Respiration compensation for mapping

Abstract

To provide systems, apparatuses and methods for compensating for respiration in mapping.SOLUTION: A system and a method are disclosed. The system and method include obtaining respiration data of a patient via at least one sensor, obtaining probe location data for a probe positioned within a cavity, generating probe-cavity location data by compensating the respiration data from the probe location data, identifying periods when the probe is stable relative to a cavity boundary on the basis of the probe-cavity location data, and capturing data on the basis of generated probe-cavity location during the identified periods to produce mapping.SELECTED DRAWING: Figure 1

Description

Technical Field

[0001] This application provides a system, apparatus, and method for compensating for respiration during mapping.

Background Art

[0002] Surgical procedures involving a cavity or cardiac chamber, such as the heart, require accurate information regarding the position of a probe or catheter within the cavity. In one aspect, mapping of the cardiac chamber is important for identifying problems associated with cardiac arrhythmias (e.g., atrial fibrillation (AF)) that can be treated via an in-body procedure.

[0003] Respiration can be divided into the phases of inhaling (inspiration or inhalation) and exhaling (expiration or exhalation). Due to human physiology, the diaphragm moves up and down during this process, periodically displacing the heart. During the expiratory phase, there is a moment called the end-expiratory phase, where this displacement from the diaphragm is minimized over a period of time.

Summary of the Invention

Means for Solving the Problems

[0004] A system and method are disclosed. The system and method include obtaining respiratory data of a patient via at least one sensor, obtaining probe location data for a probe positioned within a cavity, generating probe-cavity location data by compensating the respiratory data from the probe location data, identifying a period during which the probe is stable relative to the cavity boundary based on the probe-cavity location data, and capturing data based on the probe-cavity locations generated during the identified period to create a mapping.

Brief Description of the Drawings

[0005] A more detailed understanding will be possible from the following description given by way of example in conjunction with the accompanying drawings.

Figure 1

Figure 2

Figure 3

Figure 4A

Figure 4B

Figure 5

Figure 6

Figure 7A

Figure 7B

Figure 8

Figure 9

Figure 10

Figure 11

Figure 12

Figure 13

Figure 14

Figure 15

Figure 16

Best Mode for Carrying Out the Invention

[0006] As disclosed herein, a system, apparatus, and method are provided for determining or estimating the movement of a probe while it is inside a heart chamber by considering respiratory movement. The term "probe" is used interchangeably herein with the term "catheter," and one of ordinary skill in the art will understand that any type of location-sensing device can be implemented to have the configuration disclosed herein.

[0007] A system and method are disclosed. The system and method include obtaining respiratory data of a patient via at least one sensor, obtaining probe location data for a probe positioned within a cavity, generating probe-cavity location data by compensating the respiratory data from the probe location data, identifying a period during which the probe is stable with respect to the cavity boundary based on the probe-cavity location data, and capturing data based on the probe-cavity positions generated during the identified period to produce a mapping.

[0008] This method includes obtaining a patient's respiratory data via at least one sensor, obtaining probe location data for a probe positioned within a cavity, generating compensated location data by compensating the obtained probe location data with the obtained respiratory data, creating a mapping based on the generated compensated location data, and may include applying a filter to the respiratory data. Obtaining the respiratory data may include obtaining a respiratory indicator (abbreviated herein as "RI") via at least one sensor. Obtaining the respiratory data may include transforming the respiratory indicator by using singular value decomposition (SVD) to obtain a first eigenvector and a first eigenvector derivative, where the first eigenvector corresponds to a primary respiratory signal and the first eigenvector derivative corresponds to a phase-shifted respiratory signal. The primary respiratory signal and the phase-shifted respiratory signal transition from a two-dimensional to a three-dimensional ellipsoid based on a correlation matrix, which is determined based on a weighting factor including at least one of the age of the respiratory data, the speed of the probe, and the depth of respiration. The method may include generating an image including an estimated respiratory motion based on the respiratory data, a probe motion based on the probe location data, and a probe-cavity motion based on the probe-cavity location data. The method may also include notifying a surgeon of a period during which the probe is stable with respect to the cavity boundary. Notifying the surgeon includes notifying the surgeon of the period during which the probe is stable with respect to the cardiac cavity boundary via a visual indicator displayed on a monitor. The method may include identifying a site stability site based on a period during which the probe speed is less than 2 mm / second over at least 3 seconds.

