characteristics. After forming a rectified envelope, much of the frequency content of the echo has been destroyed, and a lower resampling rate is adequate to represent the shape of the envelope. Ideally, three points of the envelope are preserved after re-sampling so that the peak of the envelope can be distinguished from the edges. To obtain three samples for each 4 MHz echo, the data must be re-sampled at 3*4 MHz or 12 MHz. This is one- fourth the original sample rate of 48 MHz.
  • the signal is decimated by a factor of four. Before re-sampling, however, it is preferable to first low pass filter the envelope data to prevent anti-aliasing. Then the data is re-sampled, and every fourth point is output to the delay shift filter. Even at a fairly high resolution, e.g., a one-cycle signal with a center frequency of 4 MHz 3 a 4:1 decimation provides three sample points per echo, which is sufficient to enable peak detection.
  • the next processing stage is delay shift filtering which enhances the signal-to-noise ratio of the moving object echoes.
  • delay shift filtering which enhances the signal-to-noise ratio of the moving object echoes.
  • tests were performed comparing delay shift filtering before forming an envelope and decimating the signal (so that the delay shift filter is performed on all of the acquired echo line data) with filtering the decimated envelope data.
  • delay shift filtering using all the original sample data provided somewhat better signal-to-noise improvements, but at 16 times more computational cost.
  • RF delay shift filtering of the decimated envelope samples actually provided better noise reduction. For most applications, the somewhat improved sensitivity does not outweight the significantly reduced computational cost achieved with filtering just the decimated envelope samples.
  • Delay shift filtering shifts in time the echoes corresponding to the same moving object by different delay shift amounts in order to determine the optimal shift value that allows the same object echoes to be evaluated over a coherent track.
  • the echoes of faster moving objects require more shifting than the echoes of slower moving objects.
  • Temp[m][t][n] X[t][n-d]
  • Temp refers to an array in memory for storing a series of lines being processed
  • m is a parameter associated with the amount d by which the range n has been shifted
  • the memory array Temp[m][t][n] stores the time-shifted values X[t][n-d], which are then evaluated statistically to produce another temporary memory array Temp2[m][n].
  • the number of delays used in the delay shift filtering is determined by the range of possible velocities of the moving object.
  • a range of velocities of the emboli might be 0.1 m/s to 1 m/s.
  • the formula below converts these velocities into a slope given in units of samples per ping or line:
  • t ping is the pulse repetition period
  • c is the speed of sound in tissue (1540 m/s)
  • f s is the decimated sample rate of 12 MHz
  • V f is the flow velocity of the emboli
  • ⁇ is the transducer angle relative to the flow direction.
  • the range of delay shifts "m" is 0.8 samples/ping to 8 samples/ping. The lower limit is rounded down to zero so that nine delay shifts are needed for the delay shift filter, ranging from 0 to 8 samples/ping or line. Finer resolution shifting may also be used, but finer resolution may not translate into greater detector sensitivity.
  • the value of the time shift m that provides optimal enhancement of the echo signal-to-noise ratio is then chosen to produce a new, noise-reduced line Y[n] that is a composite of the T lines X[t][n] at that selected delay shift.
  • Three methods for determining the delay shift for the optional composite line Y[n] with statistically reduced noise as compared to each of the lines X[t][n] are described:
  • the following table illustrates how the different delay shifts may be statistically combined to achieve enhanced signal-to-noise ratio for a composite line representing the three lines as compared to any one of the three individual lines.
  • the three statistical combinations of averaging, finding the median, and determining a minimum variance are employed.
  • the variance is computed against the median, but it could also be computed against the average.
  • Figure 8B helps illustrate how the minimum covariance delay shift selection produces a more optimal echo composite signal when clutter (larger amplitude curves) is present near the echoes (smaller amplitude curves).
  • the maximum median filter is somewhat more sensitive than the minimum variance filter, but at the cost of introducing additional noise. Thus, the maximum median filter is more likely to produce false positive readings, particularly in cluttered environments, but more likely to detect weak signals.
