Reducing electromagnetic noise in ultrasound signals
By learning the noise correlation between different channels and utilizing the catheter geometry and transmission characteristics, efficient removal of noise signals in ultrasonic catheters is achieved, the problem of electromagnetic noise interference is solved, and the accuracy and quality of signal processing is improved.
Patent Information
- Application Number
- CN202410572521.8
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Priority Date
- 2023-05-11
- Filing Date
- 2024-05-10
- Publication Date
- 2024-11-12
AI Technical Summary
In the ultrasonic catheter, electromagnetic noise generated by nearby electronic components is superimposed on the electrical signal, resulting in serious noise interference in signal processing, and it is difficult for the prior art to effectively remove it.
Learn the correlation of noise between different channels through a computer processor, estimate noise on a given channel using noise signals and correlations, and remove noise in real time when receiving signals, using the geometry of the catheter and the ultrasonic transducer. Transmission characteristics optimize signal reception.
It effectively reduces noise interference in the signal, improves the quality and accuracy of ultrasonic signal processing, and enhances the accuracy of estimation of subject tissue parameters and images and anatomical mapping.
Smart Images

Figure CN118924423A_ABST
Abstract
Description
Technical Field
[0001] The present disclosure relates generally to the field of signal processing, and in particular to the processing of ultrasound signals. Background Art
[0002] Ultrasonic transducers generate ultrasonic waves by converting electrical signals into mechanical energy, and converting the reflections of the waves back into electrical signals. BRIEF DESCRIPTION OF THE DRAWINGS
[0003] The present disclosure will be more fully understood through the following detailed description of examples of the present disclosure in conjunction with the accompanying drawings, in which:
[0004] Figure 1 is a schematic diagram of an exemplary electrophysiological mapping and ablation system according to some examples of the present disclosure;
[0005] Figure 2 is a schematic diagram of removing noise from a signal according to some examples of the present disclosure;
[0006] Figure 3 is a schematic diagram of a technique for calculating one or more noise estimation functions according to some examples of the present disclosure;
[0007] Figure 4 is a flow chart of an algorithm for calculating one or more noise estimation functions according to some examples of the present disclosure;
[0008] Figure 5 is a schematic diagram of the calculation of an estimated noise signal according to some examples of the present disclosure;
[0009] Figure 6 is a flow chart of an algorithm for reducing noise in a signal according to some examples of the present disclosure;
[0010] Figure 7 is a schematic diagram of the calculation of one or more noise estimation functions according to some examples of the present disclosure;
[0011] Figure 8 is a flow chart of an algorithm for removing noise from a signal according to some examples of the present disclosure;
[0012] Fig. 9 is a flow chart of an exemplary calculation of an impulse response vector according to some examples of the present disclosure;
[0013] Fig.10 is a flow chart of an exemplary calculation of noise estimation according to some examples of the present disclosure; and
[0014] Fig.11 is a schematic diagram of an ultrasonic catheter according to some examples of the present disclosure. DETAILED DESCRIPTION
[0015] Overview
[0016] In some procedures, an ultrasound catheter including multiple ultrasound transducers is inserted into the body of a subject. The transducers then transmit ultrasound waves and receive reflections of these waves from the subject's tissue. In response to these reflections, the transducers generate electrical signals that are received through different corresponding channels and can be processed, for example, to estimate parameters of the subject's tissue (e.g., thickness or elasticity), and / or to generate images and / or anatomical maps of the tissue.
[0017] When performing such procedures, a challenge is that nearby electronic components may generate electromagnetic noise that is superimposed on the electrical signal.
[0018] To address this challenge, the inventors cleverly exploited two observations:
[0019] 1) The noise on any one channel is usually correlated with the noise on the other channels, just with different amplitudes and / or phases.
[0020] 2) In many cases, although a signal is received through one channel, at least some of the other channels carry "noise signals" that include noise but do not include any substantial transduction of ultrasound reflections. For example, the ultrasound transducers can be arranged on a convex surface (e.g., a balloon, such as an elliptical balloon) rather than distributed in a two-dimensional array. Alternatively or in addition, the distance between two ultrasound transducers can be relatively large, for example, greater than ten times the wavelength of the transmitted wave. Alternatively or in addition, the transmitted beam can be relatively narrow. Due to the convexity of the surface, the distance between the transducers, and / or the narrowness of the beam, there may be relatively little overlap in the respective fields of view of the transducers.
[0021] More specifically, based on these two observations, the inventors realized that a computer processor can be used to learn how the noise on any given channel is related to the noise on other channels. Subsequently, given a noise signal received through the given channel and one or more noise signals simultaneously received through other channels, the processor can estimate the noise on the given channel based on the noise signal and the learned correlations. The processor can then subtract this estimated noise from the signal.
[0022] In some examples of the present disclosure, the processor learns a correlation in a first execution phase based on a first noise signal received simultaneously through a given channel and one or more other channels. Subsequently, in a second execution phase, while receiving the signal, the processor removes noise from the signal based on the learned correlation and a second noise signal received through other channels.
[0023] In other examples, the processor uses only a single set of noise signals that are received through other channels while the signal is received through a given channel. In particular, the processor identifies portions of the signal that appear to not include any significant transduction of ultrasound reflections. Based on these portions and corresponding portions of the noise signal, the processor learns correlations. The processor then removes noise from the signal based on the learned correlations and the noise signal.
[0024] In some examples, the processor calculates a linear time-invariant (LTI) model describing the correlation between channels. In particular, for any given channel in the channel, the processor can calculate an impulse response vector that describes the noise on the given channel as a function of the noise on one or more other channels. The impulse response vector can then be used to calculate an estimated noise on the given channel.
[0025] For greater efficiency, some examples of the present disclosure utilize the geometry of the catheter to facilitate receiving simultaneous signals through different corresponding channels, and / or receiving a noise signal for use in estimating noise through a first channel while receiving a signal through a second channel. Alternatively or in addition, such optimization can be achieved with the aid of varying transmission characteristics (e.g., frequency) between ultrasonic transducers.
[0026] System Description
[0027] First reference Figure 1 , which is a schematic diagram of an exemplary electrophysiological mapping and ablation system 10 according to some examples of the present disclosure. A commercial product embodying elements of the system 10 is the CARTO TM 3 systems.