[0009] The system may include a probe configured to be inserted into a body cavity of a patient, at least one sensor configured to acquire respiratory data, wherein the probe and the at least one sensor are configured to acquire probe location data, a processor. The processor is configured to generate compensated location data by compensating the acquired probe location data with the acquired respiratory data, and generate a mapping based on the generated compensated location data. The processor may be configured to apply a filter to the respiratory data and the probe location data. The processor may be configured to generate a respiratory indicator based on the respiratory data from the at least one sensor. The processor may be configured to transform the respiratory indicator by using singular value decomposition (SVD) to obtain a first eigenvector and a first eigenvector derivative, wherein the first eigenvector corresponds to a primary respiratory signal and the first eigenvector derivative corresponds to a phase-shifted respiratory signal. The primary respiratory signal and the phase-shifted respiratory signal transition from a two-dimensional to a three-dimensional ellipsoid based on a correlation matrix. The correlation matrix is determined based on a weighting factor including at least one of the age of the respiratory data, the speed of the probe, and the depth of respiration. The processor may be configured to generate an image including an estimated respiratory motion based on the respiratory data, a probe motion based on the probe location data, and a probe-cavity motion based on the probe-cavity location data. The processor may be configured to notify a surgeon of a period during which the probe is stable with respect to the cavity boundary. Notifying the surgeon of a period during which the probe is stable with respect to the cardiac cavity boundary includes displaying a visual indicator on a monitor. The processor may be configured to identify a site stability site based on a period during which the probe speed is less than 2 mm / second over at least 3 seconds.

[0010] Refer to FIG. 1, which shows an exemplary system 10 (e.g., a medical device and / or a catheter-based electrophysiology mapping and ablation system) in which one or more features of the subject matter of this specification can be implemented in accordance with one or more embodiments. The whole or part of system 100 can be used to collect the information described in this specification (e.g., biological measurement data and / or training data sets) and / or implement machine learning and / or artificial intelligence algorithms. As shown, system 10 includes a recorder 11, a heart 12, a catheter 14, a model or anatomical map 20, an electrogram 21, a spline 22, a patient 23, a physician 24 (representing any medical professional, technician, or clinician), a location pad 25, one or more electrodes 26, a display device 27, a distal tip 28, a sensor 29, a coil 32, a patient interface unit (PIU) 30, an electrode skin patch 38, an ablation energy generator 50, and a workstation 55. It should be further noted that each element and / or item of system 10 represents one or more of that element and / or that item. The example of system 10 shown in FIG. 1 can be modified to implement the embodiments disclosed in this specification. The embodiments of the present disclosure can also be similarly applied using other system components and settings. Further, system 10 may include additional components such as elements for sensing electrical activity, wired or wireless connectors, a processing device, and a display device.

[0011] System 10 includes a plurality of catheters 14 that are percutaneously inserted by a physician 24 into a cardiac chamber or vascular structure of a patient's vasculature. Typically, a delivery sheath catheter is inserted into the left atrium or right atrium near a desired location within the heart 12. Thereafter, the plurality of catheters can be inserted into the delivery sheath catheter to reach the desired location. The plurality of catheters 14 can include catheters dedicated to sensing intracardiac electrogram (IEGM) signals, ablation-only catheters, and / or catheters dedicated to both sensing and ablation. Exemplary catheters 14 configured to sense IEGM are shown herein. The physician 24 contacts the distal tip 28 of the catheter 14 with the heart wall to sense a target site of the heart 12. For ablation, the physician 24 similarly moves the distal end of the ablation catheter to the target site for ablation.

[0012] Catheter 14 is an exemplary catheter that optionally distributes over a plurality of splines 22 at the distal tip 28 and includes at least one, preferably a plurality of electrodes 26 configured to sense IEGM signals. Catheter 14 may additionally include a sensor 29 embedded within or near the distal tip 28 to track the position and orientation of the distal tip 28. Optionally and preferably, the position sensor 29 is a magnetic-based position sensor that includes three magnetic coils for sensing three-dimensional (3D) position and orientation.