  • An additional advantage of the minimum variance filter is that the detected position is more coherent than with the maximum median filter, resulting in stronger RF signatures when the signal is averaged along a track.
  • the inventors performed various tests using simulated data to compare the performance of the maximum median and minimum variance filter approaches. At a signal-to-noise ratio (SNR) of 1.8 dB, the maximum median filter was swamped by noise, making detection very difficult, while the minimum variance filter did a better job of noise reduction so that the echo signal track could be separated from the noise. At a higher SNR of 3.3 dB s the maximum median filter performed better than the minimum variance filter. The maximum median filter did a better job of enhancing the composite signal so that the track is stronger than in the minimum variance filter.
  • SNR signal-to-noise ratio
  • the minimum variance filter was determined to provide better performance than the maximum median filter in cluttered environments.
  • the sensitivity advantage of the maximum median filter may be compensated for by hardware improvements that increase the signal-to-noise ratio of the input data.
  • the minimum variance filter is considerably less computationally intensive than the maximum median filter, since the minimum variance filter computations perform only about half (e.g., 5 lines of data are needed in the test example) of the number of computations required for the maximum median filter (e.g., 9 or 11 needed in the test example). This computational reduction of one half coupled with computational reduction of one sixteenth achieved in the lower sampling rate of the echo envelope results in an overall computation reduction of 1/32 when the minimum variance approach is used.
  • Tests were also performed to compare performing delay shift filtering before forming an envelope and decimating the signal (so that the delay shift filtering is performed on RF data) with the use of delay shift filtering after forming an envelope.
  • Delay shift filtering on the RF data provided somewhat better signal-to-noise ratio, but at 16 times more computational cost.
  • SNR 1.9 dB
  • an RF delay shift filtering performed on RF data provided better noise reduction than the delay shift filtering on the rectified envelope, but the signal level was higher for delay shift filtering on the rectified envelope.
  • the RF delay shift filtering performed on RF data did not increase the sensitivity of the detector enough to justify the extra computational cost.
  • a minimum variance condition is used to select delay shifts instead of a maximum signal condition used in the average and median value approaches, it may be necessary to adjust the length of the delay shift filter.
  • the maximum signal condition the best results are obtained when the median or average value is calculated over 12 or more lines because signal-to-noise improves with -JN , where N is the number of lines over which the delay shift filter is implemented.
  • N is the number of lines over which the delay shift filter is implemented.
  • the next step is to filter the composite line to smootli the echo wave shapes.
  • delay shift filtering enhancement there is still inherent system noise. Because the delay shifts are only calculated over integer values, the echo envelope peaks tend to decorrelate. That means that noise rides on top of an echo envelope peak, so that a peak detector based on finding local maxima will output a cluster of detected events around a global maximum that passes a minimum threshold test.
  • the composite lines are processed by a smoothing filter, e.g., a Gaussian smoothing filter.
  • a Gaussian smoothing function g[n] maybe derived via the following function:
  • the next step is detecting the peaks of the echoes. Peak detection provides more reliable estimation of the range of an object than threshold detectors because they detect a common feature within the echo envelope — the peak value — that does not vary in range as the amplitude of the signal changes.
  • One example of a peak detector implementation is based on a quadratic curve fit of the data over a moving N-point window. Although accurate and reliable, this implementation is computationally intensive.
  • An alternative is a peak detection algorithm based on a finite difference operator. A central difference operator is preferred because it has the same computational requirements as a forward or backward difference operator, but is less sensitive to noise. [0061] The central difference operator differentiates (the derivative operation is represented as d_) the output Y of the Gaussian smoothing operator as follows:
