Systems and methods for noise reduction in imaging
The method addresses noise interference in medical imaging by separating and correcting in-band noise using out-of-band or reference detection, improving signal quality and resolution in ultrasound and MRI systems.
Patent Information
- Application Number
- KR1020247026556
- Authority / Receiving Office
- KR · KR
- Patent Type
- Patents
- Current Assignee / Owner
- Priority Date
- 2017-02-24
- Filing Date
- 2018-02-23
- Publication Date
- 2026-11-11
- Estimated Expiration
- 2038-02-23
AI Technical Summary
Medical imaging techniques such as ultrasound and MRI face challenges in maintaining high signal-to-noise ratios due to weak ultrasound signals and noise interference from various sources, which affect image quality and resolution.
A method for denoising imaging signals involves detecting energy waves to obtain noise characteristic waveforms, filtering them to separate in-band and out-of-band noise, and applying correction factors or adaptive filtering to reduce in-band noise, using reference receiving circuits to estimate and subtract noise.
This approach effectively reduces noise in imaging signals, enhancing signal quality and maintaining resolution by estimating and correcting in-band noise based on out-of-band or reference noise detection.
Smart Images

Figure 112024085926718-PAT00134_ABST
Abstract
Description
Technology Field
[0001] This application claims priority to U.S. Provisional Application No. 62 / 463,431, filed on February 24, 2017, titled "SYSTEMS AND METHODS FOR NOISE REDUCTION IN IMAGING," the entire contents of which are incorporated herein by reference.
[0002] The present disclosure relates to imaging techniques and the processing of imaging data for noise removal. Background Technology
[0003] Medical imaging using ultrasound and MRI relies on detecting low-amplitude signals in the radio frequency spectrum, typically spanning from 2 MHz to 200 MHz. Image quality is significantly affected by the signal-to-noise ratio.
[0004] In IVUS (intravascular ultrasound), ICE (intracardiac echocardiography), and other forms of minimally invasive ultrasound, ultrasound transducers detect ultrasound signals from surrounding structures and convert acoustic energy into electrical signals. These signals are then transmitted along one or more conductive channels (such as coaxial conductors, twisted pair conductors, flex circuits, etc.). For many reasons (including cost, manufacturability, safety, biocompatibility, thermal issues, and power supply requirements), parts of minimally invasive imaging probes that can be inserted intracorporeally often do not include amplifiers to boost signal intensity. The electrical signals detected by minimally invasive ultrasound transducers can be very small (< 10 mV and more typically < 1 mV), and a significant portion of the information regarding tissue structures that can be imaged by ultrasound tends to be in the lower portion of the dynamic range of the detected electrical signals. The signal amplitude of the received ultrasound signal is limited by the mechanical efficiency of the transducer, the low amplitude of the detected acoustic signals, the small size of the transducer, and attenuation along the conductors carrying the electrical signal from the transducer out of the body. Considering this, the signals in minimally invasive ultrasound imaging systems tend to be very weak.
[0005] Noise can be introduced into the system from many sources, including wireless transmitters, power electronic devices, transmission lines, switching transistors, and other things known in the art. Noise can be introduced directly through induction by components sensitive to electromagnetic interference, or through conduction and suboptimal isolation. Some of the noise may be generated by components within the imaging system itself, such as pulse width modulators for scanning actuators and motor controllers, switched-mode power supplies, clocking circuits, and transistors in any of the electronic components of the imaging system. Furthermore, impedance monitors, (Carto ® 3, Carto ® Tracking systems (such as those found in XP or NavX™ systems), temperature sensors, infusion pumps, ablation systems, ECGs, and hemodynamic monitors, and other systems coupled to the patient or in the procedural environment can introduce noise. RFID inventory management systems used in some clinical areas can also introduce noise.
[0006] Several approaches relate to reducing the amount of noise entering the ultrasonic receiving circuitry of ultrasonic imaging systems, including the selection of system components that generate minimal RF noise, electrical isolation, shielding, proper grounding, and physically separating noise-generating components from those vulnerable to electromagnetic noise. These approaches are often difficult to implement because noise sources often possess preferred characteristics for other reasons (e.g., pulse width modulation motor controllers are energy-efficient and have good response times) or are difficult to physically isolate from one another (e.g., it may be desirable to have power electronics in close proximity to the imaging probe or its associated circuitry).
[0007] Other approaches to reduce the impact of noise on ultrasonic signal quality (and thus ultrasonic image quality) include filtering and image processing. Ultrasonic signals typically have known bandwidths, and detected ultrasonic signals can be filtered using analog or digital filtering techniques (often a combination of both). Analog or digital filtering may be applied to limit portions of the electrical signal output from the ultrasonic receiving circuit to those frequencies within the operating bandwidth (or harmonics) of the ultrasonic transducer. Selecting filters with narrow bandwidths and sharp cutoffs can reduce the amount of allowable noise in signals used to generate images or to use ultrasonic signals in other ways (such as for spectral analysis of Doppler measurements of the ultrasonic signal or evaluation of scatterer flow in the sonicated field). Notch or comb filters help remove narrowband noise within the imaging frequency range. Overly aggressive filtering can have the undesirable effect of reducing the amount of signal power accepted for generating images or for other uses of ultrasound signals. This can also negatively impact other performance aspects of the ultrasound imaging system, such as resolution. However, if the passband of the filters is too wide, more noise is accepted into the system.
[0008] Image processing can further reduce noise by filtering the generated image data, such as by averaging or removing outlier values. For example, such filtering can be applied within the image in the spatial domain by applying a Gaussian filter to a pixel and its neighboring pixels to blur or smooth out any random noise in the image. Unfortunately, this tends to reduce the spatial resolution of the image. Similarly, spatial domain filtering can be applied to structures being imaged that do not move rapidly relative to the frame repetition frequency of the imaging modality. For example, a pixel in an image frame may be the average or Gaussian-filtered result of pixels at similar locations in one or more preceding and / or trailing frames.
[0009] Similar problems apply to MRI imaging systems, where weak signals are detected in the presence of noise from unwanted radio frequency energy sources.
[0010] Very useful are methods, systems, and devices for identifying noise and actively removing noise from one or more imaging signals.
[0011] Many forms of noise enter the ultrasound receive signal chain, and especially if the noise is broadband by nature, once it enters the system, it can be difficult to remove, where some of the noise is within the passband of the ultrasound system. For example, in an imaging system having a transducer with a center frequency of 10 MHz and a passband of 7.5 to 12.5 MHz, the system may be designed to significantly filter out any portion of the noise below 7.5 MHz and any portion of the noise above 12.5 MHz. Unfortunately, the amplitude of the noise within the 7.5 to 12.5 MHz passband can often be substantial relative to the amplitude of the ultrasound signal being detected.
[0012] Many noise sources are generated as a result of rapid transients, such as when a field-effect transistor or switch is turned on or off. Electrical signals with rapid transients have a very wide frequency domain representation that may span all or part of the passband of the ultrasonic receiving signal chain. This is especially true in power supplies or pulse width modulation circuits where the noise can have an amplitude strong enough to compete with the signal being detected.
[0013] One approach to reducing broadband noise utilizes the fact that while imaging energy is primarily within a selective imaging band, broadband noise can be detected both within and outside the imaging band. In principle, by detecting or characterizing noise outside the imaging band at any given time, the broadband noise that may be within the imaging band can be estimated, and the detected signal can be modified to reduce the estimated in-band noise. By effectively generating an estimate of the in-band noise based on the out-of-band noise, a signal can be generated that estimates the desired imaging energy in the absence of the estimated in-band noise.
[0014] Another approach to reducing noise within the imaging band is to use reference receiving circuits (including resistors, capacitors, inductors, transmission lines, amplifiers, transformers, disabled transducers, or components capable of emulating transducer receiving circuits) that are sensitive to noise within the imaging band but isolated from imaging energy. By estimating the in-band imaging noise based on the in-band noise received by the reference receiving circuit, signals can be generated that estimate the desired imaging energy in the absence of the estimated in-band noise.
[0015] In one embodiment, a method for denoising imaging signals detected in the presence of broadband noise is provided, the method being:
[0016] When imaging energy is not received, the method comprises the step of detecting energy waves with an imaging transducer receiving circuit to obtain a noise characteristic waveform thereby, and filtering the noise characteristic waveform to generate an in-band noise characteristic waveform existing within the imaging band and an out-of-band noise characteristic waveform existing within a noise detection band that is at least partially outside the imaging band;
[0017] A step of segmenting in-band noise characterization waveforms and out-of-band noise characterization waveforms according to one or more time windows;
[0018] For at least one time window, a step of processing an in-band noise characterization waveform and an out-band noise characterization waveform to determine the relationship between noise in the imaging band and noise in the noise detection band;
[0019] A step of detecting imaging signals with an imaging transducer receiving circuit and thereby acquiring one or more imaging waveforms;
[0020] For at least one imaging waveform:
[0021] a) a step of filtering the imaging waveform to generate an in-band imaging waveform existing within the imaging band and an out-of-band noise detection imaging waveform existing within the noise detection band;
[0022] b) a step of segmenting in-band imaging waveforms and out-band noise detection imaging waveforms according to one or more time windows;
[0023] c) a step of estimating a measure associated with the amount of noise in an in-band imaging waveform within at least one time window using the relationship and out-of-band noise detection imaging waveform; and
[0024] d) For at least one time window processed in c), the method includes the step of applying a denoising correction to a portion of the in-band imaging waveform within the time window.
[0025] In another embodiment, a method for denoising imaging signals detected in the presence of noise is provided, the method being:
[0026] When not receiving imaging energy:
[0027] A step of detecting energy waves with an imaging transducer receiving circuit, thereby obtaining a noise characterization waveform, and filtering the noise characterization waveform to generate an in-band noise characterization waveform existing within the imaging band; and
[0028] A step of detecting noise with a reference receiving circuit configured to avoid transduction of imaging energy while detecting noise received by an imaging transducer receiving circuit, thereby obtaining a reference noise characteristic waveform;
[0029] A step of processing an in-band noise characterization waveform and a reference noise characterization waveform to determine the relationship between noise in the imaging band and noise detected by a reference receiving circuit;
[0030] A step of detecting imaging signals with an imaging transducer receiving circuit and thereby acquiring one or more imaging waveforms;
[0031] For at least one imaging waveform:
[0032] a) a step of filtering the imaging waveform to generate an in-band imaging waveform existing within the imaging band;
[0033] b) A step of detecting a reference noise detection waveform using a reference receiving circuit;
[0034] c) a step of segmenting the in-band imaging waveform and the reference noise detection waveform according to one or more time windows;
[0035] d) a step of estimating a measure associated with the amount of noise in an in-band imaging waveform within at least one time window using the relationship and reference noise detection waveform; and
[0036] e) For at least one time window processed in c), the step of applying denoising correction to a portion of the in-band imaging waveform within the time window is included.
[0037] In another embodiment, a method for denoising imaging signals detected in the presence of broadband noise is provided, the method being:
[0038] A step of detecting energy waves with an imaging transducer receiving circuit to obtain an imaging waveform thereby, and filtering the imaging waveform to generate an in-band imaging waveform existing within the imaging band and an out-of-band noise detection imaging waveform existing within a noise detection band that is at least partially outside the imaging band;
[0039] A step of detecting the in-band imaging envelope of an in-band imaging waveform;
[0040] A step of detecting the out-of-band envelope of an imaging waveform for out-of-band noise detection;
[0041] A step of applying a scaling factor to an out-of-band envelope to obtain an out-of-band envelope modified thereby; and
[0042] The method includes the step of obtaining a noise-corrected in-band envelope by combining a modified out-of-band envelope and an in-band imaging envelope;
[0043] Here, the scaling factor is selected to reduce the contribution of in-band noise in the noise-corrected in-band envelope.
[0044] In another embodiment, a method for denoising imaging signals detected in the presence of noise is provided, the method being:
[0045] A step of detecting energy waves with an imaging transducer receiving circuit to obtain an imaging waveform thereby, and filtering the imaging waveform to generate an in-band imaging waveform existing within the imaging band and an out-of-band noise detection imaging waveform existing within a noise detection band that is at least partially outside the imaging band;
[0046] A step of obtaining a modified waveform by applying a frequency shift and amplitude scaling factor to an out-of-band noise detection imaging waveform so that the modified waveform includes frequency components existing within the imaging band; and
[0047] The method includes the step of obtaining a noise-corrected in-band imaging waveform by combining a modified waveform and an in-band imaging waveform;
[0048] Here, the amplitude scaling factor is selected to reduce the contribution of in-band noise in the noise-corrected in-band imaging waveform.
[0049] In another embodiment, a method for denoising imaging signals detected in the presence of noise is provided, the method being:
[0050] A step of detecting energy waves with an imaging transducer receiving circuit, thereby acquiring an imaging waveform, and filtering the imaging waveform to generate an in-band imaging waveform existing within the imaging band; and
[0051] A step of detecting noise with a reference receiving circuit configured to avoid transduction of imaging energy while detecting noise received by an imaging transducer receiving circuit, thereby obtaining a reference noise detection waveform;
[0052] A step of detecting the in-band imaging envelope of an in-band imaging waveform;
[0053] A step of detecting the reference envelope of a reference noise detection waveform;
[0054] A step of applying a scaling factor to a reference envelope to obtain a reference envelope modified thereby; and
[0055] The method includes the step of obtaining a noise-corrected in-band envelope by combining a modified reference envelope and an in-band imaging envelope;
[0056] Here, the scaling factor is selected to reduce the contribution of in-band noise in the noise-corrected in-band envelope.
[0057] In another embodiment, a method for denoising imaging signals detected in the presence of noise is provided, the method being:
[0058] A step of detecting energy waves with an imaging transducer receiving circuit, thereby acquiring an imaging waveform, and filtering the imaging waveform to generate an in-band imaging waveform existing within the imaging band; and
[0059] A step of detecting noise with a reference receiving circuit configured to avoid transduction of imaging energy while detecting noise received by an imaging transducer receiving circuit, thereby obtaining a reference noise detection waveform;
[0060] A step of adaptively filtering a reference noise detection waveform according to one or more adaptive filter parameters; and
[0061] The method includes the step of obtaining a noise-corrected in-band imaging waveform by combining a filtered reference noise detection waveform and an in-band imaging waveform;
[0062] Here, the adaptive filter parameters are actively determined by processing the noise-corrected in-band imaging waveform to minimize the power of the noise-corrected in-band imaging waveform.
[0063] In another embodiment, a method for denoising imaging signals detected in the presence of broadband noise is provided, the method being:
[0064] A step of detecting energy waves with an imaging transducer receiving circuit to obtain an imaging waveform thereby, and filtering the imaging waveform to generate an in-band imaging waveform existing within the imaging band and an out-of-band noise detection imaging waveform existing within a noise detection band that is at least partially outside the imaging band;
[0065] A step of detecting the in-band imaging envelope of an in-band imaging waveform;
[0066] A step of detecting the out-of-band imaging envelope of an out-of-band noise detection imaging waveform;
[0067] A step of adaptively filtering an out-of-band imaging envelope according to one or more adaptive filter parameters; and
[0068] The method includes the step of obtaining a noise-corrected in-band imaging envelope by combining a filtered out-of-band imaging envelope and an in-band imaging envelope;
[0069] Here, the adaptive filter parameters are actively determined by processing the noise-corrected in-band imaging envelope to minimize the power of the noise-corrected in-band imaging envelope.
[0070] In another embodiment, a method for denoising imaging signals detected in the presence of broadband noise is provided, the method being:
[0071] A step of detecting energy waves with an imaging transducer receiving circuit to obtain an imaging waveform thereby, and filtering the imaging waveform to generate an in-band imaging waveform existing within the imaging band and an out-of-band noise detection imaging waveform existing within a noise detection band that is at least partially outside the imaging band;
[0072] A step of obtaining a modified waveform by applying a frequency shift to an out-of-band noise detection imaging waveform so that the modified waveform includes frequency components existing within the imaging band;
[0073] A step of adaptively filtering an out-of-band noise detection imaging waveform according to one or more adaptive filter parameters; and
[0074] The method includes the step of obtaining a noise-corrected in-band imaging waveform by combining a filtered modified waveform and an in-band imaging waveform;
[0075] Here, the adaptive filter parameters are actively determined by processing the noise-corrected in-band imaging waveform to minimize the power of the noise-corrected in-band imaging waveform.
[0076] In another embodiment, a method for denoising imaging signals detected in the presence of broadband noise is provided, the method being:
[0077] A step of detecting energy waves with an imaging transducer receiving circuit to obtain an imaging waveform thereby, and filtering the imaging waveform to generate an in-band imaging waveform existing within the imaging band and an out-of-band noise detection imaging waveform existing within a noise detection band that is at least partially outside the imaging band;
[0078] A step of processing an out-of-band noise detection imaging waveform to select appropriate filter parameters of a dynamic digital filter for filtering an in-band imaging waveform to remove in-band noise; and
[0079] It includes a step of filtering an in-band imaging waveform with a dynamic digital filter according to filter parameters.
[0080] In another embodiment, a method for performing noise reduction on signals acquired by a detection system characterized by one or more noise sources is provided, the method comprising:
[0081] When imaging energy is not received, the method comprises the step of detecting energy waves with an imaging transducer receiving circuit to obtain a noise characteristic waveform thereby, and filtering the noise characteristic waveform to generate an in-band noise characteristic waveform existing within the imaging band and an out-of-band noise characteristic waveform existing within a noise detection band that is at least partially outside the imaging band;
[0082] A step of segmenting in-band noise characterization waveforms and out-of-band noise characterization waveforms according to one or more time windows;
[0083] For at least one time window, a step of processing in-band noise characterization waveforms and out-of-band noise characterization waveforms according to a pattern recognition algorithm to identify noise patterns within a noise detection band that correspond to noise in an imaging band;
[0084] A step of detecting imaging signals using an imaging transducer receiving circuit to obtain an imaging waveform thereby, and filtering the imaging waveform to obtain an in-band imaging waveform existing within the imaging band and an out-of-band noise detection imaging waveform existing within the noise detection band;
[0085] A step of segmenting in-band imaging waveforms and out-band imaging waveforms according to one or more time windows;
[0086] For at least one time window, a step of processing an out-of-band noise detection imaging waveform according to a pattern recognition algorithm to detect a noise pattern; and
[0087] In the case of detecting a noise pattern, the method includes the step of applying denoising correction to a time window of an in-band imaging waveform specific to the noise pattern detected in an out-of-band noise detection imaging waveform.