[0013] Sensor 29 (e.g., a position-based or magnetic-based position sensor) may operate with a location pad 25 that includes a plurality of magnetic coils 32 configured to generate a magnetic field within a predetermined working volume. The real-time position of the distal tip 28 of the catheter 14 may be tracked based on the magnetic field generated by the location pad 25 and sensed by the sensor 29. Details of magnetic-based position sensing techniques are described in U.S. Patent Nos. 5,539,199, 5,443,489, 5,558,091, 6,172,499, 6,239,724, 6,332,089, 6,484,118, 6,618,612, 6,690,963, 6,788,967, and 6,892,091.

[0014] System 10 includes one or more electrode patches 38 positioned on the patient 23 for skin contact to establish location referencing of the location pad 25 and impedance-based tracking of the electrodes 26. For impedance-based tracking, a current is directed to the electrodes 26 and sensed at the patches 38 (e.g., electrode-skin patches), whereby the location of each electrode can be triangulated via the patches 38. Details of impedance-based position tracking techniques are described in U.S. Patent Nos. 7,536,218, 7,756,576, 7,848,787, 7,869,865, and 8,456,182, which are hereby incorporated by reference.

[0015] Recorder 11 displays an electrogram 21 captured by the electrodes 18 (e.g., body surface electrocardiogram (ECG) electrodes) and an intracardiac electrogram (IEGM) captured by the electrodes 26 of the catheter 14. Recorder 11 may include pacing capabilities for pacing the heart rhythm and / or may be electrically connected to an independent pacer.

[0016] System 10 may include an ablation energy generator 50 adapted to deliver ablation energy to one or more electrodes 26 at the distal tip 28 of a catheter 14 configured to ablate. The energy produced by ablation energy generator 50 may include radiofrequency (RF) energy, pulsed-field ablation (PFA) energy, or a combination thereof, including monopolar or bipolar high voltage DC pulses such that they can be used to effect irreversible electroporation (IRE), but are not limited thereto.

[0017] PIU30 is an interface configured to establish electrical communication between a catheter, electrophysiology equipment, a power source, and a workstation 55 that controls the operation of system 10. The electrophysiology equipment of system 10 may include, for example, a plurality of catheters 14, location pads 25, body surface ECG electrodes 18, electrode patches 38, ablation energy generator 50, and recorder 11. Optionally and preferably, PIU30 additionally includes processing capabilities to implement real-time calculation of catheter location and perform ECG calculations.

[0018] The workstation 55 includes a memory, a processor unit having a memory or storage device loaded with appropriate operating software, and a user interface function. The workstation 55 optionally provides a plurality of functions including: (1) modeling the endocardial anatomical structure in three dimensions (3D) and rendering it to display a model or anatomical map 20 on a display device 27; (2) displaying on the display device 27 the activation sequence (or other data) compiled from the recorded electrogram 21 as a representative visual indicator or image superimposed on the rendered anatomical map 20; (3) displaying the real-time locations and orientations of a plurality of catheters within the heart chamber; and (5) displaying on the display device 27 a site of interest such as a location where ablation energy has been applied. One commercially available product embodying the elements of the system (10) is available as the CARTO (trademark) 3 system, commercially available from Biosense Webster, Inc., 31A Technology Drive, Irvine, CA, 92618.

[0019] For example, system 10 can be part of a surgical system (e.g., the CARTO® system sold by Biosense Webster) configured to acquire biometric data (e.g., anatomical and electrical measurements of a patient's organ such as heart 12 as described herein) and perform a cardiac ablation procedure. More specifically, in the treatment of cardiac conditions such as arrhythmias, it is often required to obtain a detailed mapping of the cardiac tissue, chambers, veins, arteries, and / or electrical pathways. For example, as a prerequisite for successfully performing catheter ablation (as described herein), the cause of the arrhythmia may need to be accurately located within the chambers of heart 12. Such localization is performed by an electrophysiological examination, during which spatially resolved potentials can be detected by a mapping catheter (e.g., catheter 14) introduced into the chambers of heart 12. This electrophysiological examination, so-called electroanatomical mapping, provides 3D mapping data that can be displayed on display device 27. In many cases, the mapping function and the treatment function (e.g., ablation) are provided by a single catheter or a group of catheters, and the mapping catheter also operates as a treatment (e.g., ablation) catheter simultaneously.