[0028] The system 10 includes one or more catheters that are percutaneously inserted into the body of a subject 23 by a physician 24. After inserting each catheter, the physician 24 navigates the catheter through the vascular system of the subject 23 into a chamber or vascular structure of the subject's heart 12. Typically, the catheter is navigated through a delivery sheath 14. The catheters may include catheters for sensing signals from cardiac tissue (such as intracardiac electrogram (IEGM) signals), catheters for ablating cardiac tissue, and / or catheters for both sensing and ablation.
[0029] Typically, each catheter includes a tracking sensor for tracking the three-dimensional (3D) position and orientation of the distal end of the catheter. Typically, the tracking sensor includes three magnetic coils. A positioning pad 25 including a plurality of magnetic coils 32 configured to generate a magnetic field in a predetermined working space is positioned near (e.g., below) the subject 23. Based on the signal induced in the coil by the magnetic field, the position and orientation of the distal end of the catheter are tracked. The details of such magnetic-based tracking are described in U.S. Patents Nos. 5,5391,199, 5,443,489, 5,558,091, 6,172,499, 6,239,724, 6,332,089, 6,484,118, 6,618,612, 6,690,963, 6,788,967, and 6,892,091.
[0030] Typically, the system 10 also includes one or more electrode patches 38 configured to contact the skin of the subject 23. The patches 38 can establish a position reference for the positioning pad 25. In addition, the current from the electrodes disposed on any catheter can be sensed at the patches 38, and the position of each electrode can be triangulated in response thereto. This position information can be combined with the information derived from the magnetic-based tracking described above to increase the accuracy of catheter tracking. Details of such impedance-based tracking techniques are described in U.S. Patent Nos. 7,536,218, 7,756,576, 7,848,787, 7,869,865, and 8,456,182.
[0031] The system 10 also includes a recorder 11 and a display 27. The recorder 11 is configured to record electrocardiogram (ECG) signals 21 acquired by the surface ECG electrodes 18 and / or IEGM signals acquired by the in-vivo catheters, and optionally display these signals on the display 27. The recorder 11 may also be configured to pace the heart 12 and / or may be electrically connected to an independent pacemaker.
[0032] The system 10 also includes an ablation energy generator 50 configured to conduct ablation energy to one or more electrodes at the distal end of the ablation catheter. The energy generated by the ablation energy generator 50 may include, but is not limited to, radio frequency (RF) energy and / or pulsed field ablation (PFA) energy, including unipolar or bipolar high voltage direct current (DC) pulses for achieving irreversible electroporation, or a combination thereof. The ablation energy generator 50 includes an energy generating circuit configured to generate energy and a controller configured to control the circuit and optionally perform other computing functions.
[0033] The system 10 also includes a workstation 55 that includes a processor 34, volatile memory and / or nonvolatile memory, including volatile memory and / or nonvolatile memory that can store appropriate software instructions and / or data. The processor 34 can be configured to perform a variety of functions, including, for example: (1) mapping the endocardial anatomy of the heart 12 in 3D and rendering the resulting anatomical map 20 for display on the display 27; (2) displaying the activation sequence and / or other data compiled from the ECG signal 21 on the display 27 as representative visual markers or images superimposed on the anatomical map 20; (3) displaying the real-time position and orientation of one or more catheters within the body of the subject 23; and (4) displaying a site of interest (such as a site to which ablation energy has been applied).
[0034] The system 10 also includes a patient interface unit (PIU) 30 that is configured to establish electrical communication between a power source, a workstation 55, and an electrophysiology device belonging to the system. The electrophysiology device may include, for example, one or more catheters, a positioning pad 25, an ECG electrode 18, an electrode patch 38, an ablation energy generator 50, and / or a recorder 11. Typically, the PIU 30 also includes another processor that is configured to calculate the position and orientation of each of the catheters and perform any related calculations based on the ECG signal 21.
[0035] In some examples, at least one of the catheters is an ultrasound catheter 16 on which a plurality of ultrasound transducers 44 are disposed. The ultrasound transducers 44 are connected to different corresponding (wired) channels 49 extending to the proximal end of the catheter 16. The circuit 47 receives electrical signals through the channels 49, which include transduction of ultrasound reflections and electromagnetic noise from the tissue of the subject 23. Typically, the circuit 47 includes an anti-aliasing filter that limits the signal bandwidth, an analog-to-digital (A / D) conversion circuit configured to digitize the electrical signal, and a noise filtering circuit configured to remove some noise from the signal. The processor 48 is configured to receive signals (usually in digital form) from the circuit 47 and process the signals as described in detail below with reference to subsequent figures. The output of the processing can be transmitted to the processor 34, which can use the output to facilitate the construction of the anatomical map 20 and / or display an ultrasound image on the display 27.
[0036] Typically, the circuitry 47 and processor 48 are located in an ultrasound unit 46 , which is connected to the catheter 16 via a cable 51 .
[0037] exist Figure 1In the example shown, the catheter 16 includes a balloon 40 and a plurality of sensing units 42 disposed on the balloon 40. Each sensing unit 42 includes an ultrasound transducer 44 and optionally includes an electrode for collecting IEGM signals. As described above, each transducer is connected to a circuit 47 via one or more wires passing through the catheter 16. Each electrode is connected to the PIU 30 via one or more other wires passing through the catheter 16.
[0038] In other examples, the catheter 16 includes a plurality of splines (eg, in a basket format), or any other suitable structure for supporting the ultrasonic transducer 44 .
[0039] As described in the above overview, various components of system 10 (such as PIU 30) may generate electromagnetic noise that is superimposed on the signal received from transducer 44 and cannot be completely removed by circuit 47. Typically, processor 48 is configured to reduce such noise, as described in detail below with reference to subsequent figures. Alternatively, this noise reduction function may be performed by another processor belonging to system 10 (such as processor 34 or a processor of PIU 30). As another option, the function may be performed in coordination by multiple processors belonging to system 10. Therefore, it should be understood that the term "processor" as used below may refer to any single processor or group of multiple processors belonging to system 10.
[0040] In general, the functions of each processor in the processors described herein may be implemented only in hardware, for example, using one or more fixed functions or general purpose integrated circuits, application specific integrated circuits (ASICs) and / or field programmable gate arrays (FPGAs). Alternatively, the function may be implemented at least in part in software. For example, a processor may be embodied as a programmed processor including, for example, a central processing unit (CPU) and / or a graphics processing unit (GPU). Program code (including software programs and / or data) may be loaded to be executed and processed by a CPU and / or GPU. For example, program code and / or data may be downloaded to a processor in electronic form via a network. Alternatively or in addition thereto, program code and / or data may be provided and / or stored on a non-transient tangible medium, such as a magnetic, optical or electronic memory. Such program code and / or data, when provided to a processor, generates a machine or a special purpose computer configured to perform the tasks described herein.