[0088] In another embodiment, a method for performing noise reduction on signals acquired by a detection system characterized by one or more known noise sources is provided, the method comprising:
[0089] When not receiving imaging energy:
[0090] A step of detecting energy waves with an imaging transducer, thereby obtaining a noise characterization waveform, and filtering the noise characterization waveform to generate an in-band noise characterization waveform existing within the imaging band; and
[0091] A step of detecting noise with a reference receiving circuit configured to avoid transduction of imaging energy while detecting noise received by an imaging transducer receiving circuit, thereby obtaining a reference noise characteristic waveform;
[0092] A step of segmenting an in-band noise characterization waveform and a reference noise characterization waveform according to one or more time windows;
[0093] A step of processing an in-band noise characterization waveform and a reference noise characterization waveform to determine the relationship between noise in the imaging band and noise detected by a reference receiving circuit for at least one time window;
[0094] A step of processing an in-band noise characterization waveform and a reference noise characterization waveform according to a pattern recognition algorithm to identify the existence of a noise pattern within a reference noise characterization waveform that is correlated with the noise in the in-band noise characterization waveform;
[0095] A step of detecting imaging signals with an imaging transducer receiving circuit to acquire an imaging waveform, while also detecting a reference noise detection waveform with a reference receiving circuit, and filtering the imaging waveform to acquire an in-band imaging waveform existing within the imaging band;
[0096] A step of segmenting in-band imaging waveforms and reference noise detection waveforms according to one or more time windows;
[0097] For at least one time window, a step of processing a reference noise detection waveform according to a pattern recognition algorithm to detect the presence of a noise pattern; and
[0098] In the case of detecting a noise pattern, the method includes the step of applying denoising correction to a time window of an in-band imaging waveform specific to the noise pattern detected from a reference noise detection waveform.
[0099] In another embodiment, a method for denoising imaging signals detected in the presence of noise is provided, the method being:
[0100] When imaging energy is not received, the method comprises the step of detecting energy waves with an imaging transducer receiving circuit to obtain a noise characteristic waveform thereby, and filtering the noise characteristic waveform to generate an in-band noise characteristic waveform existing within the imaging band and an out-of-band noise characteristic waveform existing within a noise detection band that is at least partially outside the imaging band;
[0101] A step of detecting imaging signals using an imaging transducer receiving circuit, thereby acquiring an imaging waveform, and filtering the imaging waveform to generate an in-band imaging waveform existing within the imaging band and an out-of-band noise detection imaging waveform existing within the noise detection band;
[0102] A step of performing cross-correlation between an out-band imaging waveform and an out-band noise characterization waveform to determine the time delay associated with the maximum cross-correlation; and
[0103] It includes the step of applying a time delay and amplitude adjustment to an in-band noise characteristic waveform to obtain a modified in-band noise characteristic waveform, and subtracting the modified in-band noise characteristic waveform from an in-band imaging waveform.
[0104] In another embodiment, a method for denoising imaging signals detected in the presence of noise is provided, the method being:
[0105] When not receiving imaging energy:
[0106] A step of detecting energy waves with an imaging transducer receiving circuit, thereby obtaining a noise characterization waveform, and filtering the noise characterization waveform to generate an in-band noise characterization waveform existing within the imaging band; and
[0107] A step of detecting noise with a reference receiving circuit configured to avoid transduction of imaging energy while detecting noise received by an imaging transducer receiving circuit, thereby obtaining a reference noise characterization waveform;
[0108] A step of detecting imaging signals with an imaging transducer receiving circuit to acquire an imaging waveform, while also detecting a reference noise detection waveform with a reference receiving circuit, and filtering the imaging waveform to acquire an in-band imaging waveform existing within the imaging band;
[0109] A step of performing cross-correlation between a reference noise detection waveform and a reference noise characterization waveform to determine the time delay associated with the maximum cross-correlation; and
[0110] It includes the step of applying a time delay and amplitude adjustment to an in-band noise characteristic waveform to obtain a modified in-band noise characteristic waveform, and subtracting the modified in-band noise characteristic waveform from an in-band imaging waveform.
[0111] In another embodiment, a method for denoising imaging signals detected in the presence of noise is provided, the method being:
[0112] A step of detecting imaging signals along a plurality of adjacent scan lines using an imaging transducer receiving circuit, thereby acquiring a plurality of imaging waveforms;
[0113] For at least two adjacent scan lines:
[0114] A step of filtering the respective associated imaging waveforms to generate an in-band imaging waveform existing within the imaging band and an out-band noise detection imaging waveform existing within the noise detection band that is at least partially outside the imaging band;
[0115] A step of segmenting in-band imaging waveforms and out-band noise detection imaging waveforms according to a series of time windows;
[0116] For at least one window:
[0117] A step of processing an out-of-band noise detection imaging waveform to determine whether a corresponding windowed portion of an in-band imaging waveform should be noise corrected; and
[0118] When the in-band imaging waveform within the time window is considered suitable for noise correction, a step of applying denoising correction to the in-band imaging waveform within the time window—the denoising correction for each sample within the window is based on one or more statistical measures associated with samples in in-band imaging waveforms from two or more adjacent windows, and each adjacent window exists on its respective adjacent scan line—; and
[0119] It includes the step of generating an image based on multiple scan lines and their respective associated denoised in-band imaging waveforms.
[0120] In another embodiment, a method for denoising imaging signals detected in the presence of noise is provided, the method being:
[0121] A step of detecting energy waves with an imaging transducer receiving circuit to obtain an imaging waveform thereby, and filtering the imaging waveform to generate an in-band imaging waveform existing within the imaging band and an out-of-band noise detection imaging waveform existing within a noise detection band that is at least partially outside the imaging band;
[0122] A step of detecting imaging signals along a plurality of adjacent scan lines with an imaging transducer receiving circuit and thereby acquiring a plurality of imaging waveforms;
[0123] A step of processing one or more out-of-band noise detection imaging waveforms to determine the periodicity of the noise source;
[0124] A step of adjusting the scan rate so that noise in in-band imaging waveforms associated with adjacent scan lines is not temporally synchronized;
[0125] For at least two adjacent scan lines:
[0126] A step of segmenting an in-band imaging waveform according to a series of time windows;
[0127] For at least one window:
[0128] A step of applying denoising correction to in-band imaging waveforms within a time window - the denoising correction for each sample within the window is based on one or more statistical measures associated with samples in in-band imaging waveforms from two or more adjacent windows, each adjacent window existing on its respective adjacent scan line -; and
[0129] It includes the step of generating an image based on multiple scan lines and their respective associated denoised in-band imaging waveforms.
[0130] In another embodiment, a method for denoising imaging signals detected in the presence of noise is provided, the method being:
[0131] For at least two adjacent scan lines:
[0132] A step of detecting imaging signals with an imaging transducer receiving circuit to acquire an imaging waveform, while also detecting a reference noise detection waveform with a reference receiving circuit, and filtering the imaging waveform to acquire an in-band imaging waveform existing within the imaging band;
[0133] A step of segmenting the in-band imaging waveform and the reference noise detection waveform according to a series of time windows;
[0134] For at least one window:
[0135] A step of processing a reference noise detection waveform to determine whether a corresponding windowed portion of an in-band imaging waveform needs to be noise corrected; and
[0136] When the in-band imaging waveform within the time window is considered suitable for noise correction, a step of applying denoising correction to the in-band imaging waveform within the time window—the denoising correction for each sample within the window is based on one or more statistical measures associated with samples in in-band imaging waveforms from two or more adjacent windows, and each adjacent window exists on its respective adjacent scan line—; and
[0137] It includes the step of generating an image based on multiple scan lines and their respective associated denoised in-band imaging waveforms.
[0138] Further understanding of the functional and advantageous aspects of the present disclosure can be achieved by referring to the following detailed description and drawings. Brief explanation of the drawing
[0139] Examples will now be described by way of example with reference to the drawings: Figure 1a illustrates an example of an ultrasonic imaging system configured for noise suppression. FIG. 1b illustrates an example of a conventional ultrasonic receiving signal chain for processing an ultrasonic signal before converting it into an ultrasonic image. FIG. 1c illustrates an example of an ultrasound imaging system including an in-body imaging probe. FIG. 1d illustrates an example of an ultrasonic imaging system including a reference transducer for detecting in-band noise. FIG. 1e illustrates an example of an ultrasonic imaging system comprising a second imaging transducer having an imaging band outside the imaging band of a first imaging transducer. The second imaging transducer may be part of a circuit suitable for detecting in-band noise affecting signals received from the first imaging transducer. FIG. 1f illustrates an example of an ultrasonic imaging system including a reference receiving circuit for detecting in-band noise, wherein the reference receiving circuit extends to a location within the imaging probe. FIG. 1g illustrates an example of an ultrasonic imaging system including a reference receiving circuit for detecting in-band noise, wherein the reference receiving circuit is located in one or more parts of the system outside the imaging probe. FIG. 2a illustrates an exemplary system configuration for noise reduction on the envelope of an input waveform through the estimation and suppression of estimated in-band noise, wherein the in-band noise is estimated by performing envelope detection of the out-of-band waveform, followed by delay, scaling, and shape adjustment prior to subtraction. FIG. 2b illustrates an exemplary system configuration for noise reduction of an input waveform through the estimation and subtraction of in-band noise, wherein the in-band noise is estimated by frequency-shifting the out-of-band waveform, filtering the frequency-shifted out-of-band waveform, and subsequent delay, scaling, and shape adjustment prior to subtraction. FIG. 2c illustrates an example of a system configuration for noise reduction using a reference receiving circuit that is sensitive to some or all of the in-band noise, which is at least partially isolated from the imaging signals detected by the imaging transducer receiving circuit, but which is sensitive to the imaging transducer receiving circuit. Subtracting the noise signals detected by the reference receiving circuit from the signals received by the imaging transducer receiving circuit reduces the noise in the output signal. FIG. 3a illustrates an example of a system configuration for noise reduction through active noise cancellation, wherein the cancellation waveform for active noise cancellation is obtained from a reference receiving circuit. FIG. 3b illustrates an example of a system configuration for noise reduction of an input waveform through active noise cancellation, wherein the cancellation waveform for active noise cancellation is obtained by envelope detection of an out-of-band waveform. FIG. 3c illustrates an example of a system configuration for noise reduction of an input waveform through active noise cancellation, wherein the cancellation waveform for active noise cancellation is obtained by frequency-shifting an out-of-band waveform and filtering the frequency-shifted out-of-band waveform. FIG. 4 illustrates an example of a system configuration for noise reduction of an input waveform through filtering of an in-band waveform, wherein the filtering is controlled based on feedback parameters obtained by a filter update algorithm that determines one or more parameters of a filter based on one or more characteristics of a noise detection waveform. FIG. 5a schematically illustrates an exemplary system configuration for noise reduction of an input waveform by applying one or more noise reduction algorithms based on noise parameters acquired during a first measurement stage when imaging signals are absent, and using the aforementioned noise parameters during a second measurement stage when imaging signals are collected. FIG. 5b schematically illustrates an exemplary alternative system in which a reference receiving channel is used to detect in-band noise. FIGS. 6a and 6c illustrate an exemplary system configuration for noise reduction for an input waveform based on the detection of noise in an out-of-band waveform, wherein different time windows of an in-band waveform are suppressed based on the processing of respective windows of an out-of-band waveform, wherein the noise windows of the in-band waveform are corrected by subtracting using a subtrahend value that depends on the amount of power within the windows of the out-of-band waveform. Figure 6b illustrates an exemplary scatter plot showing the signal power of the in-band waveform windows versus the signal power of the respective out-band waveform windows in the noise characterization stage. FIG. 6d illustrates an exemplary system configuration for noise reduction for an input waveform based on the detection of noise in an out-of-band waveform, wherein different time windows of an in-band waveform are suppressed based on the processing of each window of an out-of-band waveform, wherein the noise windows of the in-band waveform are corrected by multiplying by an attenuation factor that depends on the amount of power within the window of the out-of-band waveform. Figures 6e and 6f illustrate charts regarding a method in which different time windows of in-band waveforms are initially identified as primarily signal or noise, and then noise windows surrounded by signal windows are identified as possibly incorrect and reclassified, and conversely, signal windows surrounded by noise windows are identified as incorrect and reclassified as noise. FIGS. 6g and FIGS. 6h illustrate an exemplary system configuration for noise reduction for an input waveform based on the detection of noise in a filtered reference waveform measured by a reference receiving channel, wherein different time windows of the in-band waveform are suppressed based on the processing of the respective windows of the reference waveform, and wherein the noise windows of the in-band waveform are corrected by subtracting using a reduction value that depends on the amount of power within the window of the filtered reference waveform. FIG. 6i illustrates an exemplary system configuration for noise reduction for an input waveform based on noise measured in a filtered reference waveform measured by a reference receiving channel, wherein different time windows of an in-band imaging waveform are corrected based on the processing of each window of a filtered reference waveform, wherein the noise windows of the in-band imaging waveform are corrected by an attenuation factor that depends on the amount of power within the window of the filtered reference waveform. FIGS. 7a and 7b illustrate an exemplary system configuration for noise reduction of an input waveform based on noise detected in one or more noise detection waveforms—at least one of which includes signals outside the band of the imaging band. Different time windows of the in-band waveform undergo noise reduction according to one or more patterns identified by processing one or more noise detection waveforms. FIGS. 7c and 7d illustrate an exemplary system configuration for reducing an input waveform based on noise detected in a reference waveform, wherein different time windows of an in-band imaging waveform undergo noise reduction according to one or more patterns identified by processing one or more reference waveforms. FIGS. 8a and 8b illustrate an exemplary system configuration for noise reduction for an input waveform based on noise detected in an out-of-band waveform, wherein different time windows of the in-band waveform undergo noise reduction according to the estimated in-band noise that is temporally aligned prior to reduction. FIGS. 8C and 8D illustrate an exemplary system configuration for noise reduction of an input waveform based on noise detected in a filtered reference waveform, wherein different time windows of the in-band waveform undergo noise reduction according to estimated in-band noise that is temporally aligned prior to reduction. FIG. 8e illustrates an exemplary system configuration for noise reduction of an input waveform based on noise detected in an out-of-band waveform, in which measures from adjacent or replicate scan lines are used when performing noise correction. FIG. 8f illustrates an exemplary system configuration for noise reduction of an input waveform based on noise detected in a filtered reference waveform, wherein scales from adjacent or duplicated scan lines are used when performing noise correction. Figure 9 illustrates an example of a magnetic resonance imaging system configured for noise suppression. FIGS. 10a to 10c illustrate exemplary images obtained using an intra-cardiac echo system showing (A) an image obtained in the absence of a noise source; (B) an image obtained in the presence of noise generated through an electroanatomic mapping system; and (C) an image obtained in the presence of noise generated from an ablation generator. FIGS. 11a and FIGS. 11b illustrate images obtained in the presence of noise from an electroanatomical mapping system, (A) without noise reduction and (B) with noise reduction. FIGS. 11c to 11e illustrate images obtained in the presence of noise from an electroanatomical mapping system after noise reduction by attenuation, where the relaxation parameter was set to 0.5(C), 1(D), and 1.5(E). FIGS. 12a and FIGS. 12b illustrate images obtained in the presence of noise from an ablation generator, (A) without the application of a noise reduction method and (B) with the application of a noise reduction method. FIGS. 13a and FIGS. 13b illustrate images obtained in the presence of noise from a magnetic tracking system, (A) without the application of a noise reduction method and (B) with the application of a noise reduction method. FIG. 14 illustrates phrases used to refer to waveforms in the imaging band, noise detection band, and waveforms from the reference receiving circuit. Specific details for implementing the invention
[0140] Various embodiments and aspects of the present disclosure will be described with reference to the details discussed below. The following description and drawings are illustrative of the present disclosure and should not be construed as limiting. Numerous specific details are described to provide a thorough understanding of the various embodiments of the present disclosure. However, in certain cases, well-known or prior details are not described to provide a brief discussion of the embodiments of the present disclosure.
[0141] As used herein, the terms “comprising” and “comprising” should be interpreted as inclusive, open, and non-exclusive. Specifically, as used in the specification and claims, the terms “comprising” and “comprising” and variations thereof mean that the specified features, steps, or components are included. These terms should not be interpreted as excluding the presence of other features, steps, or components.
[0142] As used herein, the term "exemplary" means "serving as an example, illustration, or model" and should not be interpreted as being more desirable or advantageous than other configurations disclosed herein.
[0143] As used herein, the terms “about” and “approximately” are intended to cover variations that may exist in the upper and lower limits of ranges of values, such as variations in attributes, parameters, and dimensions. Unless otherwise specified, the terms “about” and “approximately” mean + / - 25% or less.
[0144] Unless otherwise specified, it should be understood that any specified range or group is a shortened form of referring to each and every possible sub-range or sub-group encompassed therein, as well as referring individually to each and every member of each range or group, and is similarly applicable to any sub-ranges or sub-groups within it. Unless otherwise specified, the present disclosure relates to and explicitly includes each and every specific member and combination of sub-ranges or sub-groups.
[0145] As used herein, the term “degree of”, when used in relation to a quantity or parameter, refers to a range spanning from approximately one-tenth to ten times the mentioned quantity or parameter.
[0146] Ultrasonic imaging relies on transmitting a narrow acoustic pulse in a specific direction in an optional medium and then receiving echoes from the medium. As used herein, the term "scan line" refers to a line indicating a spatial direction in the medium—from which imaging energy will be received. A 2D image is acquired by receiving echoes from a plurality of scan lines within the medium. The inventors have devised, developed, and tested various methods and systems for effectively reducing broadband noise from an ultrasonic acquisition and / or processing system.
[0147] Referring now to FIG. 1a, an exemplary ultrasonic imaging system is illustrated in which one or more ultrasonic transducers (10) are controlled to perform ultrasonic imaging over a plurality of scan lines (12). The transducers (10) interface with control and processing hardware (100), and the control and processing hardware (100) arbitrarily controls a transmitter (15) for the generation and emission of imaging energy by the transducers (10). The control and processing hardware (100) is configured to receive ultrasonic energy signals detected by the transducers (10), which are typically routed to one or more amplifiers (20) via a Tx / Rx (transmit / receive) switch (25).
[0148] The ultrasonic transducer(s) (10) may be arbitrarily configured to image a spatial region associated with a plurality of scan lines (12), for example, through mechanical scanning of the transducer (10), or, for example, as a non-limiting example, through electronic scanning using an array of imaging elements such as a phased array, a ring array, a linear array, a matrix array, or a curvilinear array. In the latter case, a transmit beamformer (26) and a receive beamformer (27) may be used to generate a plurality of transmit signals and to beamform a plurality of received signals.
[0149] As used herein, the term “receive circuit” generally refers to components such as transmission lines (e.g., coaxial, PCB tracings, others), connectors, mux / demux, RX / TX switches (25), amplifiers (20), slip rings, transformers, and other components known in the art.
[0150] As used herein, the term "transducer receive circuit" may include a receiving circuit connected to one or more ultrasonic transducer elements (10) configured to receive ultrasonic signals when in use.
[0151] As used herein, the phrase “ultrasound receive signal chain” includes a receiving circuit, but may include additional components such as analog-to-digital converters (ADCs) and a noise removal processing module (150), but is not limited to, additional digital processing components and / or processing logic before the signal enters a process of being converted into an image (such as through scan conversion) and subsequent image processing.