[0020] In another aspect, at least one uniaxial magnetic sensor mounted on the catheter tip is configured to operate in combination with at least one external magnetic sensor of a patient pad (i.e., under the patient). In one embodiment, there are three magnetic sensors mounted on the catheter oriented in three different directions (i.e., 120 degrees apart) and configured to operate in combination with three external magnetic sensors. Details of such techniques are provided in the following documents, which are incorporated herein by reference as if fully set forth herein: U.S. Patent No. 5,391,199, U.S. Patent No. 5,443,489, U.S. Patent No. 5,558,091, U.S. Patent No. 6,172,499, U.S. Patent No. 6,177,792, U.S. Patent No. 6,690,963, U.S. Patent No. 6,788,967, and U.S. Patent No. 6,892,091.

[0021] In another aspect, an impedance sensor on the catheter is provided that is configured to be used without an external sensor. Details of such techniques are provided in the following documents, which are incorporated herein by reference as if fully set forth herein: U.S. Patent No. 5,944,022, U.S. Patent No. 5,983,126, and U.S. Patent No. 6,456,864. As used herein, the term sensor refers to any of the sensor configurations described herein, whether the sensor is integrated with the catheter or includes a combination of a sensor integrated with the catheter and additional sensors such as external sensors.

[0022] Regardless of the sensor configuration, these sensors assist in modeling the patient's respiratory cycle and identifying when the patient's lungs are inhaling or exhaling. In one embodiment, impedance increases as the lungs expand or are filled with air. The sensors can also generate a magnetic field that can be used to detect the absolute position of catheter 14. Those skilled in the art will understand from this disclosure that various methods and sensors can be used to determine the patient's respiratory cycle or the position of the catheter. Using the respiratory motion collected from the sensors, an ellipsoid (i.e., element 120 of FIG. 2) can be generated that provides a model of the respiratory cycle.

[0023] FIG. 2 shows an exemplary image 110 that includes data regarding respiratory motion 120, catheter motion 130, reconstructed motion 140, and intracardiac motion 150. This information is configured to be displayed on monitor 3. Each of these components will be described in more detail herein.

[0024] The term respiratory motion 120 refers to motion based on respiratory data, which can be collected or gathered in various ways. For example, in one embodiment, a probe, sensor, or patch (such as sensor 18 of FIG. 1) is attached to the patient's body or the sensor is integrated with catheter 14 that can be used in conjunction with an external sensor 18. This data reflects changes in the volume of the patient's lungs. Respiratory motion 120 can be based on the movement of the patient's diaphragm during respiration. Signals can be extracted by placing electrodes on the body surface. In one embodiment, respiratory motion 120 is based on an electrical signal between an electrode on the probe and a patch on the body surface. In another aspect, the probe can measure the displacement of one part of the body relative to another part of the body. This information is collected and further processed using the processes disclosed herein. In one aspect, a primary or main signal is generated by a sensor (such as sensor 18 of FIG. 1). This primary or main signal is then further processed to determine the derivative of the signal. Respiratory indicators and respiratory vectors related to respiratory motion 120 are described herein.

[0025] The catheter movement 130, also known as probe or sensor movement or original catheter movement, shown in FIG. 2, reflects data regarding the absolute position of the catheter 14. This data can be generated using any known tracking or sensing configuration for the catheter, probe, sensor, etc. As a result, in one aspect, a magnetic field and sensors 29 on the catheter 14 are used to detect the position of the catheter 14. For example, catheter movement data can be obtained using a magnetic transmitter external to the patient that generates a signal based on the position of the catheter 14. In one aspect, the sensor configuration includes at least three magnetic sensors oriented in three different directions relative to each other, mounted at the distal end of the catheter and acting in conjunction with a plurality of sensors external to the patient. In another aspect, an impedance sensor is provided and no external sensors are required. The catheter movement data can be further processed through any combination of filtering, smoothing, or post-acquisition signal processing.