[0041] Noise reduction
[0042] Reference now Figure 2 , Figure 2 is a schematic diagram of removing noise from a signal 52 according to some examples of the present disclosure. (Although for ease of explanation, this figure uses a signal envelope to describe each signal, it should be noted that the signal processing described herein is generally for the signal from the circuit 47 ( Figure 1) is performed on the original signal received, which usually includes radio frequency oscillations within the signal envelope.
[0043] As above reference Figure 1 As described, the signal 52 (including the transducer 54 of the ultrasonic reflection) can be received by the circuit 47 through any given channel in the channel 49. In addition to receiving the signal 52 from the circuit, the processor also reduces the noise in the signal. In particular, the processor uses one or more noise estimation functions to calculate an estimated noise signal, and then subtracts the estimated noise signal from the signal 52, thereby obtaining a signal 52' with less noise.
[0044] As further described below, the processor calculates one or more noise estimation functions for a given channel based on one or more noise signals that are received through other channels in channel 49 and include less transduction of ultrasound reflections than signal 52. Each of the noise estimation functions is applicable to a corresponding subset of the noise signals so as to return an estimate of the noise received through the given channel when the subset of noise signals is received. For example, the processor may calculate a single noise estimation function that is applicable to all noise signals in the noise signals (i.e., the aforementioned subsets are not necessarily strict subsets). Alternatively, for example, the number of noise estimation functions may be the same as the number of noise signals, each noise estimation function being applicable to a different corresponding one of the noise signals.
[0045] 1. Two-stage technology
[0046] In some examples, the processor performs the above functions in two stages using two different sets of noise signals. In particular, one or more "first noise signals" received simultaneously from a given channel and one or more other channels are used to calculate a noise estimation function. The noise estimation function is then applied to one or more "second noise signals" that are received from the same other channels when receiving signal 52 in order to calculate an estimated noise signal.
[0047] Each of these stages is described in detail herein.
[0048] (i) Stage 1: Calculation of the noise estimation function
[0049] Reference now Figure 3 , Figure 3 is a schematic diagram of a technique for computing one or more noise estimation functions 56 based on a first noise signal 58 according to some examples of the present disclosure.
[0050] like Figure 3As shown, corresponding first noise signals 58_1, 58_2, ... 58_N are simultaneously received through N channels in the channel. (Typically, the first noise signal 58 does not include any transduction of ultrasound reflection.) Based on the first noise signal 58, the processor calculates one or more noise estimation functions 56 for at least one channel in the N channels. For example, the processor can calculate N groups of one or more noise estimation functions, each group of noise estimation functions is used to reduce the noise received through a different corresponding one of the N channels.
[0051] In some examples, the first noise signal is received before the catheter enters the body of the subject during a calibration procedure. In other examples, the first noise signal is received while the catheter is inside the body of the subject, for example, just before transduction of ultrasound reflections begins. In other examples, the first noise signal is received while the catheter is inside or outside the body, after the signal 52 ( Figure 2 ) after receiving the first noise signal. For example, the processor may separate the signal received through one of the channels into two parts: (a) signal 52, which includes transduction of ultrasound reflections; and (b) the first noise signal after signal 52, which is known not to include any such transduction. Fig.11 Other options for receiving the first noise signal are described.
[0052] For an exemplary implementation of this functionality, reference is now additionally made to Figure 4 , Figure 4 is a flow chart of an algorithm 86 for calculating one or more noise estimation functions according to some examples of the present disclosure.
[0053] The algorithm 86 starts with a signal receiving step 88, where the processor receives the first noise signal 58 from the circuit, which is received simultaneously through different corresponding channels. After the signal receiving step 88, the processor selects one of the channels at a channel selection step 90.
[0054] Next, the processor calculates one or more noise estimation functions, each noise estimation function being configured to estimate the noise received by the selected channel based on the noise (which is based on the noise received on the different corresponding subgroups of other channels receiving the first noise signal). The processor selects each such subgroup at the subgroup selection step 92. Next, at the function calculation step 94, the processor calculates the noise estimation function of the selected subgroup (that is, the noise estimation function with the noise received by the selected channel subgroup as input) according to the corresponding first noise signal (that is, the first noise signal received by the subgroup of the channel) and the first noise signal received by the selected channel. After the calculation, the processor checks at the checking step 96 whether there are still channel subgroups to choose from. If so, the processor returns to the subgroup selection step 92.
[0055] In some examples, the processor computes only a single noise estimation function, and therefore selects only a single subset for the selected channel.
[0056] When it is determined at check step 96 that no unselected subgroups remain, the processor checks at another check step 98 whether any channels remain for selection. If so, the processor returns to channel selection step 90. Otherwise, execution of algorithm 86 ends.
[0057] (ii) Second stage: noise removal
[0058] Reference now Figure 5 , Figure 5 is a schematic diagram of the calculation of the estimated noise signal 62 according to some examples of the present disclosure.
[0059] For example, Figure 5 It shows that when signal 52 ( Figure 2 ) is received through the first channel, an estimated noise signal 62 of the first channel of the N channels is calculated based on the second noise signal 60 received through the second to Nth channels. In particular, to calculate the estimated noise signal, the processor applies one or more noise estimation functions calculated for the first channel to the second noise signals 60_2, 60_3, ... 60_N. Subsequently, as described above with reference to Figure 2 The processor uses the estimated noise signal to reduce noise in a signal received via the first channel.
[0060] The second noise signal typically comprises at most only the transduced energy of a small number of ultrasound reflections. For example, for each of the second noise signals, the energy of the noise in the signal may exceed the energy of any ultrasound reflection in the signal, for example by at least five times, such as at least ten times.
[0061] For an example implementation of this functionality, reference is now additionally made to Figure 6 , Figure 6 is a flow chart of an algorithm 100 for reducing noise in a signal according to some examples of the present disclosure.