  • a peak is defined at any point where d_Y n > 0 and d_Y n+1 ⁇ 0, and where Y n exceeds a threshold value defined by the background noise characteristics of the ultrasonic system. Testing indicates that the central difference operator. detects peaks equivalently to the quadratic curve fit so long as the detection threshold is raised by approximately 25%. The higher threshold is required because the quadratic curve returns an estimated peak that is lower than the peak in the detected data, while the central difference operator returns the actual peak value.
  • line tracking is performed to associate detected object echoes along the moving trajectory of that object so that multiple detections from a single object are not counted more than once.
  • the line tracker also provides additional filtering of spurious noise events and provides a convenient way to acquire an RF signature of an object echo. As described below, the RF signature may then be used to classify the object by size, composition, or other characteristics, if desired.
  • the line tracking algorithm For each line of echo data, the line tracking algorithm associates detected echo peaks with existing object tracks and maintains the data in a track table in memory. As shown in Figure 12, the example line tracking algorithm operates in three phases I, II, and III. In phase I, detected events (i.e., objects) from one cycle are checked against all tracks (step SlO) and associated with one or more deferred or active tracks. The detected event is associated with a track if the range of this object is "close" to the estimated range in the track table (step S 12), and if the measured delay shift or "slope" (obtained from the delay shift filter output) is "close” to the estimated velocity. If not, a current line tracking flag value is maintained (step S 13).
  • detected events i.e., objects
  • step SlO all tracks
  • the detected event is associated with a track if the range of this object is "close” to the estimated range in the track table (step S 12), and if the measured delay shift or "slope" (obtained from the
  • Closeness is defined by an error bar around the difference between the estimated range and measured range and an error bar between the measured delay shift/slope and the estimated velocity.
  • a small error bar or +/- 2 range values may be appropriate, as large error bars may create false tracks.
  • a larger error bar of +/- 3 may be better as smaller errors tend to create false negatives.
  • the flag is incremented by 1 (step S 14).
  • range and magnitude data for the peak are added to the track table, and the original RF echo data, obtained at the output of the stationary object canceller, is added to the accumulated echo stored within the track table.
  • This accumulated echo signature is used later for classification of the moving object.
  • a check is made for overlapping tracks, and if overlap is detected, the shorter of the overlapping tracks is terminated (prior to updating track state in Phase II) or deleted, depending on the length of the track (step S 15).
  • the track state (deferred, active, or terminated) is updated, the estimated position and velocity of the moving object is updated, and old events are rotated out of the track table.
  • the track state may be updated (step S 16) using the following non-limiting example procedures:
  • the future position/range estimate X_est for the moving object and the future velocity estimate V_est for the moving object are estimated for the next cycle based on the current track for that moving object (step S 17).
  • that estimation may be determined using a Kalman filter as follows:
  • the factors g and h are smoothing factors that reduce the jitter in the estimate based on measurement error. The choice of g and h depends on how much the object's velocity is expected to change from line to line, and what the measurement error is in determining the range of the object.
  • Phase III peaks that were not associated with tracks in Phase I are used to initiate new tracks (step S 18).
  • the delay shift data from the delay shift filter is use to give an initial V_est and X_est.
  • the next procedure is signature classification. Before counting a track output as an object, e.g., an embolus, the track data is used to pull an RF signature from stationary object cancellation (SOC) data in memory. These RF signatures are averaged along the track, and the signature may then be analyzed to determine if the track consists of coherent — and therefore valid — RF signatures or if it consists primarily of noise. This is illustrated conceptually in Figure 13.
  • a threshold detector in which the signature passes the test if the signature passes a certain threshold, can be used to measure the degree of coherence. Further classification of the signal may be possible through detailed analysis of the time- domain signature, by performing a fast Fourier transform on the signal to analyze its spectral content, or any number of other signal characterization techniques. Other classification methods may be employed such as those described in USP 5,441,051.