[0152] As used herein, the term “channel” may refer to conductive electrical circuits, wireless channels, optical channels, or other signal paths. For example, an imaging receiving channel, which denotes the path traversed by detected imaging signals, is illustrated in FIGS. 1a through 13. The system may utilize a single receiving channel per transducer or several receiving channels per transducer (as may be the case for an array transducer where there may be a channel for each piezoelectric transducer element or group of piezoelectric transducer elements in the array). ASICs and other devices may be used along the signal receiving chain to multiplex signals along channels from more than one piezoelectric transducer element.
[0153] In "imaging mode," the system may be configured to control the transducer (10) to transmit energy to a medium at will and to detect imaging energy within an imaging frequency band (hereinafter referred to as the "imaging band"). The imaging band may consist of a single continuous frequency band in which imaging energy is detected, or two or more frequency intervals (such as in harmonic imaging). Imaging energy or noise within the imaging band is hereinafter referred to as "in-band."
[0154] Figure 14 shows the grouping and terminology used to describe various waveforms.
[0155] This system may also be configured to detect energy in one or more additional frequency bands through one or more channels connected to the transducer, wherein at least one frequency band is, at least partially, out of the imaging band. These one or more additional frequency bands are hereinafter referred to as "detection bands." A waveform that is, at least partially, out of the imaging band is hereinafter referred to as "out-of-band." In some cases, the detection band may be within the imaging band. A waveform that is entirely within the imaging band, where frequency components are within the entire imaging band or in subbands within the imaging band, is hereinafter referred to as "within-band." Noise detection bands may be out-of-band or within-band. At least one detection band may be selected such that when the transducer is used in imaging mode (i.e., when the transducer detects imaging energy), the signal-to-noise ratio within the detection band is substantially smaller than the signal-to-noise ratio in the imaging band. For example, the detection band may be outside the full-width (half-maximum) bandwidth of the imaging band or another bandwidth corresponding to a threshold value below the maximum intensity of the signal used.
[0156] As used herein, "imaging waveform" refers to a waveform (analog or digitally sampled) obtained from an imaging transducer receiving circuit when the imaging transducer is receiving imaging energy or is expected to receive imaging energy.
[0157] As used herein, the phrase “in-band imaging waveform” refers to an imaging waveform (sampled analogously or digitally) that is within an imaging band. An in-band imaging waveform is expected to contain imaging energy and may also contain unwanted noise energy. In various exemplary embodiments of this disclosure, the in-band imaging waveform is processed to remove noise energy for generating a denoised image.
[0158] As used herein, the phrase “detection-band imaging waveform” refers to a waveform acquired from an imaging transducer receiving channel and existing within one or more noise detection bands. The detection-band imaging waveform may be out-of-band or within-band. For example, an within-band noise-detection imaging waveform may be used to confirm the presence of noise within the imaging band. More specifically, an within-band noise-detection imaging waveform may be used to confirm that a noise source having a noise component outside the imaging band also has a noise component within the imaging band. At least partially, a detection-band imaging waveform outside the imaging band is referred to as an “out-of-band noise-detection waveform.”
[0159] Referring to FIG. 1a, the system can be configured to be in a "noise-characterization mode," during which the transducer (10) does not transmit energy to the medium and does not detect imaging energy from the medium.
[0160] As used herein, the phrase "noise-characterization waveform" refers to a waveform acquired when the imaging transducer is not receiving imaging energy.
[0161] As used herein, the phrase “in-band noise-characterization waveform” refers to a waveform present in the imaging band that is acquired from the imaging transducer receiving channel when the imaging transducer is not receiving imaging energy.
[0162] As used herein, the phrase “detection-band noise-characterization waveform” refers to a waveform existing in the noise detection band, obtained from the imaging transducer receiving channel when the imaging transducer is not receiving imaging energy. A detection-band noise-characterization waveform that is at least partially outside the imaging band is referred to as an “out-of-band noise characterization waveform.” A detection-band noise-characterization waveform that is entirely within the imaging band is referred to as a “within-band noise characterization waveform.”
[0163] As used herein, the phrase "baseline noise-characterization waveform" refers to a waveform acquired when the imaging transducer is not receiving imaging energy and when the selected noise source is expected to be off (i.e., absence of noise generation), such that the baseline noise-characterization waveform provides a baseline for a selected noise source. A baseline noise-characterization waveform within the imaging band is referred to as an "in-band baseline noise-characterization waveform." A baseline noise-characterization waveform within the noise detection band is referred to as a "detection-band baseline noise-characterization waveform."
[0164] Referring again to FIG. 1a, an optional reference receiving circuit (11) may be provided, comprising a receiving circuit configured not to receive reflected ultrasonic signals during imaging, while being capable of detecting noise energy similar to noise coupled to one or more transducer receiving circuits during imaging. The reference receiving circuit may utilize one or more components of the transducer receiving circuit (for example, the reference receiving circuit and the transducer receiving circuit may utilize different channels of an amplifier or an ADC).
[0165] In one exemplary embodiment, the system may be configured to detect noise within the imaging band through one or more reference receiving channels that are acoustically isolated or deactivated so as not to transduce reflected ultrasound waves but are optionally connected to a noise-sensitive reference ultrasound transducer (not shown) received by an imaging transducer receiving circuit. One or more imaging transducers (10) and one or more reference transducers do not need to be oriented in a common spatial direction.
[0166] Signals received (via a reference receiving channel) by one or more reference transducer receiving circuits or reference electrical receiving circuits are hereinafter referred to as reference waveforms. Reference waveforms are primarily noise and not imaging energy.
[0167] As used herein, the phrase "reference waveform" refers to a waveform obtained from one or more reference receiving channels. The reference waveform may be filtered to exist within and / or outside the imaging band.
[0168] As used herein, the phrase “reference noise-detection waveform” refers to a reference waveform obtained from a reference receiving channel when an imaging transducer is receiving or is expected to receive imaging energy.
[0169] As used herein, the phrase "reference noise-characterization waveform" refers to a reference waveform obtained from a reference receiving channel when the imaging transducer is not receiving imaging energy.
[0170] This system may be arbitrarily configured to suppress noise using a combination of detection band waveforms and reference waveforms. As used herein, the phrase "noise-detection waveform" refers to a reference waveform or a detection band waveform. When this system is in imaging mode and the imaging transducer receiving circuit is receiving or is expected to receive imaging energy, the noise detection waveform is referred to as a "noise-detection imaging waveform." When this system is in noise characterization mode and the imaging transducer receiving circuit is not receiving imaging energy, the noise detection waveform is referred to as a "noise-detection characterization waveform."
[0171] Although FIG. 1a illustrates a single transducer element, it will be understood that the embodiment illustrated in FIG. 1a provides only one non-limiting exemplary configuration, and transducers having multiple piezoelectric elements may be used. For example, in one exemplary embodiment, multiple transducer elements may be controlled as a phase array, a linear array, or a 2D array. Furthermore, the transducer may not be limited to transmitting imaging energy for the purpose of generating multidimensional 2D cross-sectional images or 3D volumes (including 4D imaging data sets containing 3D images over time), and may include transducers used for Doppler evaluation of flow, transducers used as ultrasonic beacons (e.g., as described in U.S. Patent Publication No. 2016 / 0045184, whose title of invention is “Active localization and visualization of minimally invasive devices using ultrasound,” to which the whole is incorporated by reference), or ultrasonic transducers used to detect the position of moving elements (e.g., as described in U.S. Patent Publication No. 2012 / 0197113, whose title of invention is “Ultrasonic probe with ultrasonic transducers addressable on common electrical channel,” to which the whole is incorporated by reference).
[0172] Transducers may not be limited to transmitting and receiving imaging energy as illustrated in FIG. 1a, and may include transducers that receive ultrasonic energy from a medium excited by other means, such as optical energy (photoacoustic imaging), or by a separate ultrasonic transducer. Furthermore, although FIG. 1a illustrates a configuration for imaging a spatial region associated with multiple scan lines in different directions, the scan lines may be unidirectional, such as in M-mode imaging or during certain Doppler modalities, such as pulsed or continuous wave Doppler for evaluating flow.
[0173] In one exemplary embodiment, a single transducer receiving channel is configured to receive imaging energy within the imaging band and, coincidentally, also receive additional energy within one or more noise detection bands—at least one of which contains out-of-band noise. In another example, one or more imaging transducer receiving channels may be used to receive imaging energy within the imaging band, and one or more transducer receiving channels may be used to receive additional energy within one or more noise detection bands—at least one of which contains out-of-band noise. In yet another example, one or more reference receiving channels may be used to receive noise energy (i.e., reference noise detection waveforms) while being isolated from imaging energy in the imaging band. The reference receiving channels may be filtered in a manner similar to the imaging transducer receiving channels using imaging bandpass filters. Alternatively, in some embodiments, the reference receiving channels may not be filtered at all, or may have filters different from the imaging bandpass filters to better facilitate the estimation of noise within the imaging band.
[0174] The control and processing hardware (100) may include, for example, one or more processors (110), memory (115), system bus (105), one or more input / output devices (120), and a plurality of optional additional devices such as a communication interface (135), a data acquisition interface (140), a display (125), and external storage (130).
[0175] It should be understood that the exemplary system illustrated in FIG. 1a illustrates non-limiting exemplary embodiments and is not intended to be limited to the components illustrated. For example, the system may include one or more additional processor and memory devices. Furthermore, one or more components of the control and processing hardware (100) may be provided as external components that interface with the processing device. For example, as illustrated in the drawing, an arbitrary transmitting beamformer (26) and an arbitrary receiving beamformer (27) may be included as components of the control and processing hardware (100) (as illustrated within the dashed line) or provided as one or more external devices.
[0176] Some aspects of the present disclosure may be embodied, at least partially, in software that, when executed on a computing system, constitutes a special-purpose computing system capable of performing the signal processing and noise reduction methods disclosed herein, or variations thereof. That is, these techniques may be performed in response to a processor, such as a microprocessor, CPU, or GPU, in a computer system or other data processing system, executing a sequence of instructions contained in memory, such as ROM, volatile RAM, non-volatile memory, cache, magnetic and optical disks, cloud processors, or remote storage devices. Furthermore, instructions may be downloaded to a computing device over a data network, such as in the form of compiled and linked versions. Alternatively, the logic for performing the processes discussed above may be implemented in additional computer and / or machine-readable media, such as individual hardware components as LSIs (large-scale integrated circuits), ASICs (application-specific integrated circuits), or firmware such as EEPROMs (electrically erasable programmable read-only memory) and FPGAs (field-programmable gate arrays).
[0177] A computer-readable medium may be used to store software and data that, when executed by a data processing system, cause the system to perform various methods. Executable software and data may be stored in various locations, for example, including ROM, volatile RAM, non-volatile memory and / or cache. Parts of this software and / or data may be stored in any of these storage devices. Generally, a machine-readable medium includes any mechanism that provides (i.e., stores and / or transmits) information in a form accessible by a machine (e.g., a computer, a network device, a personal digital assistant, a manufacturing tool, any device having a set of one or more processors, etc.).
[0178] Examples of computer-readable media include, but are not limited to, volatile and non-volatile memory devices, read-only memory (ROM), random access memory (RAM), flash memory devices, floppy and other removable disks, magnetic disk storage media, optical storage media (e.g., CDs (compact discs), DVDs (digital versatile disks), etc.), recordable and non-recordable types of media such as network attached storage and cloud storage. Instructions may be embodied in digital and analog communication links for electrical, optical, acoustic, or other forms of propagation signals such as carrier waves, infrared signals, digital signals, and similar ones. As used herein, the phrases "computer-readable material" and "computer-readable storage medium" refer to all computer-readable media, excluding transitory propagating signals themselves.
[0179] Many of the embodiments described herein utilize the adjustment of a noise reduction filter based on noise detected in the environment. In some exemplary implementations, one or more of the waveforms, data, filter parameters, and other related information described in the exemplary embodiments below may be transmitted over a network and evaluated remotely for further analysis and / or optimization of the noise reduction filter implementation. Once optimized, noise reduction filter algorithms and / or parameters may subsequently be transmitted to this system to enable improved noise reduction.
[0180] As illustrated in FIG. 1a, exemplary control and processing hardware (100) includes an imaging processing module (145) and a noise suppression module (150). The image processing engine (145) may be configured or programmed to execute known image processing methods, such as scan conversion.
[0181] Although some of the embodiments described herein have been illustrated and described in a manner that enables real-time noise reduction, it should be understood that noise reduction may occur in a post-processing manner. For example, data on a transducer receiving channel or a reference receiving channel may be digitized and stored before or after any filtering, envelope detection, shifting, shape / phase or delay adjustments, signal characterization, attenuation, subtraction, or other steps in the described embodiments of the present invention.
[0182] FIG. 1b illustrates examples of steps that may be utilized by control and processing hardware (100) and a receiving channel to process a detected in-band imaging waveform from an imaging transducer receiving channel prior to image generation. The detected waveform from the imaging transducer receiving circuit may be amplified (201) and filtered (202) prior to analog-to-digital conversion (203). Once digitized, a band-pass filter (200) (which may utilize multiple passbands and stopbands) is used to filter the detected waveform and retain the signal in the imaging band. The envelope of the filtered waveform is then generated through an envelope detector (210). The resulting envelope of the detected waveform is then optionally decimated or expanded (220) and provided to an image processing module (230) for image generation.
[0183] Referring again to FIG. 1a, exemplary control and processing hardware (100) includes one or more noise suppression modules (150), and the noise suppression modules (150) include instructions for processing detected data (e.g., raw RF data, envelope data, or image data) to reduce the contribution of noise according to noise reduction algorithms described in detail below. As described below, the noise suppression algorithms disclosed herein (and represented as the noise suppression module (150) in FIG. 1a) may be used to remove or reduce noise in several potential steps during the processing flow illustrated in FIG. 1b, based on processing one or more noise detection waveforms. In the case of systems using array transducers, noise suppression may occur before or after beamforming (or both). In various exemplary embodiments described in detail below, noise reduction of imaging data (including, but not limited to, raw waveforms, sampled waveforms, envelope waveforms, Fourier transformed signals, and processed image data) is performed based on measurements of signal energy (power, amplitude, intensity, or other measures of signal strength) or waveform patterns of noise detection waveforms (such as out-of-band noise detection imaging waveforms detected through imaging receiving channels or reference noise detection waveforms detected on reference receiving channels). In many embodiments, the noise detection waveform, which is substantially free of imaging energy, may include noise that is correlated with or coincident with noise within the imaging band (or imaging bands). One or more relationships between imaging band noise and noise detected by the noise detection waveform may be used for correction of imaging band signals for the removal or reduction (e.g., suppression) of imaging band noise.
[0184] Implementations of reference receiving circuits for in-body imaging
[0185] Now, referring to FIG. 1c, an exemplary system for ultrasound imaging of an area using an intracorporeal imaging probe (350) connected to control and processing hardware (100) via a patient interface module (PIM) (300) is illustrated. The intracorporeal ultrasound imaging device may be configured to receive acoustic imaging energy from a one-dimensional, two-dimensional, or three-dimensional area through optional mechanical or electronic scanning.
[0186] An imaging probe (350) comprises an imaging assembly (353) located far from its proximal end, having an electric and / or optical channel (354) passing through an arbitrary conduit (354) along at least a portion of its length, and a connector (351) located at its proximal end. For the purposes of this disclosure, the imaging assembly (353) generally refers to a component of the imaging probe (350) or a set of components in which imaging energy (e.g., acoustic or optical signals) is detected for the purpose of imaging an area adjacent to the imaging assembly. The imaging assembly may optionally include one or more imaging energy emitters and at least one imaging energy receiver. For example, the imaging assembly may include an ultrasonic imaging transducer (10) that is both an emitter and a receiver of acoustic energy. The ultrasound imaging transducer may be mounted on an imaging assembly that is optionally attached to or otherwise connected to a rotatable conduit or shaft (e.g., a torque cable) (352) that is housed within the hollow sheath of an ultrasound imaging probe in vivo to facilitate mechanical scanning.
[0187] An optional PIM (300) facilitates the transmission of signals from any wires or conduits to an appropriate image processing unit (100) via a PIM cable (320), such as when the imaging probe (350) is not directly connected to the control and processing hardware (100). The PIM may include one or more amplifiers (20) to amplify signals from one or more transducer receiving channels. The PIM may optionally include a motor drive unit (301) that imparts rotational motion to a rotatable conduit (354). The motor drive unit (301) may include a slip ring, a rotary transformer, or other component that couples the signals from the probe (350) to the control and processing hardware (100), thereby allowing the imaging conduit to rotate while the PIM cable (320) is not rotating. The PIM (300) may also optionally include a pullback mechanism (302) or a reciprocating push-pull mechanism to facilitate longitudinal translation of the imaging assembly (353). Such longitudinal translation of the imaging assembly may occur with longitudinal translation of the outer shaft surrounding the imaging conduit, or may occur within a relatively stationary outer shaft.
[0188] Many electrical components within the imaging system can pick up unwanted energy from environmental noise sources. Examples of such components include the imaging assembly (353), the imaging conduit (352), the motor drive unit (301), and the PIM cable (320). One or more reference receiving circuits that detect noise correlated with the noise detected by the imaging receiving circuit may be useful for suppressing in-band noise. The following are exemplary implementations of reference receiving circuits for noise reduction of the imaging signal in an ultrasonic imaging system.
[0189] FIG. 1d illustrates an exemplary embodiment in which an imaging transducer is replicated by one or more non-imaging reference transducers (361) present within an imaging probe (350). The reference transducer has its own electrical channel (360) passing through an optional conduit (352) and a connector (351). The reference transducer (361) may be coated with epoxy or any other acoustic damping material (362) so as to be acoustically isolated from receiving imaging energy. Alternatively, the piezoelectric may be depoled to make the piezoelectric inactive or replaced in other ways with an acoustically insensitive substrate. This implementation may be extended to array transducers, which may have multiple ultrasonic transducer elements configured to receive acoustic imaging energy. One or more elements of the array may be acoustically insensitive so as not to transduce acoustic imaging energy and may function as a reference transducer receiving circuit.
[0190] FIG. 1e illustrates an exemplary embodiment in which an imaging probe comprises two or more imaging transducers, each having an individual electrical channel (370 and 354). The two or more imaging transducers may be sensitive to receiving acoustic imaging energy in substantially non-overlapping spectral bandwidths. For example, the first transducer may be configured to receive acoustic energy for frequencies centered at 10 MHz, and the second imaging transducer may be configured to receive imaging energy for frequencies centered at 40 MHz. The 40 MHz band of the 10 MHz transducer may function as a reference noise channel for the 40 MHz transducer, and similarly, the 10 MHz band of the 40 MHz transducer may function as a reference noise channel for the 10 MHz transducer.