[0062] Algorithm 100 begins at a signal receiving step 102 where the processor receives a signal from circuit 47 ( Figure 1 ) receives a signal 52 received through any given channel of the channel ( Figure 2 ) and a second noise signal 60 received through one or more other channels while receiving the signal 52. (As mentioned above, the signal 52 and the second noise signal can be received before the first noise signal, that is, the signal receiving step 102 can be performed at Figure 4) After the signal receiving step 102 , the processor selects a subgroup of other channels at the subgroup selecting step 92 .
[0063] For each selected subgroup, at an estimate calculation step 104, the processor calculates a noise estimate for the selected subgroup from the corresponding second noise signal using the corresponding noise estimation function. In other words, using the noise estimation function calculated for the selected subgroup and the given channel, the processor calculates a noise estimate in the signal 52 based on the second noise signal received from the selected subgroup. Next, at a check step 96, the processor checks whether there are more subgroups to be selected. If so, the processor returns to the subgroup selection step 92.
[0064] As described above, in some examples, there is only a single subset of other channels, so that the processor calculates only a single noise estimate based on all noise signals in the second noise signals. In other examples, the processor calculates multiple noise estimates from different corresponding subsets of the second noise signals using different corresponding noise estimation functions. For example, the processor can calculate a separate noise estimate from each second noise signal in the second noise signals.
[0065] Thus, when it is determined at check step 96 that no subgroups remain to be selected, the processor checks at another check step 106 whether multiple noise estimates have been calculated. If so, then at estimated noise signal calculation step 108, the processor calculates the estimated noise signal 62 based on the noise estimates. For example, the processor may calculate the estimated noise signal by applying a median filter to the noise estimates. Otherwise, it is assumed that the estimated noise signal has already been calculated at estimate calculation step 104, and the estimated noise signal calculation step 108 is omitted.
[0066] After the estimated noise signal calculation step 108, or if this step is omitted, the processor subtracts the estimated noise signal 62 from the signal 52 at a subtraction step 110, thereby obtaining a less noisy signal 52' ( Figure 2 ).
[0067] II. Phase I Technology
[0068] In other examples, the processor uses only one set of noise signals. In other words, to calculate the estimated noise signal, the processor applies the noise estimation function to the same noise signal used to calculate the noise estimation function. In such examples, typically, the noise signal includes at most only a small amount of transduced energy of ultrasound reflections. For example, for each noise signal in the noise signal, the energy of the noise in the signal may exceed the energy of any ultrasound reflection in the signal, for example by at least five times, such as by at least ten times.
[0069] For further details, refer now to Figure 7 , Figure 7is a schematic diagram of the calculation of one or more noise estimation functions 56 according to some examples of the present disclosure.
[0070] In some examples, after receiving signal 52 through any given channel, the processor extracts noise signal 64 from signal 52. The processor then calculates one or more noise estimation functions for the given channel using noise signal 64 and one or more noise signals 66 received through corresponding other channels when signal 52 was received through the given channel.
[0071] For example, the processor may first segment the signal 52 into a plurality of segments 65. Subsequently, the processor may identify one or more segments 65 that include less transduced ultrasound reflections than other segments, i.e., are noisier than other segments. Next, the processor may construct a noise signal 64 from the identified segments, for example, by invalidating the unidentified (less noisy) segments in the signal 52.
[0072] Typically, before calculating noise estimation function 56, the processor modifies noise signal 66 by removing noise from any time windows in noise signal 64 that do not represent signal 52. For example, Figure 7 As shown, if the noise signal 64 is zero between t0 and t1, the processor may invalidate the portion of each noise signal 66 from t0 to t1. The processor then calculates the noise estimation function 56 based on the modified noise signal 66 and the noise signal 64.
[0073] In some examples, to identify the noisier segments 65, the processor first calculates a corresponding per-segment noise estimation function for each segment 65 based on the corresponding portion of the noise signal 66. For example, if a particular segment runs from t2 to t3, the processor can calculate the per-segment noise estimation function based on the corresponding portion of the noise signal 66 running from t2 to t3. The processor then clusters the per-segment noise estimation functions, such as using a k-means algorithm or a density-based spatial clustering of noise applications (DBSCAN) algorithm. (For example, if each per-segment noise estimation function includes multiplying by an impulse response vector, as described below with reference to Fig. 9 As described, the processor can cluster the impulse response vectors. ) Next, the processor identifies the segment corresponding to the largest cluster of per-segment noise estimation functions.
[0074] After calculation of the noise estimation function 56, the processor applies the noise estimation function to the (unmodified) noise signal in order to calculate an estimated noise signal.
[0075] For example implementations of the above functions, see Figure 8 , Figure 8 is a flow chart of an algorithm 112 for removing noise from a signal according to some examples of the present disclosure.
[0076] The algorithm 112 begins with a signal receiving step 102 where the processor receives from the circuit a signal 52 ( Figure 2 ) and the noise signal 66 received through one or more other channels. After the signal receiving step 102, the processor segments the signal 52 into a plurality of segments at a segmentation step 114. Subsequently, at a segment identification step 116, the processor identifies the segment with the most noise in the signal. Next, at an invalidation step 118, the processor invalidates the other (less noisy) segments (thereby obtaining an extracted noise signal) and invalidates corresponding portions of the noise signal 66 (thereby obtaining a modified noise signal).
[0077] Next, each subgroup of the other channels is selected at a subgroup selection step 92. For each selected subgroup, the processor calculates a noise estimation function based on the modified noise signal and the extracted noise signal corresponding to the subgroup at a function calculation step 94. The processor then performs an estimate calculation step 104 to calculate a noise estimate based on the corresponding (unmodified) noise signal using the noise estimation function. After the estimate calculation step 104, the processor performs a check step 96.
[0078] When it is determined at the checking step that there are no more subgroups to be selected, algorithm 112 is similar to algorithm 100 ( Figure 6 )
[0079] Compute and use the noise estimation function
[0080] As described above, the processor is configured to calculate one or more noise estimation functions for any given channel, each such function being applicable to one or more noise signals received through corresponding other channels in order to estimate the noise received through the given channel. In some examples, the processor calculates the noise estimation function by calculating a corresponding impulse response vector, such that each noise estimation function includes a multiplication with the impulse response vector.
[0081] In this regard, now refer to Fig. 9 , Fig. 9 is a flow chart of an exemplary calculation 68 of an impulse response vector according to some examples of the present disclosure. The calculation 68 may be performed at Figure 4 and Figure 8 The function calculation step 94 is performed during execution. Fig.10 , Fig.10 is a flow chart of an exemplary calculation 78 of noise estimation according to some examples of the present disclosure. The calculation 78 may be performed at Figure 6 and Figure 8 The estimation calculation step 104 is performed during the execution of the estimation calculation step 104.