[0026] The reconstructed movement 140 of FIG. 2 is, in one aspect, a control element. In other words, the reconstructed movement 140 is generated to provide a validity and error check as to whether there is an undesired amount of difference between itself and the original catheter movement 130. In one aspect, the reconstructed movement 140 is equal to the original catheter movement 30 added to the respiratory movement 120. An error variable, also referred to as ERR and described in more detail herein, can be determined based on the reconstructed movement 140 from the original catheter movement 130. Such determination based on the reconstructed movement 140 can be by subtraction or addition to the original catheter movement 130.

[0027] Intracardiac motion 150 is also referred to as catheter-heart motion or probe-cavity motion. In one aspect, intracardiac motion 150 is measured in millimeters, although any metric can be used for measurement. This motion data can generally refer to any type of motion data of any wall of the cardiac cavity or a probe, catheter, or sensor with respect to the cardiac cavity. This data can ultimately be used to determine site stability, particularly with respect to ablation in a patient's heart. Intracardiac motion 150 is typically important for a surgeon as this information indicates when the catheter is stable with respect to the cardiac cavity wall. With intracardiac motion 150, it is possible to find a stable point, also known as a stability location or site. This information is important for determining various parameters regarding ablation lesions such as location, size, etc. During respiration, it is generally assumed that the catheter and the heart move up and down based on diaphragmatic contraction. If the catheter is not stable with respect to the cardiac wall or boundary during ablation (i.e., the catheter is not in contact with the cardiac wall or boundary), the ablation is likely to be unsuccessful or not ablate the desired target location. In one aspect, the present disclosure relates to providing a more reliable and accurate mapping of intracardiac motion 150 (i.e., catheter-heart motion or probe-cavity motion) based at least in part on respiratory motion and other characteristics, which addresses issues related to any sudden movements that occur between end-expirations. In other words, by providing more accurate information regarding the relative position of the catheter and the cavity boundary, any movement that occurs during end-expiration can be considered by the surgeon and used to facilitate decisions regarding the ablation procedure.

[0028] In one aspect, the present system, method, and process are based on the following three features: (1) signal processing, (2) respiratory compensation, and (3) local discovery or site stability location. One of ordinary skill in the art will understand that any one or more of these features can be used independently of the other features. For example, features (1) and (2) can be implemented without feature (3). In another aspect, feature (3) is specifically implemented for procedures involving the heart, specifically procedures involving ablation of a specific region of the heart. None of the features described herein are specifically limited to being used or adapted to determine the aspect of the cardiac chamber and being specifically implemented only with respect to the cardiac chamber (i.e., navigation in the lung).

[0029] In one aspect, signal processing includes converting a respiration indicator (RI) into a primary respiratory signal and its derivative, i.e., a respiration vector (RV). Respiratory data is collected based on sensors in one embodiment, and the sensors can include a catheter integrated with the sensor, an external sensor, or any combination thereof. In one aspect, before converting the RI to the RV, ablation effects can be considered by ablation offset correction, and low-pass filtering can be used to reduce cardiac motion. In one aspect, the disclosed subject matter is based on decomposing catheter motion into two components, i.e., respiratory motion and catheter-heart motion. Respiratory motion is generally assumed to be correlated with the RV via a correlation matrix A.

[0030] Catheter-heart motion is assumed to be smoothed motion and can be modeled via a spline curve. Based on the above, Equation 1 is provided. Measured motion = Respiratory motion (A × RV)+Catheter-heart motion + ERR (Equation 1)

[0031] Here, the correlation matrix A and the spline parameters are adjusted incrementally and simultaneously to minimize the root mean square (RMS) deviation between the measured motion after low-pass filtering and the reconstructed motion in order to minimize the error.