[0191] FIG. 1f illustrates an exemplary embodiment in which an imaging transducer receiving channel is replicated by a reference receiving circuit present within an imaging probe. The reference receiving circuit may have any combination of resistors, inductors, capacitors, and / or other components configured such that the electrical impedance of the reference electric circuit (381) matches the impedance of the imaging transducer receiving circuit, or such that the sensitivity of the reference electric circuit to noise is more similar to the sensitivity of the transducer receiving channel to noise. The advantage of such embodiments is that they may be cheaper, easier to manufacture, and easier to miniaturize than some of the components of the reference electric circuit (381) by not requiring an actual ultrasonic transducer. Furthermore, parts of the reference receiving circuit may serve additional purposes, such as transmitting energy to drive an actuator (such as, but not limited to, a magnetic actuator) or transmitting a signal (including, but not limited to, a current generated from an electromagnetic field for sensing temperature, pressure, or position). Such use of a reference receiving circuit for additional purposes other than merely collecting a reference noise detection waveform to reduce noise in an imaging signal can enable easier miniaturization, lower cost, and / or improved functionality.
[0192] FIG. 1g illustrates an exemplary embodiment in which a reference receiving circuit (391) is terminated within a PIM that is affected by noise received by a motor drive unit and a PIM cable, similar to an imaging receiving channel. Here, only a portion of the transducer receiving circuit is replicated. Reference noise detection may be optionally used in combination with detection band (out-of-band or in-band) noise detection to further reduce noise in the in-band imaging waveform from the imaging transducer receiving channel.
[0193] It should be noted that embodiments in which a detection band waveform is used as a noise detection waveform may be less expensive to generate than those utilizing a reference circuit or reference transducer to generate a reference waveform, because the former does not require a physical implementation of a reference channel, such as in cases where the imaging probe or part thereof is not used repeatedly across different patients. It should also be noted that while a noise detection waveform from a reference channel may be more effective in reducing noise in some imaging systems because it can provide information regarding noise present within the imaging band, an out-of-band noise detection waveform does not provide a direct estimate of in-band noise; instead, it relies on noise whose in-band properties can be predicted based on its out-of-band properties, at least partially.
[0194] It is also noted that noise estimates obtained using either a detection band waveform or a reference noise detection waveform can be used to reduce noise in more than one imaging channel. For example, in a phase array transducer with multiple piezoelectric elements, it is possible to use a single reference receiving channel or a single out-of-band noise detection waveform to estimate the noise that may be collected by all or a subset of the piezoelectric elements, and thus to apply the same noise estimation scheme to the signals collected from all or a subset of the piezoelectric elements.
[0195] Noise measurements
[0196] The following sections of the present disclosure describe several different exemplary embodiments for performing noise reduction of in-band image data based on out-of-band noise detection, or reference channel noise detection, or a combination of both out-of-band noise detection (possibly further supplemented by in-band noise detection) and reference channel noise detection.
[0197] As described in connection with exemplary embodiments of the following examples, various noise measurements and / or noise characteristics may be determined from measurements in the noise detection waveform to increase the signal-to-noise ratio within the imaging band. Non-limiting examples of noise measurements include one or more of the following:
[0198] Measurements of energy (amplitude, root-mean-square amplitude, average power) in the noise detection imaging waveform and the in-band imaging waveform;
[0199] Energy measurements in two or more different noise detection bands;
[0200] Temporal, spectral, and / or time-frequency properties of the noise detection waveform; or spatial or spatiotemporal patterns in images generated using the noise detection waveform (including features or parameters describing or characterizing such patterns);
[0201] Temporal, spectral, and / or time-frequency properties of the noise detection imaging waveform and co-incidental patterns in the in-band imaging waveform (including features describing such patterns); spatial or spatiotemporal patterns in images generated using the noise detection waveform and co-incidental spatial patterns in images generated from the in-band imaging waveform from the imaging transducer receiving channel (including features or parameters describing or characterizing such patterns);
[0202] Filter parameters determined or controlled using parameters obtained based on waveform characteristics, such as the spectral spacing of energy and harmonic peaks detected in a noise detection waveform.
[0203] In some exemplary embodiments, the estimation of noise characteristics may be performed when the imaging transducer is not receiving imaging energy (e.g., after the ultrasonic energy from the most recent emission of an ultrasonic pulse is expected to dissipate from the environment). Alternatively, the estimation of noise characteristics may be performed during imaging when imaging energy is expected to be detected (e.g., when the transducer is in image acquisition mode). In some exemplary embodiments where noise characteristics are measured in the absence of imaging energy, such noise characteristics may be updated intermittently to adapt to and compensate for time-dependent variations in the noise characteristics.
[0204] Noise reduction using measurements in the absence of imaging energy
[0205] Although some noise reduction embodiments involve the measurement and use of energy from a noise detection waveform while the imaging transducer is receiving imaging energy, alternative embodiments may utilize measurements from a noise detection waveform acquired during periods of absence of imaging energy, or a combination of both.
[0206] FIG. 5a illustrates an example of such an embodiment in which noise characterization is performed while the transducer receiving channel is not receiving imaging energy, and the resulting noise characterization can be used to enable noise reduction of in-band imaging waveforms acquired while the transducer receiving channel is receiving imaging energy. Typically, the noise characterization step occurs prior to the acquisition and processing of noise-reduced imaging data, but with proper recording of the imaging data, information collected during noise characterization can be used post-hoc for the recorded imaging data.
[0207] According to the exemplary method, energy is detected in a transducer receiving circuit during a first time window when at least one transducer receiving circuit is expected or known not to be receiving imaging energy, so that the waveform detected by the imaging transducer receiving channel (13) is thought to be noise (405). The detected waveform is filtered at 200 and 410 to generate an in-band noise characterization waveform (407) and a detection band noise characterization waveform (408).
[0208] The in-band noise characterization waveform and the detection band noise characterization waveform are processed to characterize their noise properties, as illustrated in 420. The noise characterization (420) can be used, for example, to generate characteristic parameters (430) that characterize the noise. Examples of suitable noise characterization parameters are provided in the following exemplary embodiments.
[0209] Optionally, energy may be detected in the transducer receiving circuit during an additional baseline noise characterization stage when at least one transducer receiving circuit is expected or known not to be receiving imaging energy and a specific noise source is known to be optionally off, so that the waveform detected by the imaging transducer receiving channel (13) is considered to be baseline noise for a selected noise source. The detected baseline noise characterization waveform is filtered at 200 and 410 to generate an in-band baseline noise characterization waveform (407) and a detection band baseline noise characterization waveform. It should be understood that the noise parameters (430) may include parameters acquired during the baseline noise characterization stage.
[0210] Characteristic noise parameters can be calculated before or during the imaging session, or retrieved from a pre-stored database located on a local or remote storage drive (network drive, cloud, etc.).
[0211] After characterizing noise in the absence of imaging energy, characteristic parameters (430) can be used to perform noise reduction of the in-band imaging waveform (437) detected while the transducers are receiving imaging energy. The waveform detected from the imaging transducer receiving channel during imaging (435), which includes imaging energy and noise, is filtered at 200 and 410 to generate the in-band imaging waveform (437) and the detected band imaging waveform (438). Thus, the in-band imaging waveform (437) includes the detected imaging energy and noise, and the detected band imaging waveform (438) includes information associated with the possibility of the presence of noise in the in-band imaging waveform (437). The characteristic parameters (430) obtained during the noise characterization stage can then be used for the detection and / or estimation of noise (440) within the in-band imaging waveform and for noise suppression (500) of the in-band imaging waveform. Examples of suitable noise characterization parameters and how they are used to reduce noise are provided in the following exemplary embodiments.
[0212] FIG. 5b illustrates an alternative embodiment in which a reference receiving channel, such as the reference receiving channel described with reference to FIG. 1a, is used to detect noise and generate a reference noise characterization waveform (406). This reference noise characterization waveform is filtered in 202 to generate a filtered reference noise characterization waveform (409). In one exemplary embodiment, the reference channel filter may be an imaging bandpass filter. Alternatively, if the noise estimation gains from the input of out-of-band noise, the reference channel filter may differ from an imaging bandpass filter. The reference noise characterization waveform (409), and optionally the in-band noise characterization waveform (407), are processed in step (420) to provide noise characterization parameters (430).
[0213] During imaging, a reference waveform (436) is detected and randomly filtered to generate a filtered reference noise detection waveform (439). The characteristic parameters (430) obtained during the noise characterization stage may then be used to detect and / or estimate noise (440) within the in-band imaging waveform (437) and to perform noise suppression (500) of the in-band imaging waveform (437). In another exemplary embodiment, both the detected band imaging waveform (438) (shown in FIG. 5a) and the filtered reference noise detection waveform (439) are processed to provide information regarding the possibility of the presence of noise in the in-band imaging waveform (437).
[0214] In some exemplary embodiments, noise suppression may be achieved by processing an in-band imaging waveform using one or more of the following methods: subtracting noise estimated from a signal in the imaging band; attenuating the estimated noise energy by multiplying the signal in the imaging band by an attenuation factor; and filtering the signal in the imaging band. For example, the subtraction value may be proportional to the amount of power detected in the out-of-band noise detection imaging waveform. In another example, the attenuation factor may be inversely proportional to a scale associated with the amount of noise in the in-band imaging waveform, thereby attenuating the portions of the in-band imaging waveform associated with the noise.
[0215] Noise characterization (as illustrated in 420 in FIGS. 5a and 5b, for example) can be performed once, or alternatively, multiple times or consecutively. For example, to adapt to and compensate for time-dependent changes in noise characteristics, noise characterization can be performed intermittently over time (e.g., at periodic or non-periodic intervals).
[0216] When noise characterization is occurring, the noise characterization waveforms collected for noise characterization can be digitized and collected in a number of individual arrays, such as arrays that are each long enough to store imaging data along a single scan line in ultrasound imaging, or can be collected in a more continuous manner as one or more data streams stored in a large array, circular buffer, or other data structure.
[0217] In some exemplary implementations, noise characterization may be user-initiated, for example, at the start of an imaging session or when the user observes or suspects a degradation in image quality (e.g., by pressing a button).
[0218] In other exemplary implementations, noise characterization may be triggered automatically or after prompting the user when the absence of imaging energy is detected. For example, periods of the absence of imaging energy may be detected when the relative energy between the in-band imaging waveform and the noise detection imaging waveform remains unchanged for a predetermined time period and is within a predetermined range. For example, periods of the absence of imaging energy may also be determined when the energy in the in-band imaging waveform after noise correction in 501 is below a predetermined threshold—which indicates the absence of imaging energy.
[0219] The noise characterization step (420) may also be useful for warning the user or the system that the noise profile has changed in a way that could cause the noise reduction algorithm to have an adverse effect on the system (if a new isolated noise source is detected, the noise suppression module (500) may mistakenly cause suppression of the in-band signal or be less effective at suppressing in-band noise). For example, a noise-detection waveform (e.g., the out-of-band noise detection waveform (438) of FIG. 5a or the reference noise detection waveform (439) of FIG. 5b) may be processed to determine that the noise characteristics have changed. For example, while performing noise reduction, an arbitrary noise monitoring module may be used to monitor the characteristics of the noise detection waveform (such as peak energy, power, frequency content slope, skew, kurtosis, histogram, or any other metrics). If the characteristics of the noise detection waveform change (e.g., if the peak energy exceeds a threshold value), the noise monitoring module can communicate with other parts of the system (such as through messages, interrupts, alarms, or others) to warn that the noise content has changed.
[0220] In another implementation, an error value may be evaluated at the noise characterization stage, where noise suppression is performed on the in-band imaging waveform in the absence of imaging energy, and the error value is the energy of the in-band imaging waveform after noise correction. If the error value exceeds a predefined threshold, a warning is generated. The warning may prompt for re-characterization of the noise, or the system may choose to ignore one or more out-of-band noise detection bands or one or more reference receiving channels in its noise removal algorithm.
[0221] Noise characterization (420) may also be useful for determining noise sources in the environment. Noise sources may be determined by a pattern recognizer, for example, as described in step (570) of FIG. 7a (described in more detail in Example 4). For example, information about noise sources may be used to access a database (local or networked) to select parameters for noise suppression or to determine a sequence of noise reduction methods to be used. For example, it may be desirable to remove periodic noise first (as described in detail in Example 6, FIGs. 8a through 8d), and then remove less periodic noise.
[0222] As an additional example, noise characterization may be useful for detecting the type of electroanatomical mapping system used during an ablation procedure, or for detecting the activation and deactivation of an ablation catheter, such as a catheter that uses radio frequency energy to perform ablation to treat arrhythmias. This can be achieved, for example, by a pattern recognizer such as that described in step (570) of FIG. 7a. One or more noise detection waveforms may additionally detect the duration, relative intensity, or frequency of the applied ablation energy. Such information may be useful for an intracardiac imaging system because it can facilitate annotating imaging data sets with information regarding when a noise source, such as an ablation catheter, was activated during the procedure.
[0223] Noise suppression (as illustrated in 500 in FIGS. 5a and 5b, for example) can be performed once, multiple times, intermittently, or continuously. For example, noise suppression can be initiated by the user. Alternatively, to compensate for time-dependent noise sources, noise suppression can be performed intermittently over time (e.g., at periodic or non-periodic intervals). In another example, noise suppression can be controlled by an external device emitting noise. For example, noise suppression can be enabled or disabled by controls of an RF ablation generator so that noise suppression is performed when RF energy is being delivered.
[0224] The following embodiments are described for one imaging waveform and one noise detection waveform. It should be understood that these embodiments may be extended to multiple imaging waveforms and / or multiple noise detection waveforms.
[0225] Example 1: Noise reduction based on suppressing envelope-detected out-of-band noise using arbitrary amplitude, shape, and delay corrections
[0226] Now, referring to FIG. 2a, an exemplary method is illustrated in which an out-of-band noise detection waveform (438) is used to perform noise reduction of an in-band imaging waveform (437) through a suppression operator (525). In one exemplary embodiment, the suppression operator may be a subtractor that subtracts the noise estimated from the in-band imaging waveform (437). In another exemplary embodiment, the suppression operator may be an attenuator that attenuates the in-band imaging waveform with an attenuation factor derived from the estimated noise. One or more transducer receiving channels are used to detect imaging energy, wherein the detected energy includes both the imaging band and the noise detection band. The waveforms may be digitally sampled, divided (or copied), and filtered so as to obtain the in-band imaging waveform (437) sampled thereby and the out-of-band noise detection imaging waveform (438) sampled thereby. Sampled waveforms can be detected as a set of samples received in a time window (listening window). For example, in the case of ultrasonic imaging, the listening window may occur immediately after or after pulsing the ultrasonic transducer so that the ultrasonic transducer releases energy into the adjacent environment. Pulsing may correspond to transmitting one or more pulses.
[0227] In the exemplary embodiment illustrated in FIG. 2a, the imaging waveform (435) is filtered (digitally or analogously) using an imaging bandpass filter (200) and a noise detection band filter (410) spanning frequencies outside the imaging band. Envelope detection may subsequently be performed on the in-band imaging waveform (437) and the out-band noise detection imaging waveform (438), respectively, as illustrated in 210 and 411.
[0228] In the exemplary embodiment illustrated in FIG. 2a, the out-of-band noise detection imaging waveform (438) is used to reduce the noise of the in-band imaging waveform (437). Prior to suppression, the amplitude of the envelope-detected out-of-band noise detection imaging waveform (438) is arbitrarily scaled through an amplitude adjustment factor, as illustrated in 510, to compensate for the differences between the noise power within the imaging band and the noise power within the noise detection band. In one exemplary embodiment, the amplitude adjustment factor may be determined based on the power spectrum of the noise, as determined in the absence of imaging energy, i.e., at the noise characterization stage. In another exemplary embodiment, the amplitude adjustment factor may be selected or modified by an operator to provide a desired level of noise reduction, or may be determined after cross-correlation in 580 (described below).
[0229] Before suppression, in order to compensate for differences in the shape of noise waveforms between the imaging band and the noise detection band, the envelope-detected out-of-band imaging waveform can be temporally expanded, compressed, or shaped using any other linear or non-linear time scaling function, as illustrated in 510.
[0230] As illustrated in 510, it may also be advantageous to apply delay correction to the envelope-detected out-of-band noise detection imaging waveform prior to suppression. For example, two bandpass filters (200 and 410) may not transform the input waveforms similarly. Either the properties of the bandpass filters or the noise itself may result in an offset of the noise when propagated through the bandpass filters. In the absence of delay correction, the noise may be incorrectly shifted to suppression, which may have a negative effect on the noise-reduced signal.
[0231] In one exemplary implementation, delay adjustment can be achieved by calculating the cross-correlation between the in-band imaging waveform and the out-band imaging waveform, and aligning the waveforms at the point where the cross-correlation is maximum. In other words, the cross-correlation can be used to determine a time delay correction value to correct the relative temporal misalignment of the envelopes of the in-band imaging waveform and the out-band imaging waveform.
[0232] Time delay correction values and amplitude correction values may be calculated using a plurality of sampled in-band imaging waveforms and (simultaneously occurring) sampled out-band noise detection imaging waveforms (which may be referred to as "arrays"), or in one or more time windows in the sampled in-band imaging waveforms and (simultaneously occurring) sampled out-band noise detection imaging waveforms.
[0233] In one exemplary implementation, out-of-band noise may be interrogated as a plurality of noise detection bands, and the power dependence with respect to frequency between the plurality of noise detection bands may be used to select an appropriate amplitude adjustment for estimating the noise power present in the imaging band for suppression in step (525). For example, the average noise power within the plurality of noise detection bands may be fitted according to a functional dependence with respect to frequency, such as a linear fit, to estimate the noise power within the imaging band. This functional dependence with respect to frequency may be determined in the absence of imaging energy, i.e., in the noise characterization stage (420).
[0234] Alternative embodiments: Noise reduction based on using frequency-shifted out-of-band noise detection waveforms with arbitrary amplitude, shape, and delay corrections
[0235] FIG. 2b illustrates an exemplary embodiment of an alternative to a subtractive or signal attenuation noise correction method in which an out-of-band noise detection imaging waveform is frequency-shifted prior to delay and amplitude adjustment. As shown in FIG. 530, a frequency-shift operation (e.g., by multiplying by a complex exponent) is performed on the out-of-band noise detection imaging waveform to shift the spectrum of the out-of-band noise detection imaging waveform so as to be within the imaging band or to overlap with the imaging band.
[0236] The frequency shift operation (530) can be performed such that the center frequency of the frequency-shifted noise detection waveform matches or is approximately the same as the center frequency of the imaging band. For example, if the imaging band is in the range of 7 to 13 MHz, the center frequency is f c1 = 10 MHz. If the noise detection band is in the range of 15 to 25 MHz, the center frequency of the noise detection band is f c2 = 20 MHz. Accordingly, the out-of-band noise detection imaging waveform is f c2 -f c1 A frequency shift operation can be performed to shift by -10 MHz. Alternatively, the frequency shift operation can be performed so that the center frequency of the frequency-shifted out-of-band noise detection imaging waveform matches or is nearly identical to a frequency within the imaging band where a portion of the in-band noise is predicted or known to exist.