[0082] In the description Figures 9 and 10 Previously, the following symbols were introduced:
[0083] • M is the number of noise signals in the subset of noise signals used to compute the impulse response vector (and hence the noise estimation function).
[0084] ·s j (j=1…M) is the jth noise signal in the subgroup, which can be modified relative to the original jth noise signal for the one-stage technique, as described above with reference to Figure 7 described.
[0085] y is a representation of the noise received through a given channel when the noise signal in the subgroup is received. In particular, for the two-stage technique, y is the first noise signal 58 ( Figure 3 ). For the one-stage technique, y is the extracted noise signal 64 ( Figure 7 ).
[0086] h is the impulse response vector.
[0087] z is signal 52 ( Figure 2 ).
[0088] For the two-stage technique, w j is the jth second noise signal 60( Figure 5 ), that is, by receiving s j The second noise signal received on the same channel. For the one-stage technique, w j is the original (unmodified) jth noise signal in the subgroup.
[0089] b is the noise estimate.
[0090] (It should be noted that s j ,y,z,w j Each of a and b is a vector. )
[0091] according to Fig. 9 , calculation 68 begins with a matrix calculation step 70, where the processor calculates the matrix Such that for every matrix in matrix, Next, at the splicing step 72, the processor calculates another matrix A1 as splicing, that is, the processor calculates A1 so that Next, at the pseudo-inverse calculation step 74, the processor calculates the pseudo-inverse A1 of A1 + ; For example, A1 + can be the Moore-Penrose pseudoinverse of A1, that is, Finally, at impulse response calculation step 76, the processor calculates h=A1 + *y.
[0092] refer to Fig.10 , the calculation 78 begins with another matrix calculation step 80, where the processor calculates the matrix Such that for every matrix in matrix, Then, at another splicing step 82, the processor calculates another matrix A2 as splicing, that is, the processor calculates A2 so that Finally, at an impulse response application step 84, the processor calculates b=A2*h.
[0093] In some examples, corresponding signals are received through N channels (at different corresponding times), and for each of the N channels, the removal of noise is based on corresponding noise signals received through N-1 other channels. In such examples, the processor can calculate the impulse response vectors for all channels at once in the form of an impulse response matrix. (Thus, the processor can perform a single function calculation step for all channels, rather than performing multiple calculations as shown in FIG. 1 .) Figure 4 ) Using the impulse response matrix, the processor can remove noise from all signals at once, rather than removing noise from each signal individually. (Thus, after receiving the signals of all channels and the second noise signal, the processor can perform a single estimation calculation step for all channels, rather than performing a single estimation calculation step for all channels. Figure 6 As shown, the estimation calculation step 104 is performed separately for each given channel.)
[0094] For further details, the following notations are introduced:
[0095] · where s j (j=1…N) is the first noise signal 58 received through the jth channel ( Figure 3 (S is therefore of size N×L, where L is the length of each first noise signal in samples.)
[0096] · yes The matrix of .
[0097] · in, yes splicing.
[0098] · where z j (j=1…N) is the signal received through the jth channel. (So the size of Z is N×L.)
[0099] ·w j,k Is receiving z j The kth second noise signal received at time .
[0100] · yes The matrix of .
[0101] · in, yes splicing.
[0102] First, the processor can calculate That is, the pseudo-inverse of B1 (e.g., the Moore-Penrose pseudo-inverse). Next, the processor can calculate the impulse response matrix where h1…h N is the impulse response vector of each of the N channels. The processor can further calculate the estimated noise matrix, such as And then subtract this matrix from Z.
[0103] optimization
[0104] Reference now Fig.11 , Fig.11 is a schematic diagram of an ultrasonic catheter 16 according to some examples of the present disclosure.
[0105] In some examples, removing noise from a given channel does not rely on all other channels, but rather relies only on a group of other channels that does not include all other channels. For example, those channels connected to an ultrasonic transducer that are closer to an ultrasonic transducer connected to a given channel may be excluded from the group because they may carry a greater amount of reflections of ultrasonic waves emitted by the ultrasonic transducer connected to the given channel. Thus, for example, each of the ultrasonic transducers connected to any one of the channels in the group may be farther from the ultrasonic transducer connected to the given channel than at least one of the ultrasonic transducers connected to any one of the channels not in the group.
[0106] For example, since the ultrasonic transducer 44b is relatively close to the ultrasonic transducer 44a, the ultrasonic transducer 44b may receive reflections of ultrasonic waves emitted by the ultrasonic transducer 44a and transduce these reflections, that is, when receiving signals through the channel of the ultrasonic transducer 44a, it may be difficult or impossible to receive noise signals through the channel of the ultrasonic transducer 44b (that is, the channel connected to it). Therefore, a group of channels for removing noise from the channel of the ultrasonic transducer 44a may exclude the channel of the ultrasonic transducer 44b. On the other hand, the group may include the channel of another ultrasonic transducer 44c far from the ultrasonic transducer 44a.
[0107] Alternatively or in addition, the channels of those ultrasonic transducers that are relatively far from the ultrasonic transducer connected to a given channel can be excluded from the group. For example, the channel of another ultrasonic transducer 44d can be excluded from the group of the channels of the ultrasonic transducer 44a.
[0108] In such examples, for higher efficiency, a signal may be received through a given channel and / or a noise signal for removing noise from the given channel may be received from other channels in the group, while another signal including transduction of ultrasound reflections is received through at least one channel that is not in the group.
[0109] For example, in order to calculate the noise estimation function of the channel of ultrasonic transducer 44a, the corresponding first noise signals can be received through the channels of ultrasonic transducer 44a and ultrasonic transducer 44c. When receiving these noise signals, a signal can be received through the channel of ultrasonic transducer 44d, which is not in the group of other channels of the channel of transducer 44a.
[0110] As another example, while a signal is received through a channel of ultrasonic transducer 44a and a noise signal (e.g., a second noise signal for a two-stage technique) is received through a channel of ultrasonic transducer 44c, another signal may be received through a channel of ultrasonic transducer 44b and / or a channel of ultrasonic transducer 44d, each of which is not in the group of other channels of the channel of transducer 44a.