Landscapes

  • Engineering & Computer Science (AREA)
  • Radar, Positioning & Navigation (AREA)
  • Remote Sensing (AREA)
  • Physics & Mathematics (AREA)
  • Health & Medical Sciences (AREA)
  • Life Sciences & Earth Sciences (AREA)
  • Computer Networks & Wireless Communication (AREA)
  • General Physics & Mathematics (AREA)
  • Radiology & Medical Imaging (AREA)
  • Medical Informatics (AREA)
  • Nuclear Medicine, Radiotherapy & Molecular Imaging (AREA)
  • Pathology (AREA)
  • Hematology (AREA)
  • Biomedical Technology (AREA)
  • Heart & Thoracic Surgery (AREA)
  • Biophysics (AREA)
  • Molecular Biology (AREA)
  • Surgery (AREA)
  • Animal Behavior & Ethology (AREA)
  • General Health & Medical Sciences (AREA)
  • Public Health (AREA)
  • Veterinary Medicine (AREA)
  • Acoustics & Sound (AREA)
  • Measurement Of Velocity Or Position Using Acoustic Or Ultrasonic Waves (AREA)

Abstract

L'invention concerne un appareil d'écho d'impulsion ultrasonore qui détecte un objet se déplaçant par rapport à des objets stationnaires. Un transducteur ultrasonore transmet une série d'impulsions ultrasonores dans une direction intersectant le trajet de l'objet mobile. Un récepteur ultrasonore reçoit une série de lignes d'échos emis par les objets situés dans le champ du signal ultrasonore. Chaque ligne d'écho correspond à l'une des impulsions ultrasonores. Un processeur de signal traite les lignes d'écho provenant de l'objet mobile. Les lignes d'écho sont échelonnées dans le temps et combinées à différents décalages dans le temps pour produire différentes lignes composites. La ligne composite ayant un rapport signal/bruit optimal est sélectionnée. D'autres perfectionnements de traitement de signal sont réalisés.
PCT/US2006/017654 2006-05-08 2006-05-08 Procédé et appareil pour améliorer la détection et le suivi d'objets mobiles à l'aide d'ultrasons WO2008013514A1 (fr)

Priority Applications (2)

Application Number Priority Date Filing Date Title
PCT/US2006/017654 WO2008013514A1 (fr) 2006-05-08 2006-05-08 Procédé et appareil pour améliorer la détection et le suivi d'objets mobiles à l'aide d'ultrasons
EP06851391A EP2020921A1 (fr) 2006-05-08 2006-05-08 Procédé et appareil pour améliorer la détection et le suivi d'objets mobiles à l'aide d'ultrasons

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
PCT/US2006/017654 WO2008013514A1 (fr) 2006-05-08 2006-05-08 Procédé et appareil pour améliorer la détection et le suivi d'objets mobiles à l'aide d'ultrasons

Publications (1)

Publication Number Publication Date
WO2008013514A1 true WO2008013514A1 (fr) 2008-01-31

Family

ID=38981747

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/US2006/017654 WO2008013514A1 (fr) 2006-05-08 2006-05-08 Procédé et appareil pour améliorer la détection et le suivi d'objets mobiles à l'aide d'ultrasons

Country Status (2)

Country Link
EP (1) EP2020921A1 (fr)
WO (1) WO2008013514A1 (fr)

Cited By (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
EP2765919A4 (fr) * 2011-10-14 2015-08-05 Jointvue Llc Reconstruction ultrasonore 3d en temps réel du genou et ses complications pour implants spécifiques à un patient et injections articulaires 3d
CN110868166A (zh) * 2019-10-30 2020-03-06 深圳市智微智能科技开发有限公司 多媒体电子白板超声波信号处理电路

Citations (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US5761155A (en) * 1994-02-28 1998-06-02 Siemens Aktiengesellschaft Method for determining the position of an object relative to the background with ultrasound
US20060052699A1 (en) * 2003-05-30 2006-03-09 Angelsen Bjern A Acoustic imaging by nonlinear low frequency manipulation of high frequency scattering and propagation properties