[0237] After frequency shifting, the summed frequency artifact (in the example described above, the summed frequency is f c2Another bandpass filtering stage (203) is performed to filter out (+10 MHz = 30 MHz). Frequency shifting may be advantageous compared to the envelope detection embodiment illustrated in FIG. 2a because frequency shifting can result in better correlation between noise in the imaging band and noise detection band, which can result in better noise suppression. Note that in the exemplary embodiment illustrated in FIG. 2b, envelope detection (210) may be applied to the output signal (501) after the output signal (501) has undergone noise reduction to obtain a noise-reduced signal envelope (520).
[0238] Alternative embodiments based on the use of a reference noise signal
[0239] Now, referring to FIG. 2c, an exemplary method is illustrated in which in-band noise detected through a reference receiving channel (using a reference receiving circuit) is used to reduce noise in an in-band imaging waveform through a suppression operator (i.e., a subtractor or attenuator). One or more imaging transducer receiving channels are used to receive imaging energy, and one or more reference receiving channels are used to receive noise energy (i.e., reference waveforms) that is expected to correlate with the noise energy received by the imaging transducer receiving channels.
[0240] These waveforms can be digitally sampled to obtain in-band imaging waveforms sampled by this, and sampled filtered reference noise detection waveforms. Alternatively, noise suppression can be performed with analog electronic devices, such as by using an analog signal adder, where the input from the reference receiving circuit to the adder is inverted in the delay, scale, and shape adjustment block (510), thereby resulting in the subtraction of the estimated noise. As a further alternative embodiment for analog signal suppression, the suppression can be embodied as an amplifier having a time-varying gain, wherein the gain is modulated by the noise detected in the reference receiving channel.
[0241] In the exemplary embodiment illustrated in FIG. 2c, input waveforms from the imaging transducer receiving channel and the reference receiving channel are filtered (digitally or analogously) using an imaging bandpass filter (200) and an arbitrary reference channel filter (202) to respectively provide an in-band imaging waveform and a filtered reference noise detection waveform. As previously mentioned, the reference channel filter (202) may be similar to the imaging bandpass filter (200). As illustrated in 210 and 411, envelope detection is subsequently performed arbitrarily on the filtered signals. In the exemplary embodiment illustrated in FIG. 2c, the reference noise detection waveform measured by the reference receiving channel is used to reduce noise in the in-band imaging waveform. Before subtraction from the in-band imaging waveform (or its envelope) or attenuation of the in-band imaging waveform (or its envelope), the amplitude of the filtered reference noise detection waveform (or its envelope) is arbitrarily scaled through an amplitude adjustment factor, as shown in 510, to compensate for the differences in noise power within the filtered reference noise detection waveform and the in-band imaging waveform. In one exemplary implementation, the amplitude adjustment factor may be determined based on the power spectrum of the noise, such as that determined in the noise characterization stage, in the absence of imaging energy. In another exemplary implementation, the amplitude adjustment factor may be selected by the operator to provide a desired level of noise reduction, or may be determined after cross-correlation in 580.
[0242] As illustrated in 510, it may also be advantageous to apply delay correction to the envelope-detected filtered reference noise detection waveform prior to subtraction. In one exemplary implementation, delay adjustment can be achieved by calculating the cross-correlation between the in-band imaging waveform and the filtered reference noise detection waveform, and aligning the waveforms at the point where the cross-correlation is maximum. Similar to the previous embodiment, the time delay correction value and the amplitude correction value may be calculated using a plurality of in-band imaging waveforms and (simultaneous) reference noise detection waveforms, or using one or more time windows of the in-band imaging waveform and (simultaneous) reference noise detection waveform.
[0243] Example 2: Noise reduction using a reference noise detection waveform from a reference receiving channel as an input to an adaptive filter
[0244] FIG. 3a illustrates an exemplary embodiment of a noise correction method in which an adaptive filter is used by applying noise reduction correction to an in-band imaging waveform based on a reference noise detection waveform in an active noise control (ANC) scheme, wherein the reference noise detection waveform corresponds to noise detected by an imaging transducer receiving circuit. The reference noise detection waveform is filtered at 202, and the waveform from the imaging transducer receiving channel is filtered at 200. In a preferred embodiment, the in-band imaging waveform and the reference noise detection waveform are filtered within the same band (e.g., 7 to 13 MHz in the case of an exemplary intracardiac echocardiography imaging system).
[0245] An adaptive filter is a linear filter having means to adjust its parameters according to an optimization algorithm and a transfer function controlled by variable parameters. Adaptive filters are typically digital finite-impulse-response (FIR) or infinite-impulse-response (IIR) filters. An active noise control (ANC) scheme is provided for a primary input receiving a signal (S) from a signal source that is corrupted by the presence of noise (N) unrelated to the signal. The reference input is noise (N) that is unrelated to the signal but is correlated in some way with the primary input noise (N). r ) is received. The reference noise passes through an adaptive filter and is an output noise (N) which is an estimate of the first-order input noise (N). estimate Generates ). The noise estimate is subtracted from the damaged signal to generate the noise-reduced signal estimate (S estimate ) generates. The adaptive filter outputs power E[s estimate 2 His coefficients are actively adjusted to minimize ]. The signal (S) is N and N r While it is not related to, noise (N) is noise (N r Since it is correlated with ), minimizing the total output power maximizes the signal-to-noise ratio. Minimization algorithms, such as the stochastic Least Mean Squares (LMS) algorithm or the deterministic recursive Least Squares (RLS) algorithm, can be used to find filter coefficients that minimize the output noise power.
[0246] In the exemplary ANC scheme illustrated in FIG. 3a, reference noise is measured through a reference receiving channel using a reference receiving circuit isolated from imaging energy. A primary input is obtained by applying an imaging bandpass filter (200) to the input waveform from the imaging transducer receiving channel (13). Reference input (N r ) is obtained by applying a reference channel filter (202) to a reference noise detection waveform obtained through a reference receiving channel.
[0247] Alternative embodiments: Active noise cancellation using out-of-band noise
[0248] Unlike the form of active noise control described above, in which a reference noise detection waveform and an in-band imaging waveform can be detected and processed within overlapping frequency bands (and potentially a common frequency band), FIGS. 3b and 3c illustrate exemplary embodiments of noise correction methods in which an adaptive filter (540) is used to apply noise reduction correction to an in-band imaging waveform using an out-of-band noise detection imaging waveform.
[0249] The primary input is obtained by applying an imaging bandpass filter (200) to the input waveform from the imaging transducer receiving channel and obtaining its envelope (210). In the exemplary embodiments illustrated in FIGS. 3b and 3c, the out-of-band noise detection imaging waveform is obtained by applying a detection band filter (410) to the input waveform from the imaging transducer receiving channel.
[0250] In FIG. 3b, the out-of-band noise detection imaging waveform is demodulated through envelope detection (411) (in a manner similar to the embodiment shown in FIG. 2a) to obtain a reference input for ANC.
[0251] In FIG. 3c, the out-of-band imaging waveform is frequency-shifted (in 530) to an imaging band (e.g., 7 to 13 MHz) in a manner similar to the embodiment shown in FIG. 2b and filtered using an imaging bandpass filter (203) to obtain a reference input for ANC.
[0252] Example 3: Frequency shift-based noise reduction using a detection band waveform as input to a variable filter of an in-band waveform
[0253] FIG. 4 illustrates an exemplary embodiment of a noise correction method in which a dynamic filter (550) is used to filter an in-band imaging waveform, wherein the dynamic filter is controlled by a filter update algorithm (560) that updates filter coefficients after processing an out-of-band noise detection imaging waveform containing out-of-band noise and optionally an in-band noise detection imaging waveform containing noise within all or part of the imaging band. As in previous embodiments (Figs. 2b and 3b), as illustrated in FIG. 4, the input waveform is individually filtered by an imaging bandpass filter (200) and a noise detection bandpass filter (410) to thereby generate at least one out-of-band noise detection imaging waveform containing an in-band imaging waveform (437) and out-of-band noise (438). One or more out-of-band noise detection imaging waveforms are processed by the filter update algorithm in 560.
[0254] The filter update algorithm can evaluate signal characteristics, such as by analyzing out-of-band noise, performing a Fourier transform on the waveform array, and identifying spectral maxima and their frequency of occurrence. The filter update algorithm can use this information to control the coefficients of a dynamic digital filter that filters the imaging waveform at 550.
[0255] In one exemplary implementation, the present method may be used to reduce noise in a signal containing harmonic noise. For example, in the frequency domain, harmonic noise may be generated from a switching rectangular pulse source in which the spacing of spectral lines depends on the pulse repetition frequency. If the pulse width modulation source generates noise spanning a band of 3 to 40 MHz, where the imaging band is within 7 to 13 MHz and the dynamic filter (550) is a comb or multi-notch filter, the filter update algorithm (560) may process signals from the noise detection band (e.g., evaluate spectral line spacing and positions as seen in the 15 to 25 MHz range) and use this information to control stop bands in the dynamic in-band filter (550) to remove or reduce harmonic noise.
[0256] In addition to out-of-band noise being used to update the dynamic filter, the filter update algorithm (560) and / or the dynamic filter (550) may also optionally probe the in-band noise detection imaging waveform to identify the presence of in-band noise in one or more selected subbands within the imaging band before removing or reducing the noise. For example, if the imaging band is within 7 to 13 MHz and the filter update algorithm recognizes that there is harmonic noise at 15 MHz, 18 MHz, 21 MHz, 24 MHz, and 24 MHz (integer multiples of the range from 15 to 25 MHz), the filter update algorithm may set the dynamic filter to filter out signals at 9 MHz and 12 MHz (integer multiples of 3 MHz within the imaging band). In one exemplary implementation, such a filter may be optionally applied only when the presence of signals at 9 and 12 MHz is greater than expected for other signals within the imaging band. In another exemplary implementation, a noise characterization step may be performed in the absence of imaging energy to determine whether harmonic noise is present within the imaging band.
[0257] Example 4: Noise reduction based on pattern recognition
[0258] FIGS. 7a and 7b illustrate exemplary embodiments of a noise correction method in which pattern recognition is used to detect noise and to perform noise reduction on in-band imaging waveforms. In these exemplary embodiments, a set of matched patterns in detection band noise characterization waveforms (at least one of which is, at least partially, outside the imaging band) and the associated in-band noise characterization waveforms are initially identified during a noise characterization period when imaging energy is not received, as shown in FIG. 7a. After the noise detection patterns are correlated with in-band noise patterns through a noise characterization stage, these correlations can be used during imaging to perform noise reduction on the in-band imaging waveforms based on the identification of patterns in one or more detection band imaging waveforms.
[0259] According to the first stage of the exemplary method, in the noise characterization stage, energy is detected within both the imaging band (in-band) and the noise detection band in the absence of imaging energy (such as during the non-imaging noise characterization stage), thereby obtaining correlated measurements of the in-band and detection band noise characterization waveforms. Samples from the in-band noise characterization waveform and the detection band noise characterization waveform are recorded as array pairs, wherein the array pair refers to a sampled in-band waveform and a second corresponding sampled detection band noise detection waveform recorded simultaneously.
[0260] The detection band noise characterization array and the in-band noise characterization array can be windowed in 566 and 565. The windows may be sliding windows with arbitrary overlap. Optionally, the windows may be centered at the peak noise amplitude or time-locked at the noise amplitude threshold. The windows may also be conditioned to reduce artifacts caused by windowing, such as by applying a window function, such as a Hamming window, a Blackman window, or other window functions well known in the field of signal processing techniques. Array data (or its windows) may be processed to identify the presence of one or more noise patterns in 570.
[0261] Now, referring to FIG. 7a, one or more detection band noise characterization arrays are processed to identify waveform patterns associated with patterns in the in-band noise characterization waveforms. The pattern recognizer in step (570) extracts features from the detection band characterization arrays and classifies the features into noise 'classes' using a prediction model. The extracted features may be statistical features (including, but not limited to, variance, standard deviation, power, skewness, and kurtosis) in the time domain, frequency domain (e.g., peak frequency), and time-frequency domain (e.g., wavelet coefficients). The selection of features to be extracted may be performed in advance using feature selection algorithms such as forward selection or reverse elimination methods.
[0262] The extracted features are fed to a prediction model in step (570), and the prediction model can be trained to identify patterns in the detection band waveform and assign these patterns to noise classes. For example, machine learning methods may be used to train the prediction model to recognize patterns in the detection band noise characterization array using the extracted features. The prediction model may include an unsupervised learning model (such as k-means clustering) or a supervised learning model (such as a linear classifier, an artificial neural network, or a nearest neighbor classifier). Supervised learning may be used if prior information regarding the noise sources is known; for example, the sources and sequences of noise patterns may be known in advance, and noise class labels may be assigned to the waveform patterns in the detection band noise characterization waveform. The prediction model in step (570) may also receive class weights or a priori probabilities as input. The higher the a priori probability or weight of a class, the higher the likelihood that the class will be recognized.
[0263] The database illustrated in step (575) may store in-band noise characterization waveform patterns known to be co-occurring with detection band noise characterization waveform patterns. For example, the database may store exemplary or average temporal in-band noise characterization waveform patterns that are paired with features of co-occurring detection band noise characterization patterns and noise class labels. The detection band noise characterization waveform patterns and their co-occurring in-band noise characterization waveform patterns may be determined by window or in other ways. After image generation in steps (230 and 231), additional noise characterization may be performed in the time domain or in the spatial domain, in which case spatial features may also be extracted in step (570). The database may be any suitable format used in computing, such as lookup tables.
[0264] After correlating the detection band noise patterns with the in-band waveform noise patterns through the noise characterization stage described above, these correlations can be used during imaging to perform noise reduction of the in-band imaging waveforms based on the identification of patterns in one or more detection band imaging waveforms.
[0265] Referring to FIG. 7b, the in-band imaging waveform is obtained by applying an imaging-band bandpass filter (200) to the waveform detected by the imaging transducer receiving channel while the transducer is receiving imaging energy, and optionally performing envelope detection (210). The detected band imaging waveform is obtained by applying a noise detection bandpass filter (410) to the waveform detected by the imaging transducer receiving channel, and optionally performing envelope detection (411). The detected band imaging waveforms and the in-band imaging waveforms can be windowed at 566 and 565, similar to the windowing step in the noise characterization stage. The waveforms can be sampled and represented as arrays.
[0266] Similar to the feature extraction step in the noise characterization stage, features can be extracted from the detection band imaging array (or its windows). To identify the presence of one or more patterns in the detection band imaging waveform, the extracted features and optionally class weights are used by a pattern recognizer (570) trained in the noise characterization stage (described above).
[0267] The period (i.e., repetition frequency) of the noise pattern can be used to adjust the a priori probability of the pattern class (e.g., in the case of a Bayes classifier) or to adjust the weight of the class (e.g., in the case of a Support Vector Machine) while applying the pattern classification algorithm in 570. The higher the a priori probability or weight of the class, the higher the probability that the class will be recognized. If the repetition interval of the pattern is known, the pattern is expected to exist with a higher probability at given times. The a priori probability or weight of the class can be adjusted to be higher at those times, thereby increasing the probability that the pattern classifier will recognize the noise pattern. This repetition interval can be determined at the noise characterization stage and stored in the database (578), or loaded from a pre-stored database (local, networked, or cloud storage).
[0268] Patterns in the detection band imaging waveform identified as being associated with one or more noise classes in step (570) are subsequently used to generate noise corrections for the in-band imaging waveform (e.g., in-band imaging array). These corrections may be generated based on discovering the correlated in-band patterns in step (575), wherein the matched feature sets of the detection band noise characterization waveform patterns, the in-band noise characterization waveform patterns, and the noise class labels are stored in an explorable database or other classification scheme.
[0269] In the exemplary method illustrated in FIG. 7b, in-band noise correction is generated per window and, after delay and / or amplitude adjustment and / or shape adjustment (510), which arbitrarily aligns the in-band noise pattern retrieved from the database (575) or other classification scheme with the in-band waveform in time, is subtracted from the in-band imaging waveform at 525 per window.
[0270] In one exemplary implementation, during the noise characterization stage (when imaging energy is not received), the detection band noise characterization array is first processed to extract one or more features, and these features are subsequently stored. The time intervals at which a given noise pattern is detected may also be determined in this stage and stored in 578. The corresponding correlated time pattern in the in-band noise characterization array is also stored in 575 (e.g., in a lookup table).
[0271] According to the present example, during the imaging stage where noise removal is to be implemented, detection band imaging arrays for one or more detection band imaging waveforms primarily contain noise and are processed through the same feature extraction process. A weight vector assigning weights (or a priori probabilities) to noise patterns of each class may be arbitrarily obtained. The repetition frequency of each pattern may be loaded from a database (578) generated in the noise characterization stage or from a pre-stored database. The weight for each class may be dynamically adjusted to depend on the repetition frequency of the pattern, the time instance in which the pattern was previously detected, and the certainty of the pattern previously detected. Features extracted from the detection band imaging waveform, and optionally class weights, are fed again to a trained prediction model (trained in the noise characterization stage) capable of identifying noise patterns in the detection band imaging waveform and assigning a class to them. The corresponding correlated in-band noise pattern for the noise class is then obtained from the database in 575, where, for example, the features of in-band noise waveform patterns and the detected band waveform patterns for each noise class may be stored during the noise characterization stage (e.g., in-band time waveforms are stored in a lookup table). The in-band noise waveform pattern extracted from class comparison may be, for example, the average of all concurrent in-band noise patterns for the current noise class, or the in-band noise pattern whose concurrent detected band pattern features are closest to the features of the current detected band imaging array, for example, determined by nearest neighbor calculation. This concurrent pattern is then subtracted from the input after amplitude and delay adjustment to obtain a noise-reduced in-band imaging waveform.
[0272] FIGS. 7a and 7b illustrate an exemplary implementation in which a single detection band waveform (at least partially containing energy existing outside the imaging band) is generated by a single detection bandpass filter (410)—from which an input to a pattern recognizer (570) is derived. Alternatively, multiple detection band waveforms may be generated by multiple detection bandpass filters—at least one of which is out-of-band.
[0273] In addition to at least one detection band waveform carrying out-of-band noise, one or more detection band waveforms may carry noise within all or part of the imaging band (i.e., within-band noise detection waveforms). Such within-band data may be useful for the pattern recognizer (570) to verify that the noise predicted by the out-of-band noise detection waveforms is actually present in the imaging band (either during the noise characterization stage or during imaging).