[0111] Alternatively or in addition, for greater efficiency, the processor may cause at least one of the ultrasonic transducers connected to one of the channels in the group to transmit an ultrasonic wave different from the ultrasonic transducer connected to the given channel. For example, the two ultrasonic transducers may transmit at different frequencies, for example using orthogonal frequency division multiplexing. Circuit 47 ( Figure 1 ) can thus receive another signal through the channel, the other signal including the transduction of the ultrasonic reflection. The processor can then separate the noise from the other signal based on the difference in transmission. The processor can then reduce the noise in the signal received through the given channel based on the separated noise.
[0112] For example, the processor may cause the ultrasonic transducer 44a to transmit at a frequency f1, and the ultrasonic transducer 44c to transmit at a frequency f2. The processor may then separate the noise of frequency f1 from the signal received through the channel of the transducer 44c, and use the separated noise to learn the noise estimation function of the channel of the ultrasonic transducer 44a or calculate the estimated noise signal of the channel of the ultrasonic transducer. Similarly, the processor may separate the noise of frequency f2 from the signal received through the channel of the ultrasonic transducer 44a (optionally, while receiving another signal through the channel of the ultrasonic transducer 44c), and use the separated noise to learn the noise estimation function of the channel of the ultrasonic transducer 44c or calculate the estimated noise signal of the channel of the ultrasonic transducer.
[0113] Example
[0114] The following examples relate to various non-exhaustive ways in which the teachings herein may be combined or applied. It should be understood that the following examples are not intended to limit the coverage of any claims that may be provided at any time in this patent application or in the subsequent submissions of this patent application. It is not intended to make a disclaimer. The following examples are provided only for illustrative purposes. It is envisioned that the various teachings herein may be arranged and applied in a variety of other ways. It is also envisioned that some variations may omit certain features mentioned in the following examples. Therefore, any of the aspects or features mentioned below should not be considered decisive unless otherwise explicitly indicated by the inventor or the successor with an interest in the inventor at a later date. If any claim proposed in this patent application or in the subsequent submissions related to this patent application includes additional features other than those mentioned below, these additional features should not be assumed to be added for any reason related to patentability.
[0115] Example 1
[0116] A system (10) for use with a plurality of ultrasonic transducers (44) disposed on a catheter (16) and connected to different respective channels (49) comprises circuitry (47) and a processor (48). The circuitry is configured to receive a signal (52) through any given one of the channels, the signal comprising transduction of ultrasonic reflections (54), and to receive a corresponding noise signal (58) through a group of other ones of the channels, the noise signal comprising transduction of fewer ultrasonic reflections than the signal. The processor is configured to receive the signal and the noise signal from the circuitry and to reduce noise in the signal by calculating one or more noise estimation functions (56) for the given channel based on the noise signal, each of the noise estimation functions being applicable to a respective subset of the noise signals, so as to calculate an estimated noise signal (62) using the noise estimation function and subtract the estimated noise signal from the signal, thereby returning an estimate of the noise received through the given channel when the subset of the noise signals is received.
[0117] Example 2
[0118] According to the system (10) described in embodiment 1,
[0119] wherein the circuit (47) is configured to receive the signal (52) while receiving the noise signal (58),
[0120] The processor (48) is configured to calculate the noise estimation function (56) in the following manner:
[0121] extracting another noise signal (64) from the signal, and
[0122] calculating the noise estimation function based on the received noise signal (58) and the extracted noise signal, and
[0123] Therein, the processor is configured to calculate the estimated noise signal (62) by applying the noise estimation function to the received noise signal.
[0124] Example 3
[0125] The system (10) of embodiment 2, wherein the processor (48) is configured to extract the another noise signal (64) by:
[0126] dividing the signal into a plurality of segments (65),
[0127] identifying one or more of the segments that include transducers that are less ultrasonically reflective than other of the segments, and
[0128] The further noise signal is constructed from the identified segments.
[0129] Example 4
[0130] According to the system (10) described in embodiment 1,
[0131] wherein the noise signal is a first noise signal (58),
[0132] wherein the circuit (47) is configured to receive the signal (52) before or after receiving the first noise signal,
[0133] Wherein, the circuit is further configured as follows:
[0134] receiving, simultaneously with the first noise signal, another first noise signal through the given channel, the another first noise signal including less transduced ultrasound reflections than the signal, and
[0135] while receiving the signal, receiving a corresponding second noise signal (60) through the set of other channels in the channel,
[0136] wherein the processor is further configured to receive the other first noise signal and the second noise signal from the circuit,
[0137] wherein the processor (48) is configured to calculate the noise estimation function (56) based on the first noise signal and based on the further first noise signal, and
[0138] Therein, the processor is configured to calculate the estimated noise signal (62) by applying the noise estimation function to the second noise signal.
[0139] Example 5
[0140] A system (10) according to any one of embodiments 1 to 4, wherein the processor (48) is configured to calculate the noise estimation function (56) by calculating a corresponding impulse response vector.
[0141] Example 6
[0142] The system (10) of embodiment 5, wherein the processor (48) is configured to calculate each of the impulse response vectors by:
[0143] Calculating the Matrix M is the number of the noise signals (58) in the subset of the impulse response vectors, such that for each of the matrices,
[0144] y is a representation of the noise received through the given channel when the noise signal in the subset is received, and
[0145] s j is the jth noise signal among the noise signals in the subgroup;
[0146] Compute another matrix A1 as The splicing,
[0147] Compute the pseudo-inverse of A1 + ,as well as
[0148] The impulse response vector is calculated as A1 + *y.
[0149] Example 7
[0150] The system (10) of embodiment 5, wherein the processor (48) is configured to calculate the estimated noise signal (62) for each of the impulse response vectors by calculating a corresponding noise estimate of the impulse response vector in the following manner:
[0151] Calculating the Matrix M is the number of the noise signals (58) in the subset of the impulse response vectors, such that for each of the matrices,
[0152] z is the signal (52), and
[0153] w j is the jth noise signal among the noise signals in the subgroup, or is another noise signal (60) received through the channel receiving the jth noise signal while receiving the signal,
[0154] Another matrix A2 is calculated as splicing, and
[0155] The noise estimate of the impulse response vector is calculated by multiplying the impulse response vector by A2.
[0156] Example 8
[0157] A system (10) according to any one of embodiments 1 to 7, wherein the group does not include all other channels in the channel (49).