[0032] In one aspect, the correlation matrix A is regressed at step n to match the respiratory motion at step n - 1, which is the subtraction of the intracardiac motion from the total motion at step n - 1. In other words, A is regressed in the current state to match the respiratory motion of the previous state, and the respiratory motion is equal to the subtraction of the intracardiac motion in the previous state from the total motion.

[0033] The correlation matrix A is a 2-by-3 data set. In one aspect, the correlation matrix A is configured to transform a two-dimensional RV into an ellipsoid in three-dimensional space. The matrix values are adjusted or selected to scale the RV values to the respiratory ellipsoid size (i.e., in mm) by multiplication. In one aspect, the RV values are relatively small (i.e., 1.e -3 ), and they are multiplied by large values (i.e., 3,000) within the A matrix to produce a movement in three-dimensional space (i.e., a 3.0 mm movement).

[0034] Regarding the respiratory motion, in one aspect, there are five RI channels. Those skilled in the art will understand that the number of RI channels can vary. RI generally indicates changes in lung motion and lung volume over time. RI generally contains similar signals and noise respectively. In one aspect, RI is collected by directly attaching a plurality of probes to the patient. For example, six patches or probes can be attached to the patient. The electrical signals between the six patches or probes are measured and RI is generated. The electrical signals may include voltage, resistance, and / or impedance. The RI corresponding to each patch or probe can be provided in mV in one embodiment.

[0035] Specifically, in one embodiment, RI is a measure of the change in resistance (impedance) between patches, and the resistance is used via time decomposition to extract the frequency band related to respiration. RI reflects the change in the amount of air present flowing into the lungs. The airflow into the lungs and the fluctuations of the air in the lungs can change the measured value of the resistance between patches (between six patches).

[0036] To remove noise or other elements, a filter can be applied to RI. For example, a low-pass finite impulse response (FIR) filter can be used to remove the imprint of the heartbeat on the RI signal. In one aspect, this step uses a 151-weight vector that is multiplied by 151 elements on each channel of the unprocessed data to produce one sample of the filtered data. In other words, the filter multiplies the corresponding samples in a sliding window manner. Those skilled in the art will understand that these parameters can vary.

[0037] Using singular value decomposition (SVD), weights can be found that select the most prominent respiration signal when multiplied by the scale 5RI. This filtering step and processing step provide a smoother output signal and remove unwanted noise and clutter caused by the heartbeat.

[0038] Based on RI, the main respiration vector is extracted. In one aspect, this extraction is performed by calculating the first eigenvector in the SVD.

[0039] Generally, SVD is configured to extract the most prominent common motion from several similar respiration-dependent signals. The common signal is found as a linear combination of the individual RI signals. In one aspect, to verify the consistency of the linear weights, SVD is performed three times, including a first SVD for the first 30 seconds of the RI buffer, a second SVD for the second 30 seconds, and a third SVD for the entire 60 seconds. If the three sets of weights are sufficiently similar, the last set is selected. In other words, the similarity here refers to V from the SVD. Thus, this V becomes the weight vector, i.e., the RI2RV weight vector. In one aspect, this similarity parameter has a Pearson correlation of at least 0.5.

[0040] In one aspect, this process includes determining the first eigenvector in the SVD of five RI signals over several respiratory cycles. For example, a 60-second buffer can be used, which can correspond to 10 respiratory cycles. By using multiple cycles, it can be ensured that the RI signals are accurate and reliable.

[0041] The disclosed subject matter also adds a phase-shifted signal to provide an elliptical-like motion or model as shown by the estimated respiratory motion 120 of FIG. 2. This phase-shifted signal is calculated by finding the derivative of the RI signal and combining the derivatives using the same weights. This process is similar to the phase shift between the Cos(x) function and the Sin(x) function.

[0042] In one aspect, the respiratory vector (RV) is generated based on the primary respiratory vector and the derivative of the primary respiratory vector. In other words, the five RIs are processed using the steps disclosed herein to generate or determine two RVs. Once these RVs are determined, the following equation is given. Measured motion = Respiratory motion (A × RV) + Catheter - Heart motion + ERR (Equation 2)

[0043] Equation 2 is generally similar to Equation 1, except that Equation 2 uses RV instead of RI.