[0274] For example, in-band noise detection imaging waveforms can also be used by a pattern recognizer to identify noise sources. For example, if the energy in some subbands of the imaging band is substantially different from that in one or more other subbands of the imaging band, or from the net energy within the imaging band, a noise source associated with the imaging subband can be identified. For example, a peak filter centered at 8 MHz can be used to acquire in-band noise detection waveforms within the imaging band in the range of 7 to 13 MHz, and a detection bandpass filter with a passband of 15 to 25 MHz can be used to acquire out-of-band noise detection waveforms. If the 8 MHz in-band noise detection waveform detects an increase in energy compared to the energy in the out-of-band noise detection waveform in the range of 15 to 25 MHz, the pattern recognizer can adjust its weights to preferentially detect a specific noise source (i.e., noise class). This system can better select correlated in-band noise patterns from the database in 573 to remove the 8 MHz peak than when relying only on out-of-band information for the imaging band.
[0275] Although the preceding exemplary embodiments have been disclosed in the context of detecting temporal patterns in raw or envelope-processed signals, it will be understood that the preceding algorithm can alternatively be adapted for an implementation using image data. For example, image data (e.g., B-mode image data) can be processed to determine spatial noise patterns instead of processing signals in the time domain (e.g., RF or envelope-detected). These alternative embodiments are illustrated by dashed paths in FIG. 7b, where decimation (220 and 221) and B-mode image line generation (230 and 231) are performed prior to 570 and 575.
[0276] Alternatively, when processing images in the spatial domain, 2D imaging windows can be used to detect spatial patterns, as with B-mode image data. For B-mode data, texture features can be extracted in the spatial domain (e.g., gray level co-occurrence matrices), the frequency domain (e.g., Fourier spectrum measurements), or the spatial frequency domain (e.g., the energy of 2D wavelet coefficients).
[0277] Referring now to FIG. 7d, an exemplary alternative embodiment is illustrated in which a reference noise detection waveform is used instead of the detection band imaging waveform of FIG. 7b when performing pattern recognition during imaging. Similarly, referring now to FIG. 7c, a reference noise characterization waveform may be used during an initial pattern recognition stage performed in the absence of an imaging signal. Referring to FIG. 7a and FIG. 7b, the algorithms or schemas described above may be adapted to this exemplary embodiment by replacing the detection band noise characterization waveform (and associated array scales) with the reference noise characterization waveform (illustrated in FIG. 7c), and replacing the detection band imaging waveform (and associated array scales) with the reference noise detection waveform (illustrated in FIG. 7d).
[0278] Example 5: Noise reduction based on relative energy scales
[0279] In the present exemplary embodiment, noise reduction is performed by selectively attenuating a windowed portion of an in-band imaging waveform based on criteria evaluated according to measurements from one or more detection band imaging waveforms—at least one of which is an out-band imaging waveform. Attenuation may refer, for example, to subtract a derived attenuation value from the envelope of the windowed in-band imaging waveform and / or to multiply the windowed in-band imaging waveform or its envelope by an attenuation factor, wherein the attenuation value and / or the attenuation factor are determined from measurements for the noise detection imaging waveforms or reference noise detection waveforms.
[0280] According to the first stage of the exemplary method, in the noise characterization stage, energy is detected within both the imaging band (in-band) and the noise detection band in the absence of imaging energy (such as during the non-imaging noise characterization stage), thereby obtaining correlated measurements of the in-band noise and the detection band noise. At least one noise detection band is out-of-band. Samples from the in-band noise characterization waveform and the detection band noise characterization waveform are recorded as array pairs, wherein the array pair refers to a sampled in-band waveform and a second corresponding sampled detection band waveform recorded simultaneously.
[0281] Optionally, the noise characterization stage may include an additional stage referred to as a baseline noise characterization stage when it is known or expected that the imaging transducer receiving circuit does not receive imaging energy and does not receive noise energy. As illustrated in FIG. 6a, an in-band baseline noise characterization array can be obtained by applying an imaging bandpass filter (200) to an input waveform detected from the imaging transducer receiving channel (13) when imaging energy is not received and when noise energy is not received, and by arbitrarily detecting the envelope of the waveform filtered at 210. A detection band baseline noise characterization array can be obtained by applying a detection band filter (410) to an input waveform from the imaging transducer receiving channel (13) and by arbitrarily detecting the envelope of the waveform filtered at 411. The in-band baseline noise characterization array and the detection band baseline noise characterization array, measured in the absence of imaging energy and noise energy, respectively, are denoted as Gi and Gn. A given array pair is Gi w and Gn wTo obtain pairs of windowed arrays denoted by , they may be arbitrarily segmented according to multiple time windows, as exemplified in 565 and 566 in FIG. 6a. The windows may be sliding windows having an arbitrary overlap between adjacent windows. In the baseline noise characterization stage, one or more noise measurements may be calculated from the energy measurements per window. For example, the maximum power within the windowed in-band baseline noise characterization array may be denoted as Ti. Similarly, the maximum power within the windowed out-band baseline noise characterization array may be denoted as Tn.
[0282] The noise characterization stage includes a stage where the imaging transducer receiving circuit is not receiving imaging energy and is expected to receive noise energy. Referring again to FIG. 6a, the in-band noise characterization array (407) can be obtained by applying an imaging bandpass filter (200) to an input waveform detected from the imaging transducer receiving channel (13) when imaging energy is not received, and by arbitrarily detecting the envelope of the waveform filtered at 210. The detection band noise characterization array can be obtained by applying a noise detection band filter (410) to an input waveform from the imaging transducer receiving channel (13) and by arbitrarily detecting the envelope of the waveform filtered at 411. The in-band noise characterization array (407) and the detection band noise characterization array (408), measured in the absence of imaging energy, are respectively labeled as Ci and Cn (as shown in FIG. 6a).
[0283] The given array pair is Ci w and Cn wTo obtain windowed array pairs denoted by , they may be arbitrarily segmented according to a plurality of time windows, as exemplified in 565 and 566 in FIG. 6a. The windows may be sliding windows having an arbitrary overlap between adjacent windows. In another example, the windows may be centered on the peak amplitudes of noise waveforms in one or more detection bands and / or imaging bands. In yet another example, the time windows may be time-locked to noise onsets determined when the amplitude of the noise in one or more detection bands and / or imaging bands exceeds predefined thresholds. For example, the thresholds may be proportional to the parameters Tn and / or Ti obtained during the baseline noise characterization stage.
[0284] Pairs of arrays of in-band and detection band noise characteristics (Ci w and Cn w ) is processed to obtain one or more metrics associated with energy in the imaging band and the noise detection band for each time window to characterize the relative intensity of noise within two bands. For example, as shown in FIG. 6a, each pair of windows (Ci w and Cn w For ), the power in the imaging band and the power in the noise detection band can be calculated at 570 and 572.
[0285] Noise characterization measurements may be calculated from select windows only, wherein selection criteria may be evaluated based on in-band energy measurements and, optionally, detected band energy measurements. For example, only windows where the in-band power exceeds a predefined threshold may be selected to obtain noise characterization measurements, as illustrated in FIG. 6b. The threshold may be proportional to Ti obtained during the baseline noise characterization stage. In another example, only windows where the out-of-band power exceeds a predefined threshold may be selected to obtain noise characterization measurements. The threshold may be derived from Tn obtained during the baseline noise characterization stage.
[0286] In the examples at hand, it should be understood that the maximum and minimum values may refer to either the upper and lower percentiles or the actual maximum and minimum values. For example, the 98th and 2nd percentile values may be used instead of the maximum and minimum values. Other statistical thresholds, such as the 95th and 5th percentiles, the 90th and 10th percentiles, and the 80th and 20th percentiles, or others, may be used to represent the maximum and minimum values for characterization purposes.
[0287] One or more noise measurements can be calculated from window-wise energy measurements in the noise characterization stage and can be used to define the relationship between power in the imaging band and power in the detection band in the presence of a noise source. For example, one or more pairs of in-band and detection band power values can be selected as inflection points to generate a piece-wise linear function to define the relationship between power in the imaging band and power in the detection band in the presence of a noise source, as illustrated in FIG. 6b.
[0288] In one example, as illustrated in FIG. 6b, a discriminative linear function defining the relationship between in-band power and detected band power in the presence of a noise source can be generated based on the maximum and minimum power values obtained from the noise characterization stage. The minimum and maximum in-band power values from the noise characterization stage can be evaluated (e.g., as absolute maximum / minimum values or using statistical measures), and each, Pi min and Pi max It can be denoted as. In one exemplary implementation, the in-band power is Pi min A set of windows falling within a pre-selected range (e.g., within a percentile range) relative to can be identified, and among the identified set of windows, the minimum detection band power is Pn min It can be selected as. Similarly, the in-band power is Pi max A set of windows falling within a pre-selected range based on can be identified, and among the identified set of windows, the minimum detection band power is Pn max It can be selected as. Power pairs (Pn min , Pi min ) and (Pn max , Pi max ) can be used to fit a function that defines the estimated relationship between in-band power and detection band power. The exemplary implementation is merely one non-limiting example of selecting values of in-band and detection band powers to provide appropriate fitting points and / or functional relationship between in-band power and detection band power, and other methods may be used alternatively.
[0289] Optionally, the ratio of the power in the in-band noise characteristic waveform to the power in the detection band noise characteristic waveform can be calculated per window, and the maximum ratio (R) across multiple windows off )(an exemplary use of this amount when determining whether to apply noise correction during imaging is described below) can be obtained.
[0290] One or more relationships f(Pn) between the power (Pi) in the in-band noise characterization waveform and the power (Pn) in the detection band noise characterization waveform can be obtained. For example, as shown in FIG. 6b, f(Pn) is the slope, intercept, and / or inflection points calculated at the noise characterization stage (Pn min , Pi min ) and (Pn max , Pi max It can be a discriminative linear function defined by ). In another example, f(Pn) can be a nonlinear polynomial, or a combination of one or more linear or nonlinear polynomials. In yet another example, f() can be a set of values defined for one or more ranges of Pn values. For example, f(Pn) is a1 ≤ Pn <a2에 대해 값 Pi a It can be assigned, and f(Pn) is b1≤Pn <b2에 대해 값 Pi b It can be assigned, and likewise, where [a1, a2] and [b1, b2], etc., are non-overlapping intervals of Pn. A set of windows of a detection band noise characterization array with power between a1 and a2 can be identified, and the associated windows of an in-band noise characterization array are Pi for the interval [a1, a2]. a It can be used to determine the value of. For example, Pi aThe detection band power (Pn) may be a representative power value (such as maximum, mean, median, or any other measure) calculated from the in-band power measurements of all windows within the range [a1, a2].
[0291] Noise measurements obtained in the noise characterization stage can be used to perform noise reduction of the in-band imaging waveform obtained during imaging, as illustrated in FIGS. 6c and 6d. Optionally, after envelope detection at 210 and 411, input waveforms are filtered as illustrated in 200 and 410 to provide the in-band imaging waveform (407) and the detection band imaging waveform (408). Waveforms can be sampled to obtain the in-band imaging arrays and the detection band imaging arrays.
[0292] In the present exemplary embodiments, the term "in-band imaging array" is used to refer to a sampled in-band imaging waveform. The term "detection-band imaging array" is used to refer to a sampled detection-band imaging waveform. A set of arrays may be recorded, wherein each array may be associated with a given scan line. For example, a first in-band imaging array may be associated with a first scan line, and a second in-band imaging array may be associated with a second scan line, and so on. The in-band and detection-band imaging arrays, each, and It is denoted as, and here is an index that identifies a given acquisition period, such as one corresponding to a scan line.
[0293] The arrays can be windowed, as illustrated in 565 and 566 in FIG. 6c, using the same window properties as those used to window the noise characterization arrays in the characterization stage. The in-band imaging arrays and detection band imaging arrays, which are temporally segmented according to the windows, are each, and It is denoted as, where the subscript w is an integer representing the window number. For example, Qi 1,10 ... refers to the tenth window portion of the in-band imaging array corresponding to the first scan line. For each windowed portion of the in-band imaging array and the detection band imaging array, power or other suitable energy measurements may be calculated as shown in 570 and 572. The in-band and detection band power values, respectively, per window, and It can be calculated as. These energy measurements can then be used to suppress noise based on measurements obtained in the noise characterization stage.
[0294] In one exemplary implementation illustrated in FIG. 6c, the power value at 525 is an in-band imaging array Noise can be suppressed by subtracting from the envelope. Noise detection band power is, based on the function f() obtained in the noise characterization stage, the in-band imaging array window Noise energy within It can be used to estimate. For example, Noise energy within is from 575 It can be estimated as. Estimated in-band noise can be arbitrarily scaled by multiplying by a scaling factor β (where 0 ≤ β ≤ 1). Scaled estimated noise is to obtain the subtraction value , or It can be arbitrarily clamped below an upper limit, such as but not limited to. The reduction value is from 525 to obtain a noise-reduced waveform envelope. It can be reduced from the envelope.
[0295] In another exemplary implementation illustrated in FIG. 6d, an in-band imaging array at 526 Noise can be suppressed by multiplying the elements of with a damping factor. During imaging, imaging window The in-band power at It can be calculated as. Noise detection band power is, based on the function f() obtained in the noise characterization stage, the in-band imaging array window Noise energy within It can be used to estimate. For example, Noise energy within is from 575 It can be estimated as. Estimated in-band noise It can be arbitrarily scaled by multiplying by a scaling factor β (where 0 ≤ β ≤ 1). The damping factor is It can be selected in proportion to.
[0296] In some exemplary implementations, the scaling factor β may be selected to be between 0 and 1 (unity). In ultrasound, where the attenuation of ultrasound energy reduces imaging energy over time, the determination of β may depend on the depth of the window within the waveform (thus corresponding to the depth within the imaged tissue). The parameter β may be user-controlled at will. The attenuation factor may be arbitrarily clamped below an upper limit, such as, but not limited to, 1, 0.95, 0.9, or 0.8. The attenuation factor may additionally or alternatively be clamped above a lower limit, such as, but not limited to, 0, 0.01, 0.05, or 0.1. To obtain a noise-reduced array, the attenuation factor in 526 array Or it can be multiplied by his envelope.
[0297] Pattern recognition for class-specific noise reduction
[0298] During the noise characterization stage (e.g., see FIG. 6a), the system arbitrarily array pairs Cn w and Ci w It can be configured to group into one or more categories referred to as classes. Referring again to the pattern recognizer described in step (570) of FIG. 7a, one or more detection band noise characterization waveforms can be processed by the pattern recognizer to identify one or more noise pattern classes and to assign the classes to time windows associated with the identified noise patterns. A set of windows of detection band noise characterization waveforms belonging to class k is selected and { w_k It can be denoted as}. In-band and out-of-band power measures for windows within the set {w_k}, using methods similar to those described in the preceding paragraphs, functional relationships specific to class k It can be used to derive.
[0299] This system may be configured to use a pattern recognizer to identify noise patterns when the imaging transducer circuit is receiving imaging energy. Using methods similar to those described for step (570) of FIG. 7b, one or more detection band imaging waveforms may be processed by the pattern recognizer to identify one or more noise pattern classes and to assign the classes to time windows associated with the identified noise patterns. The estimated in-band noise power may be derived from a functional relationship f() specific to the identified class. For example, for windows belonging to class 1, the function f1() is the estimated in-band noise (i.e., It can be used to derive ). Similarly, for windows belonging to Class 2, the function f2() can be used to derive estimated in-band noise, etc., where f1(), f2(), f3(), etc. are obtained in the noise characterization stage. As previously described, a given window of the in-band imaging waveform The noise at is the subtract value By subtracting from the envelope of, or It can be reduced by multiplying by a damping factor, where the reduction value or damping factor is It is derived from. As an additional specific example, noise classified as Class 1 by the pattern recognizer may originate from an electroanatomical mapping system, and noise classified as Class 2 may originate from an ablation energy generator. Thus, f1() can be used to estimate in-band noise generated by the electroanatomical mapping system based on power in the noise detection band when the noise is recognized by this system as coming from the mapping system, and f2() can be used to estimate in-band noise generated by the ablation generator based on power in the noise detection band when the noise is recognized by this system as coming from the ablation generator.
[0300] In another example, noise-emitting systems may be monitored to determine which functional relationship to use for noise estimation and suppression. For instance, the controls of an ablation generator may be monitored to enable a binary gating signal when the generator actively generates energy (and associated noise). This gating signal can be used to determine the time periods during which the function f2() should be used for in-band noise estimation.
[0301] Selectively perform noise reduction
[0302] Energy measurements can be used arbitrarily to estimate whether an in-band imaging array is likely to exhibit a low signal-to-noise ratio. In other words, energy measurements can be used to classify windows regarding whether to apply noise reduction correction (by multiplying with a subtraction or attenuation factor).
[0303] In one exemplary implementation, the decision regarding whether to apply noise reduction correction for a given window (w) can be made based on the power detected in the noise detection band. If it exceeds a predefined threshold, Noise correction is applied to it. The threshold value can be obtained in the noise characterization stage.
[0304] In other cases, the decision regarding whether to apply noise reduction correction for a given window may be based on the ratio of power in the imaging band to power in the noise detection band. For example, in the noise characterization stage, for each window of the pair of in-band noise characterization arrays and detection band noise characterization arrays, the ratio of power in the imaging band to power in the detection band may be obtained. R Off The representative maximum ratio across all windows, denoted as , can be used as a threshold to determine whether to apply noise reduction correction when the imaging transducer receiving circuit receives imaging energy, as described below.
[0305] The decision regarding whether to apply noise reduction correction to a given window (w) of the in-band imaging waveform is R On w This can be based on the ratio of power in the imaging band to power in the noise detection band, denoted as . In one exemplary implementation, R On w go Compared with, and here is the relaxation parameter and R Off is the representative maximum ratio calculated in the noise characterization stage. R On w this If it is determined to be larger, the signal-to-noise ratio of the in-band signal is estimated to be sufficiently high, and noise reduction correction is not applied. Conversely, R On w this If it is determined to be smaller than or equal to, the signal-to-noise ratio is estimated to be sufficiently low to guarantee the application of noise reduction correction.
[0306] The value of can be used to adjust sensitivity to the signal-to-noise ratio, and in some cases, can be used as an adjustment factor for cases where weak parts of the signal would otherwise be suppressed. Lowering the value of lowers the threshold for applying noise reduction, reducing the number of windows undergoing noise reduction and allowing more imaging energy (and noise) to persist in the final output.
[0307] In some exemplary implementations, can be selected to be between 0 and 1. In ultrasound, if the attenuation of ultrasound energy reduces the imaging energy over time, The determination of may depend on the depth of the window within the waveform (thus corresponding to the depth within the imaged tissue). Parameter This can be arbitrarily controlled by the user. For example, in cases where the texture or other structural aspects of the image are perceived to be unnecessarily or excessively attenuated, the user may reduce the value of this parameter to reduce the effect of noise reduction.