[0158] Example 9
[0159] A system (10) according to Example 8, wherein each of the ultrasonic transducers (44) connected to any one of the channels (49) in the group is farther away from the ultrasonic transducer connected to the given channel than at least one of the ultrasonic transducers connected to any one of the channels not in the group.
[0160] Example 10
[0161] A system (10) according to any one of embodiments 8 to 9, wherein the circuit (47) is configured to receive the noise signal (58) through at least one of the channels (49) that is not in the group, while receiving another signal, the other signal comprising transduction of another ultrasonic reflection.
[0162] Embodiment 11
[0163] A system (10) according to any one of embodiments 8 to 10, wherein the circuit (47) is configured to receive the signal (52) through at least one of the channels (49) that is not in the group, while receiving another signal, the other signal comprising transduction of another ultrasonic reflection.
[0164] Example 12
[0165] The system (10) according to any one of embodiments 1 to 11,
[0166] wherein the processor (48) is further configured to cause at least one of the ultrasonic transducers (44) connected to one of the channels (49) in the group to emit an ultrasonic wave different from that of the ultrasonic transducer connected to the given channel,
[0167] wherein the circuit (47) is further configured to receive another signal through the channel in the group, the other signal comprising a transducer of a reflection of the ultrasonic wave,
[0168] Wherein, the processor is further configured to:
[0169] receiving the other signal from the circuit, and
[0170] separating the noise from the other signal based on the transmission difference, and
[0171] The processor is configured to reduce the noise in the signal (52) based on the separated noise.
[0172] Embodiment 13
[0173] According to the system (10) of any one of embodiments 1 to 12, the one or more noise estimation functions (56) consist of a single noise estimation function, which is applicable to all noise signals in the noise signals (58).
[0174] Embodiment 14
[0175] The system (10) of any one of embodiments 1 to 12, wherein the one or more noise estimation functions (56) include a plurality of noise estimation functions, and wherein the processor (48) is configured to calculate the estimated noise signal (62) by:
[0176] using the noise estimation function (56) to calculate a corresponding noise estimate, and
[0177] The estimated noise signal is calculated based on the noise estimate.
[0178] Embodiment 15
[0179] A method for use with a plurality of ultrasonic transducers (44) disposed on a catheter (16) and connected to different respective channels (49) comprises receiving a signal (52) through any given one of the channels, the signal comprising transduction of ultrasonic reflections (54). The method further comprises receiving a corresponding noise signal (58) through a group of other ones of the channels, the noise signal comprising transduction of fewer ultrasonic reflections than the signal. The method further comprises reducing noise in the signal by calculating one or more noise estimation functions (56) for the given channel based on the noise signal, each of the noise estimation functions being applicable to a respective subset of the noise signals so as to calculate an estimated noise signal (62) using the noise estimation function and subtract the estimated noise signal from the signal, thereby returning an estimate of the noise received through the given channel when receiving the subset of the noise signals.
[0180] It should be understood by those skilled in the art that the present disclosure is not limited to the contents specifically shown and described above. On the contrary, the scope of the present disclosure includes the combinations and sub-combinations of the various features described above, as well as variations and modifications that are not within the scope of the prior art that may occur to those skilled in the art when reading the above description. The documents incorporated by reference into this patent application are considered an integral part of this application, unless any term defined in these incorporated documents conflicts with the definition explicitly or implicitly given in this specification, then only the definition in this specification shall be considered.
Claims
1. A system for use with a plurality of ultrasonic transducers disposed on a catheter and connected to different corresponding channels, the system include: A circuit, the circuit being configured as: receiving a signal through any given one of the channels, the signal comprising transduction of an ultrasonic reflection, and receiving a corresponding noise signal through a group of other ones of the channels, the noise signal including less transduction of ultrasonic reflections than the signal; and A processor, the processor being configured to: receiving the signal and the noise signal from the circuit, and Reduce the noise in the signal by: computing one or more noise estimation functions for the given channel based on the noise signal, each of the noise estimation functions being applicable to a respective subset of the noise signal so as to return an estimate of the noise received through the given channel when the subset of the noise signal is received, calculating an estimated noise signal using the noise estimation function, and The estimated noise signal is subtracted from the signal.
2. The system according to claim 1, in, The circuit is configured to receive the signal while receiving the noise signal, The processor is configured to calculate the noise estimation function in the following manner: extracting another noise signal from the signal, and calculating the noise estimation function based on the received noise signal and the extracted noise signal, and Therein, the processor is configured to calculate the estimated noise signal by applying the noise estimation function to the received noise signal.
3. The system according to claim 2, in, The processor is configured to extract the another noise signal by: dividing the signal into a plurality of segments, identifying one or more of the segments that include transducers that are less ultrasonically reflective than other of the segments, and The further noise signal is constructed from the identified segments.
4. The system according to claim 1, in, The noise signal is a first noise signal, wherein the circuit is configured to receive the signal before or after receiving the first noise signal, Wherein, the circuit is further configured as follows: receiving, simultaneously with the first noise signal, another first noise signal through the given channel, the another first noise signal including less transduced ultrasound reflections than the signal, and while receiving the signal, receiving a corresponding second noise signal through the group of other channels in the channel, wherein the processor is further configured to receive the other first noise signal and the second noise signal from the circuit, wherein the processor is configured to calculate the noise estimation function based on the first noise signal and based on the further first noise signal, and Therein, the processor is configured to calculate the estimated noise signal by applying the noise estimation function to the second noise signal.
5. The system according to claim 1, in, The processor is configured to calculate the noise estimation function by calculating a corresponding impulse response vector.
6. The system according to claim 5, in, The processor is configured to calculate each of the impulse response vectors by: Calculating the Matrix j=1...M, M being the number of the noise signals in the subset of the impulse response vectors, such that for each of the matrices, y is a representation of the noise received through the given channel when the noise signal in the subset is received, and s j is the jth noise signal among the noise signals in the subgroup; Another matrix A 1 Calculated as The splicing, Calculate A 1 The pseudo-inverse A 1 + ,as well as The impulse response vector is calculated as A 1 + *y.
7. The system according to claim 5, in, The processor is configured to calculate, for each of the impulse response vectors, the estimated noise signal by calculating a corresponding noise estimate of the impulse response vector in the following manner: Calculating the Matrix j=1...M, M being the number of the noise signals in the subset of the impulse response vectors, such that for each of the matrices, z is the signal, and w j is the j-th noise signal among the noise signals in the subgroup, or is another noise signal received through the channel receiving the j-th noise signal while receiving the signal, Another matrix A 2 Calculated as splicing, and By multiplying the impulse response vector by A 2 To calculate the noise estimate of the impulse response vector.