Patent Citations (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US5761155A (en) * 1994-02-28 1998-06-02 Siemens Aktiengesellschaft Method for determining the position of an object relative to the background with ultrasound
US20060052699A1 (en) * 2003-05-30 2006-03-09 Angelsen Bjern A Acoustic imaging by nonlinear low frequency manipulation of high frequency scattering and propagation properties

Cited By (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
EP2765919A4 (fr) * 2011-10-14 2015-08-05 Jointvue Llc Reconstruction ultrasonore 3d en temps réel du genou et ses complications pour implants spécifiques à un patient et injections articulaires 3d
CN110868166A (zh) * 2019-10-30 2020-03-06 深圳市智微智能科技开发有限公司 多媒体电子白板超声波信号处理电路

Also Published As

Publication number Publication date
EP2020921A1 (fr) 2009-02-11

Similar Documents

Publication Publication Date Title
US7894874B2 (en) Method and apparatus for enhancing the detecting and tracking of moving objects using ultrasound
US11054520B2 (en) Method and system for ultrasound time-of-flight measurement
US7798968B2 (en) Automatic detection system and method of spectral Doppler blood flow velocity
Shiina et al. Strain imaging using combined RF and envelope autocorrelation processing
JP4531894B2 (ja) 超音波イメージングにおける自動ドプラ角推定方法及び装置
EP1597607B1 (fr) Procede et dispositif de detection de discontinuites dans un milieu
US6447455B2 (en) Ultrasound diagnostic apparatus and method for measuring blood flow velocity using doppler effect
KR100459616B1 (ko) 도플러 효과를 이용하여 인체 조직 속도를 측정하기 위한초음파 진단 장치 및 방법
US6530890B2 (en) Ultrasound diagnostic apparatus and method for measuring blood flow velocity using doppler effect
JP2000508774A (ja) 超音波による物体の特徴付け
WO2003019230A1 (fr) Mesure des caracteristiques de l'air en basse atmosphere
US20080242997A1 (en) Method and apparatus for classifying gaseous and non-gaseous objects
AU646976B2 (en) Ultrasonic ranging devices
US7363177B2 (en) Apparatus and method for performing the time delay estimation of signals propagating through an environment
JP2000230972A (ja) レーダ信号処理装置
EP2020921A1 (fr) Procédé et appareil pour améliorer la détection et le suivi d'objets mobiles à l'aide d'ultrasons
CN110108797B (zh) 利用声阻抗变化信息的介质分界面超声检测方法
CN101889877A (zh) 基于希尔伯特变换-基频估计的平均骨小梁间距估计方法
JP6827577B2 (ja) 信号処理装置及び信号処理方法
JP3881078B2 (ja) 周波数推定方法、周波数推定装置、ドップラソナーおよび潮流計
JPH04231889A (ja) 超音波エコーグラフの固定エコーを除去するデバイス
KR101550706B1 (ko) 초음파를 이용한 두께 측정 장치 및 이의 두께 측정 방법
US20040249284A1 (en) Vector Doppler utilizing constancy of vector flow direction
CN116736255A (zh) 弱目标线谱多帧相干检测前跟踪方法
KR20220052524A (ko) 복수의 층을 통과하여 얻은 초음파 데이터를 분석하는 방법

Legal Events

Date Code Title Description
121 Ep: the epo has been informed by wipo that ep was designated in this application

Ref document number: 06851391

Country of ref document: EP

Kind code of ref document: A1

DPE1 Request for preliminary examination filed after expiration of 19th month from priority date (pct application filed from 20040101)
NENP Non-entry into the national phase

Ref country code: DE

WWE Wipo information: entry into national phase

Ref document number: 2006851391

Country of ref document: EP

DPE1 Request for preliminary examination filed after expiration of 19th month from priority date (pct application filed from 20040101)