[0308] For windows identified for noise reduction, any suitable noise reduction or suppression method may be used to reduce the noise of the in-band imaging array within the window. It will be understood that a wide variety of noise reduction corrections may be applied, such as, but not limited to, corrections involving subtraction as shown in FIG. 6c and / or multiplication with a damping factor as shown in FIG. 6d.
[0309] In some cases, noise suppression may result in incorrect noise reduction on a window-by-window basis. For example, some windows containing a small amount of imaging energy may undergo noise reduction based on an inadvertent incorrect determination of the low signal-to-noise ratio in the window (i.e., incorrect window classification). This may result in some small image "holes" in homogeneous signal regions around the image, or residual noise pixels in low-noise regions around the image. In one exemplary embodiment, the state of adjacent windows (i.e., windows in spatial neighborhoods) may be used to determine whether a window identified as suitable for noise reduction should actually undergo such a process. If a given window is identified as suitable for noise reduction according to the methods described above, adjacent windows of one or more adjacent arrays (i.e., arrays corresponding to adjacent scan lines) may be used to evaluate whether noise reduction by amplitude attenuation of the given window should be performed.
[0310] For example, for a given scan line, if a given window is identified as unsuitable for noise reduction but one or more adjacent windows are identified as suitable for noise reduction, this window may be flagged as potentially misclassified. The state of a given window may be overridden, and the given window may instead be identified as suitable for noise reduction so that noise reduction is applied to the given window. Conversely, if a given window is identified as suitable for noise reduction but adjacent windows within adjacent arrays are identified as unsuitable for noise reduction, the initial determination of the state of the given window may be overridden so that the given window is instead identified as unsuitable for noise reduction and flagged as a potentially misclassified window. Samples from the misclassified windows may be replaced with samples from one or more non-noisy neighboring windows (i.e., replaced by or interpolated from samples from adjacent windows identified as unsuitable for noise reduction according to the methods described above) after performing arbitrary delay and amplitude adjustments.
[0311] An exemplary implementation of this method is illustrated in FIG. 6e, where adjacent windows in adjacent arrays are examined to determine whether the classification of the current window matches its surrounding windows in adjacent arrays. Since two windows classified as containing mainly noise (Array 2, Window 4, and Window 5) (marked as “N”) are surrounded by windows classified as having sufficiently high signal power (marked as “S”) to avoid the need for noise reduction, as illustrated in FIG. 6f, the two “N” windows may be replaced with samples from neighboring “S” windows (e.g., by radiation or by interpolation using amplitude adjustment and / or shape adjustment), and noise reduction by noise estimation will not be performed for these windows. Although this exemplary implementation uses two adjacent windows on both sides of the array when checking the state of a given window, other embodiments may use any number of adjacent windows.
[0312] When evaluating whether the classification of a given window should be changed, one or more temporally adjacent windows before and after the given window may also be used. For example, in FIG. 6e, since the preceding window and the ongoing window of the array are marked as "N", window 4 of array 5, which was initially marked as "S", can be reclassified as "N" as shown in FIG. 6f.
[0313] Although the present exemplary embodiment involves the processing of signals prior to image processing, the present exemplary embodiment may be adapted to process image data differently from the processing of time-domain signals. For example, multiple in-band and detection-band imaging arrays representing multiple adjacent scan lines may be acquired and post-processed to acquire in-band image and detection-band image frames. The in-band and detection-band image pixels, respectively, and It is denoted as, and here represents the scan line, and d is the depth. The detection band image can be used to evaluate attenuation values (i.e., through subtraction or multiplication) for each pixel within the in-band image. These attenuation values are calculated pixel by pixel (i.e., The value at (which can be used to attenuate pixel intensity in) or by processing a region of interest in a detection band image corresponding to the local spatial neighbors of the in-band image pixels (e.g., The values of surrounding 3x3 neighbors (e.g., in polar or Cartesian coordinates) (Can be used to attenuate pixel intensity in) Can be obtained from corresponding pixels within the detection band image.
[0314] Now, referring to FIGS. 6g, 6h, and 6j, exemplary embodiments of alternatives in which a reference receiving channel is used are illustrated for determining a threshold value (Fig. 6h) or attenuation factor (Fig. 6i) suitable for noise suppression during noise characterization (Fig. 6g) and during imaging. In FIG. 6g, when performing noise characterization, a reference noise characterization waveform (409) is used instead of the detection band noise characterization waveform of FIG. 6a. The reference channel filter may be an imaging bandpass filter. Alternatively, if the noise estimation gains from the input of out-of-band noise, the reference channel filter may differ from the imaging bandpass filter. In FIG. 6h, a reference noise detection waveform (439) is used instead of the detection band imaging waveform of FIG. 6b to determine threshold values and apply noise reduction by subtraction. Similarly, in FIG. 6i, a reference noise detection waveform (439) is used instead of the detection band imaging waveform of FIG. 6b to determine and apply a damping factor for noise reduction. With reference to FIG. 6a through 6f, the methods described above can therefore be adapted to the present exemplary embodiment by replacing the detection band noise characterization waveform (408) (and associated power scales) with the reference noise characterization waveform (409) shown in FIG. 6g, and replacing the detection band imaging waveform (438) (and associated array and power scales) with the reference noise detection waveform (439) shown in FIG. 6h and FIG. 6i.
[0315] Example 6: Noise reduction of pseudo-periodic noise sources
[0316] In this exemplary embodiment, noise reduction is performed by estimating and subtracting in-band noise, where the noise is expected to originate from a pseudo-periodic noise source or a sequence of pseudo-periodic noise sources. When the imaging transducer receiving circuit is not receiving imaging energy, the in-band noise is estimated based on measurements taken during the noise characterization stage.
[0317] According to the first stage of the exemplary method, waveforms are detected and sampled within both the imaging band and the noise detection band in the absence of imaging energy (e.g., when the ultrasonic transducer is not receiving imaging energy), thereby obtaining a pair of concurrent in-band and detection band (out-of-band) noise characterization arrays.
[0318] As illustrated in FIG. 8a, the in-band noise characteristic waveform (590) can be obtained by applying an imaging bandpass filter (200) to an input waveform detected from an imaging transducer receiving channel (13) and arbitrarily detecting the envelope (210) of the filtered data. The detection band noise characteristic waveform (595) can be obtained by applying a noise detection bandpass filter (410) to an input waveform from an imaging transducer receiving channel (at least one noise detection band includes a signal from outside the imaging band) and arbitrarily detecting the envelope (411) of the filtered data. The in-band noise characteristic waveforms and the detection band noise characteristic waveforms measured in the absence of imaging energy are, respectively, Ci In-band noise characterization arrays and detection band noise characterization arrays, denoted as Ci and Cn, can be sampled to acquire them. Ci (and Cn) must capture one or more periods of a periodic noise source.
[0319] After acquiring in-band and detection band noise characterization arrays in the noise characterization stage, the correlation between the detection band characterization array and the detection band imaging array can be used to estimate adjustment parameters for subtracting the in-band characterization array from the in-band imaging array.
[0320] As illustrated in FIG. 8b, when performing imaging, input waveforms from the imaging transducer receiving circuit (13) are filtered as illustrated in 200 and 410 to provide in-band imaging waveforms and detection band imaging waveforms. These waveforms may be sampled (before or after performing envelope detection) to acquire an in-band imaging array (537) and a detection band imaging array (538). The in-band and detection band imaging arrays are denoted as Qi and Qn, respectively.
[0321] According to the present method, each in-band imaging array is processed for noise reduction using an in-band noise characterization array (Ci) to subtract noise from the in-band imaging array (Qi). However, in order to perform noise reduction through subtraction, the in-band noise characterization array (Ci) must be temporally aligned with the in-band imaging array (Qi) so that the noise occurs concurrently. Such alignment is possible in the case of a periodic noise source that generates noise in an imaging band that is correlated with the noise in the noise detection band.
[0322] Temporal alignment in 565 and 566 (as described in the preceding exemplary embodiments) can be achieved, for example, by segmenting the in-band and detection-band imaging arrays (Qi and Qn) into a plurality of time windows. The windows should preferably be long enough to capture one or more periods of a periodic noise source. Imaging arrays temporally segmented according to the windows Qiw and Q&A w It is denoted as, where the subscript w is an integer representing the window number.
[0323] In one exemplary embodiment, time alignment can be achieved per window. In this exemplary implementation, time alignment is achieved within each window, between the detection band noise characterization array (Cn) and the detection band imaging array (Qn w This can be achieved by calculating the cross-correlation between ) and selecting a relative time delay corresponding to the maximum cross-correlation, as illustrated in 580. Due to the simultaneous relationship between noise in the imaging band and noise in the noise detection band, this time delay is applied to the in-band noise characterization array (Ci) and the in-band imaging array (Qi) per window, as illustrated in 510. w It can also be applied to align with respect to ). The scaling factor can also be applied to the aligned in-band noise characterization array.
[0324] Ci w The windowed portion of the aligned in-band noise characterization array, denoted by, is followed by an in-band imaging array (Qi w Subtracted from ), and after arbitrarily subtracting and taking the absolute value or applying a floor function to remove negative values, a noise-reduced in-band imaging array ( Qi w ) results in. This process can then be repeated for each additional window where noise reduction is desired.
[0325] In one exemplary implementation where adjacent windows overlap, an in-band noise characterization array ( Ci w Aligned windowed segments of ) in-band imaging array (Qi wA scaling factor may be applied when subtracting from ). For example, the subtraction (and arbitrary modulus) can be calculated according to the following: , here is a scaling factor for considering windowing, and here and, here is the overlap factor. For example, In this case, the scaling factor is It will be. This implementation is provided to illustrate an exemplary method for scaling subtracted components, and it will be understood that other functional forms may be used alternatively.
[0326] Now, referring to FIGS. 8c and 8d, an exemplary embodiment of an alternative in which a reference receiving channel is used to determine an appropriate amplitude adjustment for noise suppression during noise characterization (Fig. 6g) and during imaging is illustrated. In FIG. 8c, when performing noise characterization, a reference noise characterization waveform (596) is used instead of the detection band noise characterization waveform (595) of FIG. 8a. Similarly, in FIG. 8d, a filtered reference noise detection waveform (439) is used instead of the detection band imaging waveform of FIG. 8b to determine and apply an amplitude adjustment. With reference to FIGS. 8a and 8b, the methods described above can therefore be adapted to the present exemplary embodiment by replacing the detection band noise characterization waveform (408) with the reference noise characterization waveform (409) shown in FIG. 8c, and replacing the detection band imaging waveform (438) (and associated arrays) with the reference noise detection waveform (439) shown in FIG. 8d.
[0327] Example 7: Noise reduction using multiple scans by changing the scan rate
[0328] This system may be configured to acquire a set of two or more in-band imaging arrays associated with imaging energy from the same scan line or scan lines having substantial spatial overlap. The imaging energy within the set of in-band imaging arrays will have redundant time / depth dependencies. Averaging the set of redundant in-band imaging arrays (or performing other statistical processing, such as evaluating minimum values) can suppress noise, provided that the noise itself is not time-locked to a trigger that prompts the imaging transducer receiving circuit to begin receiving imaging energy for each scan line. For example, if the pulse repetition frequency of the voltage pulse exciting the imaging ultrasonic transducer is 200 µs, periodic noise repeating every 2 µs will always have components at 0 µs, 2 µs, 4 µs, etc., for each in-band imaging array. However, if the pulse repetition frequency is adjusted to 199 µs, the first imaging array will have noise components at 0 µs, 2 µs, 4 µs, etc., and the second imaging array will have noise components at 1 µs, 3 µs, 5 µs, etc. After performing envelope detection, noise can be suppressed by averaging two consecutive overlapping in-band imaging arrays.
[0329] In this exemplary embodiment, noise in the detection band waveform (out-of-band) may be used to determine the period of one or more in-band noise sources. The system may be prompted to adjust its scan rate so that the imaging scan period does not become an integer multiple of the period of the noise source. For example, an auto-correlation function may be used to detect periodicity in the detection band imaging waveform while the imaging transducer receiving circuit is receiving imaging energy. Alternatively, the period of one or more noise sources may be determined from the detection band noise characterization waveform or the reference noise characterization waveform at the noise characterization stage when the imaging transducer is not receiving imaging energy, or may be loaded from a pre-stored database. The system may then be prompted to adjust its scan rate so that the imaging scan period does not become an integer multiple of the period of the noise source.
[0330] Once the optimal scan rate is determined and adjusted, in-band imaging waveforms acquired from multiple scan lines are sampled after envelope detection is performed to acquire a set of in-band imaging arrays. The in-band imaging array It is denoted as, and here is a scan line. Samples from the in-band imaging array are It is denoted as , where k = 1..K is the sample index and K is the number of samples in the array. This system can be configured to suppress noise by averaging across multiple arrays associated with adjacent scan lines or scan lines having significant spatial overlap (or performing other statistical measurements, such as taking the minimum). For example, if this system is configured to group three in-band imaging arrays, the samples Is It can be replaced by the average of. Arbitrarily, sample is a set It can be optionally retained after performing any other numerical analysis on it and determining whether the sample guarantees noise reduction. For example For , if the minimum value is greater than half the maximum value, the range of sample values may not be large enough to guarantee noise reduction by averaging, and therefore the samples remain unchanged.
[0331] Such a scheme may also be useful for MRI imaging in the presence of periodic noise sources by ensuring that RF excitation pulses are initiated at times unrelated to the timing or periodicity of noise sources in the local environment.
[0332] Additional feature: Selectively select samples to undergo noise reduction when using multiple scans
[0333] In this exemplary embodiment, the scan rate is adjusted as described above, and noise reduction is performed by selectively replacing parts of the in-band imaging array based on statistical measurements (such as average or minimum values) from arrays of multiple adjacent scan lines, where there is sufficient overlap in the scan area associated with the adjacent scan lines. Only segments of the in-band imaging array evaluated as noisy using detection band measurements are replaced.
[0334] As shown in FIG. 8e, input waveforms that are filtered as shown in 200 and 410 to provide in-band waveforms and detection band waveforms are sampled after envelope detection is performed, so that in-band imaging arrays and detection band imaging arrays can be obtained.
[0335] Data corresponding to multiple scan lines is recorded as a set of arrays for each scan line. A set of arrays for each scan line is acquired, wherein one in-band imaging array is acquired in the imaging band for each scan line, and at least one out-band imaging array is acquired for each scan line. The in-band and detection band imaging arrays within the pair are, respectively, and It is denoted as, and here is a scan line. Samples from the in-band imaging array and the detection band imaging array, respectively, and It is denoted as, where k = 1..K is the sample index and K is the number of samples in the array.
[0336] Each of the detection band imaging arrays is It is segmented into multiple windows containing J samples. The array segmented according to the windows is It is denoted as, where the subscript w is an integer representing the window number, and is an index representing a scan line, and are samples Includes, where k w is the index of the first sample in the window.
[0337] As illustrated in 600, each window in the detection band for the presence or absence of noise It is evaluated. If the noise (determined by measurements of waveform energy, such as peak, RMS, etc.) exceeds a threshold, the window is considered to be noisy. The threshold can be selected in the noise characterization stage.
[0338] According to the present exemplary embodiment, each detection band array window classified as having noise Regarding all concurrent in-band samples (i.e., ) can be identified as samples suitable for noise reduction.
[0339] When windows overlap, given sample It can be associated with more than one window. There may be cases where a sample is associated with both a noisy window and a noise-free window. In these cases, the system may be configured to pool noise evaluations from multiple out-of-band windows before determining whether the sample is suitable for noise reduction.
[0340] Samples considered suitable for noise reduction may be replaced in 620 using statistical measures (such as duplicate values, average, minimum, and similar values) from samples from arrays associated with adjacent scan lines, calculated as illustrated in 610. For example, sample If is considered suitable for noise reduction and this system is configured to group arrays from three scan lines, the sample Is It can be replaced with the minimum value.
[0341] Now, referring to FIG. 8f, an exemplary alternative embodiment is illustrated in which a reference waveform is used instead of the detection band imaging waveform of FIG. 8e when performing noise reduction during imaging. Referring to FIG. 8d, the methods described above can therefore be adapted to this exemplary embodiment by replacing the detection band imaging waveform (and associated arrays) with a filtered reference noise detection waveform.
[0342] Time Domain vs. Frequency Domain Processing
[0343] The preceding exemplary embodiments were disclosed within the context of time-domain processing. However, many of the exemplary embodiments disclosed herein may utilize frequency-domain or time-frequency-domain processing during one or more steps. For example, in FIGS. 6a and 6c, in steps (570 and 572), instead of using the ratio of maximum power and maximum power in the imaging band and the noise detection band, short-term Fourier transforms or wavelet transforms may be performed on the in-band and out-band waveforms by window or by array. Analysis of transform coefficients (e.g., average, mean square, and similar) may be used to characterize noise or detect windows when noise is present or absent. When noise is detected, instead of attenuating the signal in the time domain in 526, the transform coefficients for the current window may be attenuated. Subsequently, an inverse transform may be performed on the attenuated frequency-domain signal to obtain a noise-reduced time-domain signal.
[0344] In FIG. 2a (step (510)), FIG. 8b and FIG. 8d (step (580)), or any other embodiment where cross-correlation between two time-series waveforms is required, Fourier transform algorithms may be used for efficient calculation of cross-correlation.
[0345] In addition, as described above with reference to FIGS. 7a through 7d, machine learning algorithms may be used to classify noise patterns in step (570). These patterns may be defined by frequency domain and / or time-frequency domain features, which will require frequency domain or time-frequency domain processing of time-series waveforms.
[0346] Generalization beyond ultrasound
[0347] Although exemplary prior systems and methods for image noise reduction have been illustrated within the context of ultrasound imaging, it will be understood that the embodiments disclosed herein can be adapted to a wide variety of imaging devices, systems, and methods.
[0348] Another example of an imaging system that can be adapted for noise reduction according to the embodiments described above is a magnetic resonance imaging system. Now, referring to FIG. 9, an alternative exemplary system in which signals undergoing noise reduction are acquired from a magnetic resonance (MR) system is illustrated. The exemplary system includes a magnetic resonance scanner (50) that utilizes a main magnet (52) to generate a main magnetic field (B0) that generates polarization in a patient (60) or an examined subject. The exemplary system includes gradient coils (54) for generating magnetic field gradients. A reception coil (58) detects MR signals from the patient (60). The reception coil (58) may also be used as a transmission coil. Alternatively, a body coil (56) may be used to radiate and / or detect radio frequency (RF) pulses. RF pulses are generated by the RF unit (65), and magnetic field gradients are generated by the gradient unit (70). The manner in which MR signals are detected using a sequence of RF pulses and magnetic field gradients, and how MR images are generally reconstructed, are known to those skilled in the art.