8. The system according to claim 1, in, The group does not include all other ones of the channels.
9. The system according to claim 8, in, Each of the ultrasonic transducers connected to any one of the channels in the group is farther from the ultrasonic transducer connected to the given channel than at least one of the ultrasonic transducers connected to any one of the channels not in the group.
10. The system according to claim 8, in, The circuit is configured to receive another signal including transduction of another ultrasound reflection while receiving the noise signal through at least one of the channels that is not in the group.
11. The system according to claim 8, in, The circuit is configured to receive another signal simultaneously with receiving the signal through at least one of the channels that is not in the group, the other signal comprising a transduction of another ultrasound reflection.
12. The system according to claim 1, in, The processor is further configured to cause at least one of the ultrasonic transducers connected to one of the channels in the group to emit an ultrasonic wave different from an ultrasonic transducer connected to the given channel, wherein the circuit is further configured to receive another signal through the channel in the group, the another signal comprising a transducer of a reflection of the ultrasonic wave, Wherein, the processor is further configured to: receiving the other signal from the circuit, and separating the noise from the other signal based on the transmission difference, and The processor is configured to reduce the noise in the signal based on the separated noise.
13. The system according to claim 1, in, The one or more noise estimation functions consist of a single noise estimation function that is applicable to all of the noise signals.
14. The system according to claim 1, in, The one or more noise estimation functions include a plurality of noise estimation functions, and wherein the processor is configured to calculate the estimated noise signal by: computing a corresponding noise estimate using the noise estimation function, and The estimated noise signal is calculated based on the noise estimate.
15. A method for use with a plurality of ultrasound transducers disposed on a catheter and connected to different respective channels, the method include: receiving a signal through any given of the channels, the signal comprising transduction of ultrasonic reflections, receiving a corresponding noise signal through a group of other ones of the channels, the noise signal including less transduction of ultrasonic reflections than the signal; as well as Reduce the noise in the signal by: computing one or more noise estimation functions for the given channel based on the noise signal, each of the noise estimation functions being applicable to a respective subset of the noise signal so as to return an estimate of the noise received through the given channel when the subset of the noise signal is received, calculating an estimated noise signal using the noise estimation function, and The estimated noise signal is subtracted from the signal.
16. The method according to claim 15, in, Receiving the signal includes receiving the signal while receiving the noise signal, Wherein, calculating the noise estimation function comprises: extracting another noise signal from the signal; and calculating the noise estimation function based on the received noise signal and the extracted noise signal, and Wherein, calculating the estimated noise signal comprises calculating the estimated noise signal by applying the noise estimation function to the received noise signal.
17. The method according to claim 16, in, Extracting the another noise signal comprises: dividing the signal into a plurality of segments; identifying one or more of the segments that include less ultrasound reflective transducers than other of the segments; and The further noise signal is constructed from the identified segments.
18. The method according to claim 15, in, The noise signal is a first noise signal, wherein receiving the signal comprises receiving the signal before or after receiving the first noise signal, Wherein, the method further comprises: receiving, simultaneously with the first noise signal, another first noise signal through the given channel, the another first noise signal including less transduced ultrasonic reflections than the signal; and while receiving the signal, receiving a corresponding second noise signal through the group of other channels in the channel, wherein calculating the noise estimation function comprises calculating the noise estimation function based on the first noise signal and based on the further first noise signal, and Wherein calculating the estimated noise signal comprises calculating the estimated noise signal by applying the noise estimation function to the second noise signal.
19. The method according to claim 15, in, Calculating the noise estimation function includes calculating the noise estimation function by calculating a corresponding impulse response vector.
20. The method according to claim 19, in, Calculating each of the impulse response vectors comprises: Calculating the Matrix j=1...M, M being the number of the noise signals in the subset of the impulse response vectors, such that for each of the matrices, y is a representation of the noise received through the given channel when the noise signal in the subset is received, and s j is the jth noise signal among the noise signals in the subgroup; Another matrix A 1 Calculated as splicing; Calculate A 1 The pseudo-inverse A 1 + ;as well as The impulse response vector is calculated as A 1 + *y.
21. The method according to claim 19, in, Calculating the estimated noise signal comprises, for each impulse response vector in the impulse response vectors, calculating a corresponding noise estimate of the impulse response vector by: Calculating the Matrix j=1...M, M being the number of the noise signals in the subset of the impulse response vectors, such that for each of the matrices, z is the signal, and w j is the j-th noise signal among the noise signals in the subgroup, or is another noise signal received through the channel receiving the j-th noise signal while receiving the signal, Another matrix A 2 Calculated as splicing, and By multiplying the impulse response vector by A 2 To calculate the noise estimate of the impulse response vector.
22. The method according to claim 15, in, The group does not include all other ones of the channels.
23. The method according to claim 22, in, Each of the ultrasonic transducers connected to any one of the channels in the group is farther from the ultrasonic transducer connected to the given channel than at least one of the ultrasonic transducers connected to any one of the channels not in the group.
24. The method according to claim 22, in, Receiving the noise signal includes receiving another signal simultaneously with receiving the noise signal through at least one of the channels that is not in the group, the other signal including transduction of another ultrasound reflection.
25. The method according to claim 22, in, Receiving the signal includes receiving another signal simultaneously with receiving the signal through at least one of the channels that is not in the group, the other signal including transduction of another ultrasound reflection.
26. The method according to claim 15, further comprising: include: causing at least one of the ultrasonic transducers connected to one of the channels in the group to emit an ultrasonic wave different from that of an ultrasonic transducer connected to the given channel; receiving another signal through the channel in the group, the another signal comprising a transduction of a reflection of the ultrasonic wave, and separating the noise from the other signal based on the transmission difference, Wherein reducing the noise in the signal comprises reducing the noise in the signal based on the separated noise.
27. The method according to claim 15, in, The one or more noise estimation functions consist of a single noise estimation function that is applicable to all of the noise signals.
28. The method according to claim 15, in, The one or more noise estimation functions include a plurality of noise estimation functions, and wherein calculating the estimated noise signal includes: calculating a corresponding noise estimate using the noise estimation function; and The estimated noise signal is calculated based on the noise estimate.