[0349] The reference receiving circuit may include a coil located in the same room as the scanner but not positioned immediately adjacent to the sample being imaged (such as a patient) from which the desired MRI signals are being emitted. Electromagnetic noise traveling near the MRI machine will be detected by both the imaging receiving coil (58) and the reference receiving circuit. The coil of the reference receiving circuit may be oriented and positioned to have the potential to receive some of the same noise as the imaging receiving coil, but may be positioned far enough from the sample being imaged so that there is negligible imaging energy detected by the reference receiving circuit.
[0350] The reference receiving circuit coil can be tuned to have the same bandwidth as the imaging receiving coil (58), or it can have a different bandwidth that can still collect noise signals in an environment that is correlated with noise that can be coupled to the imaging receiving coil.
[0351] The reference receiving circuit may further include a set of receiving coils, such as three coils aligned orthogonally to each other. This will enable the collection of electromagnetic noise in such a way that the weighted sum of the noise collected from each of the three coils can be more closely matched with the noise collected from the imaging receiving coil, thus taking into account the directionality of the dominant electromagnetic noise sources in the MRI environment.
[0352] It will be understood that the MR system may have additional units or components not shown for clarity, such as but not limited to additional control or input devices, and additional sensing devices, such as devices for cardiac and / or respiratory gating. Furthermore, various units other than the described separation of individual units may be realized. It is possible for different components to be assembled into units or for different units to be combined with each other. Various units (described as functional units) may be designed as hardware, software, or a combination of hardware and software.
[0353] In the exemplary system illustrated in FIG. 9, the control and processing hardware (100) acquires magnetic resonance images of a patient (60) according to a suitable pulse sequence. The control and processing hardware (100) is interfaced with a magnetic resonance imaging scanner (50) to receive the acquired images and control the acquisition of the images. The control and processing hardware (100) receives image data from an RF unit (65) and processes the imaging data according to the methods described below.
[0354] The control and processing hardware (100) may be programmed with a set of instructions that, when executed on a processor, cause the system to perform one or more methods described in this disclosure to reduce noise in signals acquired from a magnetic resonance imaging system. For example, as illustrated in FIG. 9, the control and processing hardware (100) may be programmed with instructions in the form of a set of executable image processing modules, such as, but not limited to, a pulse sequence generation module (not shown), an image acquisition module (not shown), an image processing module (145), and a noise suppression module (150). The pulse sequence generation, image acquisition, and image processing modules may each be implemented using algorithms known to a person skilled in the art for pulse sequence generation, image acquisition, and image reconstruction. RF data is received from an RF coil (56 and / or 58) and optionally one or more reference receiving circuits. The data may be sampled and filtered to obtain an in-band waveform. In addition, a noise detection band waveform measured through a reference waveform via a reference receiving circuit or through filtering of RF from coils (56 and / or 58) is collected. One or more noise suppression methods described in FIGS. 2 through 8 may be used for noise suppression in 100. A pulse generation module establishes a sequence of RF pulses and magnetic field gradients according to a desired imaging sequence, and an image acquisition module stores the MR signals detected by the coils (56 and / or 58) in a raw data space. An image processing module (145) processes the acquired, arbitrarily noise-suppressed RF data to perform image reconstruction of the MR image.
[0355] By detecting noise correlated with noise in the bandwidth of the imaging signal (via a reference receiving channel or a noise detection band), the ability to improve the imaging signal is provided by estimating in-band noise and removing the estimated noise from the imaging signal.
[0356] This enables an improved SNR in a typical cage for shielding the MRI from environmental noise, or enables the operation of the MRI system in a more open / unshielded environment that is typically more affected by noise.
[0357] Examples
[0358] The following examples are provided to enable a person skilled in the art to understand and practice the embodiments of the present disclosure. They should not be construed as a limitation on the scope of the present disclosure, but merely as illustrative and representative examples.
[0359] Example 1: Noise reduction of an unknown noise source through a damping factor (Example of Example 5)
[0360] This example involved collecting ultrasonic data using an intra-cardiac echo (ICE) system in the presence of two noise sources. The transducer was configured to detect ultrasonic energy at a frequency of 9 MHz. Two band-pass filters were used in parallel to separate the radio frequency (RF) signal into an imaging band of 7 to 13 MHz and a noise detection band of 15 to 25 MHz, which is outside the frequency range of the emitted ultrasound.
[0361] The first noise source is the electroanatomical mapping system (Carto ®3) was. This system has an electromagnetic tracking module and an impedance-based tracking module, and for this purpose, patches to measure impedance and estimate device location are attached to the patient. These patches can couple a significant amount of noise to the imaging band of ICE images. In these experiments, a heart phantom was used in a saline bath. Electrodes from the impedance patches were immersed in the bath. The second noise source was Carto ® 3. It was to turn on the power of the ablation generator connected to the console. The noise generated from this second noise source, when the generator was powered on, Carto from the ablation generator ® It was determined that the noise was likely propagating into the saline bath through the 3 consoles and patch electrodes.
[0362] FIG. 10a illustrates an ultrasound image collected in the absence of a noise source, whereas FIG. 10b and FIG. 10c illustrate, respectively, the effects of the first and second noise sources on the imaging quality.
[0363] Noise reduction of ultrasonic waveforms detected by the ultrasonic transducer of the ICE console was performed according to the implementation of the method illustrated in FIGS. 6a and 6d.
[0364] During the first baseline noise characterization stage of this experiment, energy was detected within the imaging band when no imaging energy was received and when no noise energy was received, and an in-band baseline noise characterization array, denoted as Gi, was acquired. The in-band baseline noise characterization array was acquired by sampling the in-band waveform for 125 µs at 200 ms / s. A sliding and overlapping window (window size = 64 samples, 20% overlap) was used, and in-band power measures were calculated for each window. Representative maximum power (the 90th percentile across all windows) was calculated, and its value was assigned to the threshold value Ti.
[0365] During the second noise characterization stage of this experiment, in the absence of receiving imaging energy (i.e., when the transducer is not pulsed with voltage and therefore does not receive ultrasonic energy), energy was detected within both the imaging band and the noise detection band, thereby obtaining correlated measurements of in-band noise and noise detection band noise, denoted as Ci and Cn. The noise detection band was configured as a frequency band spanning approximately 15 to 25 MHz.
[0366] In this exemplary implementation including an ICE system, 512 waveforms (each sampled at 200 ms / s with a duration of 125 µs) were acquired. Accordingly, 512 pairs of in-band and detect-band noise characterization waveforms were used to estimate noise characteristics. Ci w and Cn w To acquire windowed pairs of in-band and detection-band noise characterization arrays, denoted as , a sliding and overlapping window (window size = 64, 20% overlap) was used for both the imaging band and the noise detection band, where the subscript w is an integer representing the window number. Ci w and Cnw For each pair, the power in the imaging band and the power in the noise detection band were calculated. For further noise characterization, only windows where the in-band power was greater than the threshold Ti were selected.
[0367] Statistical noise power measurements were calculated during noise characterization. A set of windows of the in-band noise characterization array was selected in which the power falls between the 96th and 99th percentiles of the in-band noise characterization array power values. Within this set, the window with the near minimum noise detection band power (the 20th percentile within the set) was selected as w_max, and the power values P(Cn w_max ) and P(Ci w_max ) is calculated, and each, Pn max and Pi max It was denoted as . Similarly, a set of windows of the in-band noise characterization array was selected in which the in-band power falls between the first and fifth percentiles of the in-band noise characterization array power values. Within this set, a window having nearly the minimum noise detection band power (the 20th percentile within the set) was selected as w_min, and the power values P(Cn w_min ) and P(Ci w_min ) is calculated, and each, Pn min and Pi min It was written as.
[0368] As shown in Fig. 6b, the points (Pn max , Pi max ) and (Pn min , Pi min The slope (m) and y-intercept (c) of the line passing through ). The relaxation parameter β was set to 1. The function f was defined as follows:
[0369]
[0370] Here, Pn wis the power in the detection band. This implementation is provided to exemplify an exemplary algorithm for selecting a functional relationship between relative powers in imaging and detection bands, and it will be understood that other functional forms may be used alternatively.
[0371] The f() function was used to perform noise reduction on the in-band imaging waveforms acquired during imaging (i.e., when the pulser periodically emits imaging energy and the ultrasonic energy is received by the transducer receiving circuit).
[0372] and Waveforms in the imaging band and the noise detection band were sampled to acquire pairs of in-band imaging arrays and detection band imaging arrays, denoted as. is an index that identifies the acquisition period corresponding to the scan line. Imaging arrays, and Temporarily segmented according to windows, denoted as, where the subscript w is an integer representing the window number. For each windowed portion of the in-band imaging array and the out-band imaging array, the powers and This was calculated, and each one, and It was written as.
[0373] For each window, out-of-band imaging power This in-band imaging window Noise energy within It was used to estimate. Is It was calculated as, where β was set to 1 and the parameters m, c, Pn max and Pn min It was obtained in the noise characterization stage. Is Clamped to less than. The damping factor is It was calculated as. The damping factor was clamped to greater than 0.02. The damping factor is an array Noise-reduced array multiplied by Acquired. For all 512 scan lines The image was enveloped and processed by an image generator (step (230)).
[0374] The aforementioned algorithm was used to reduce noise in in-band imaging waveforms when performing imaging in the presence of first and second noise sources. FIGS. 11a and 11b illustrate images acquired in the presence of a first noise source, (A) without the application of the noise reduction algorithm and (B) with the application of the noise reduction algorithm. FIGS. 12a and 12b illustrate images acquired in the presence of a second noise source, (A) without the application of the noise reduction algorithm and (B) with the application of the noise reduction algorithm. In both cases, a clear improvement in the signal-to-noise ratio (approximately 6 dB improvement) was observed. For further examples, FIGS. 11c, 11d, and 11e illustrate images obtained in the presence of a first noise source, having the application of the noise reduction method when the relaxation parameter β of the function f() is set to 0.5(C), 1(D), and 1.5(E).
[0375] Example 2: Noise reduction of a periodic noise source with delay correction (Example 6)
[0376] In this example, noise reduction of ultrasonic waveforms detected by the ultrasonic transducer of the ICE console was performed according to the implementation of the method illustrated in FIGS. 8a and 8b. Data for this example was collected in an ICE console where an electromagnetic tracker is present. The control unit of the electromagnetic tracker was observed to generate a pseudo-periodic noise pattern in the ultrasonic image data, as shown in FIG. 13a.
[0377] In this exemplary implementation involving an ICE system, when data is collected in raw / RF mode while the imaging transducer is not receiving imaging energy, one waveform (sampled at 200 MS / s and 125 µs in length) is acquired.
[0378] The data is stored as a pair of arrays (i.e., sampled waveforms), the first array consists of in-band noise characteristic waveforms (7 to 13 MHz) sampled after envelope detection, and the corresponding array consists of out-band noise characteristic waveforms in the 15 to 25 MHz band sampled after envelope detection. These in-band and detection band noise characteristic arrays are denoted as Ci and Cn, respectively.
[0379] Correlated in-band noise and out-of-band noise were subsequently used to perform noise reduction of the in-band imaging waveform during the acquisition of imaging data, that is, while the ultrasonic transducer was receiving ultrasonic energy.
[0380] During imaging, a pair of arrays for several corresponding acquisition periods were acquired, and each period was a scan line Corresponds to. For each array pair, one array was acquired in the imaging band (7 to 13 MHz) and one array was acquired in the noise detection band (15 to 25 MHz). Each array was sampled at 200 ms / s and had a length of 125 µs. Referred to as in-band imaging and out-of-band imaging arrays, respectively, and The arrays labeled as were measured during imaging, that is, while the transducer was receiving imaging energy.
[0381] Each in-band imaging array was processed to reduce noise using an in-band noise characterization array, thereby reducing noise from the in-band imaging array.
[0382] In-band and out-band imaging arrays and This was segmented into multiple time windows. In this exemplary implementation, the window size was 800 samples (in this case, 4 at a rate of 200 MS / s). Im), there was a 75% overlap between adjacent windows. Imaging arrays, temporally segmented according to the windows, and It is denoted as, where the subscript w is an integer representing the window number.
[0383] Temporal alignment between and Cn was achieved on a window-by-window basis. Temporal alignment was achieved between the out-of-band noise characterization array (Cn) and the out-of-band imaging array Calculate the cross-correlation between them and the relative time delay corresponding to the maximum cross-correlation. This was achieved by selecting. Due to the correlation between in-band and out-of-band noise, this time delay It was also applied to align the in-band noise characterization array with respect to the windowed in-band imaging array. The aligned in-band noise characterization array is then windowed ( Ci w (indicated as ).
[0384] In-band noise characterization array ( Ci w ) is an in-band imaging array In-band noise characterization array ( Ci w A scaling factor was applied to ). Noise-reduced in-band imaging array Is It was calculated according to, where was a scaling factor to account for windowing. Negative values were replaced with 0.
[0385] This process was repeated for each additional window and for each scan line.
[0386] The aforementioned method functions as a noise source, Aurora TM It was used to reduce noise in the acquired data when performing imaging in the presence of an electromagnetic tracking system (Northern Digital Inc). FIGS. 13a and 13b illustrate images acquired in the presence of a noise source, (A) without the application of the noise reduction method and (B) with the application of the noise reduction method. A clear improvement in the signal-to-noise ratio (approximately 5 dB) was observed when the noise reduction method was implemented.
[0387] The specific embodiments described above are illustrated by way of example and should be understood that such embodiments may be possible in various modifications and alternative forms. It should be further understood that the claims are not intended to be limited to the specific forms disclosed, but rather are intended to cover all modifications, equivalents, and alternatives that fall within the spirit and scope of the present disclosure.
Claims
Claim 1 A method for denoising imaging signals detected in the presence of broadband noise, comprising: detecting energy waves with an imaging transducer receiving circuit when imaging energy is not received, thereby obtaining a noise characteristic waveform, and processing the noise characteristic waveform to generate an in-band noise characteristic waveform existing within an imaging band and an out-of-band noise characteristic waveform existing within a noise detection band that is at least partially outside the imaging band; processing the in-band noise characteristic waveform and the out-of-band noise characteristic waveform to determine the relationship between the noise in the imaging band and the noise in the noise detection band; detecting imaging signals with the imaging transducer receiving circuit to obtain one or more imaging waveforms; and for at least one imaging waveform, processing the imaging waveform to generate an in-band imaging waveform existing within the imaging band and an out-of-band noise detection imaging waveform existing within the noise detection band. A method for denoising imaging signals, comprising the step of applying a denoising correction to at least a portion of an in-band imaging waveform using the above relationship and the above out-band noise detection imaging waveform. Claim 2 A method for denoising imaging signals according to claim 1, wherein the denoising correction is determined and applied on a time-window basis for at least two time windows of the in-band imaging waveform. Claim 3 A method for denoising imaging signals according to claim 1, wherein the step of detecting the imaging signals includes the step of detecting the imaging signals along a plurality of scan lines to an imaging transducer receiving circuit, and the method further includes the step of acquiring a plurality of imaging waveforms associated with each of the plurality of scan lines; and the step of generating an image based on denoised in-band imaging waveforms associated with each of the plurality of scan lines. Claim 4 A method for denoising imaging signals according to claim 1, wherein the denoising correction is applied within a given time window when a scale associated with the amount of noise in an in-band imaging waveform within a given time window is greater than a threshold value. Claim 5 A method for denoising imaging signals, wherein, in paragraph 4, the scale associated with the amount of noise in the in-band imaging waveform is based on the relationship and the amount of power of the in-band noise characteristic waveform. Claim 6 A method for denoising imaging signals according to claim 4, wherein, before applying the denoising correction to the given time window, the given time window having a scale exceeding the threshold is re-evaluated in relation to spatially adjacent time windows corresponding to spatially adjacent scan lines, so that samples initially estimated to be associated with the presence of noise are replaced with samples from the spatially adjacent time windows when the adjacent time windows are estimated to be associated with the absence of noise. Claim 7 A method for denoising imaging signals according to claim 4, wherein, before applying the denoising correction to the given time window, the given time window having a scale exceeding the threshold is re-evaluated in relation to temporally adjacent time windows, so that a time window initially estimated to be associated with the absence of noise is re-evaluated as associated with the presence of noise when the adjacent time windows are estimated to be associated with the presence of noise. Claim 8 A method for denoising imaging signals, wherein, in paragraph 4, the threshold value is determined based on baseline measurements taken in the absence of the imaging energy and noise. Claim 9 A method for denoising imaging signals according to claim 1, wherein the relationship between the noise in the imaging band and the noise in the noise detection band is related to the scale of relative energy in the in-band noise characteristic waveform and the out-band noise characteristic waveform. Claim 10 A method for denoising imaging signals according to claim 1, wherein the denoising correction includes a subtrahend value. Claim 11 A method for denoising imaging signals according to claim 10, wherein the denoising correction is determined and applied per time window, and the reduction value associated with a given time window increases according to the amount of energy detected in the out-of-band noise detection imaging waveform for the given time window. Claim 12 A method for denoising imaging signals according to claim 1, wherein the denoising correction includes multiplication with an attenuation factor. Claim 13 A method for denoising imaging signals according to claim 12, wherein the denoising correction is determined and applied per time window, and the attenuation factor associated with a given time window is reduced according to the amount of noise in the in-band imaging waveform for the given time window, thereby attenuating the portions of the in-band imaging waveform associated with the noise. Claim 14 A method for denoising imaging signals according to claim 1, wherein the denoising correction is determined and applied per time window, and the denoising correction applied for a given time window depends on the depth of the given time window within the in-band imaging waveform. Claim 15 A method for denoising imaging signals according to claim 1, further comprising the step of repeating noise characterization to re-establish the relationship between the noise in the imaging band and the noise in the noise detection band. Claim 16 A method for denoising imaging signals, wherein, in paragraph 15, the noise characterization is repeated according to user input. Claim 17 A method for denoising imaging signals, wherein, in paragraph 15, the noise characterization is automatically repeated when the absence of imaging energy is detected. Claim 18 A method for denoising imaging signals according to claim 1, further comprising the step of monitoring one or more parameters associated with the out-of-band noise detection imaging waveform to detect changes in noise within the noise detection band. Claim 19 A method for denoising imaging signals according to claim 18, further comprising the step of generating an alarm or message when a change in at least one parameter is detected. Claim 20 A method for denoising imaging signals according to claim 1, further comprising: a step of performing noise suppression on the in-band imaging waveform during the absence of imaging energy; a step of generating an error value based on the energy in the in-band imaging waveform after noise correction; and a step of generating an alarm for repeating noise characterization when the error value exceeds a preset threshold. Claim 21 A method for denoising imaging signals according to claim 1, wherein the imaging transducer receiving circuit includes an ultrasonic transducer. Claim 22 A method for denoising imaging signals according to claim 1, wherein the imaging transducer receiving circuit includes a coil for detecting a magnetic field.
Citation Information
Patent Citations
Ultrasonic diagnostic apparatus
JP2009261441A