[0044] To determine the catheter location, the unprocessed location data regarding the catheter is filtered. In one aspect, the unprocessed location data is filtered using the same low-pass FIR filter that was applied to the RI signal to remove cardiac motion. As a result of the processing, a synchronized and filtered respiration vector and catheter location are generated.

[0045] In summary, at the end of the signal processing, the process disclosed herein provides Signals 120 and 130 of FIG. 2.

[0046] This feature generally involves estimating the correlation matrix A and spline parameters that represent the actual motion of the catheter. Next, two RVs are transformed into three-dimensional space.

[0047] Regarding the RV, the process disclosed herein generally transforms this 2D vector into a 3D ellipse or ellipsoid in 3D space, as shown by Element 120 of FIG. 2. The ellipsoid depicted in FIG. 2 can be oriented in any direction, and the main purpose of compensating for respiration involves modeling the patient's respiratory movement on the catheter or probe.

[0048] The correlation matrix A (also referred to as the transformation matrix) transforms the two-dimensional vector into a three-dimensional ellipsoid. The correlation matrix A is determined based on regression in one aspect. The root mean square error (RMSE) is minimized between the modeled motion (i.e., respiratory motion + catheter motion) and the measured motion to provide the correlation matrix A.

[0049] As shown in FIG. 3, a system 200 including a site stability module 210 is disclosed. This system 200 can be implemented, integrated, or configured to interface with the computing system 4 shown in FIG. 1. As used herein, the term module can refer to any computing component or interface and can include any hardware component or software component.

[0050] The site stability module 210 can include any one or more of the features disclosed herein. In one aspect, the site stability module 210 is configured to receive data from various inputs. For example, location data 202, respiratory indicator data 204, status data 206 (such as location status and RI status), and initialization or algorithm parameters 208 can each be provided to the site stability module 210. In one embodiment, the data is streamed and provided at 16.67 ms intervals. One of ordinary skill in the art will understand, based on the present disclosure, that the data streaming parameters and cycles or periods can vary.

[0051] From the site stability module 210, a session stability time segment 212 (i.e., a start - end segment) and catheter - heart location data 214 are transmitted to the central module 220. In other words, the site stability module 210 provides output parameters including the last stability segment parameters of the last set of sites in the last ablation session. In one aspect, stability is a segment of duration of 3 seconds or more where the catheter - heart speed is less than a certain speed threshold. A site is part of a stable segment during ablation time, and the stable segment during ablation can be divided into several sites.

[0052] In one embodiment, the central module 220 can include any centralized computing system configured to receive, process, and / or transmit data to and from the site stability module 210. In one example, the central module 220 is a CARTO VISITAG™ processing unit. Those skilled in the art will understand that various types of configurations for communicating with the site stability module 210 can be provided.

[0053] In the central module 220, compensated location data 222 for each session can be stored and remembered. This data 222 can be further processed, filtered, etc. in the site stability module 210. Optionally, a recording module 230 can be provided. The recording module 230 can be configured to assist in debugging and post-analysis of the collected data. Element 224 in FIG. 3 represents a recalculation step, in which information from the central module 220 is continuously fed back to the site stability module 210 to refine and recalculate information regarding site stability.

[0054] Referring to FIGS. 4A and 4B, a flowchart showing various steps of a process 300 for providing site stability is provided. As shown in FIGS. 4A and 4B, the process 300 generally includes two separate stages for site stability, namely a preprocessing stage 310 and a stability detection stage 330. The preprocessing stage 310 is generally the stage in which the raw signal from the measurement is filtered and converted to provide a filtered catheter motion vector and a respiration vector. The filtered catheter motion may still include a respiration factor. The decomposition 350 during the stability detection stage 330 separates the filtered catheter motion into respiration motion and catheter motion without a ventilator. The stability detection stage 330 is the stage in which sliding motion is used to detect ablation sites within the stable segment during ablation. The preprocessing stage 310, the decomposition stage 350, and the stability detection stage 330 are described in more specific detail below.

[0055] In the preprocessing stage 310, location data 312 and RI 314 are provided. In other words, the preprocessing stage 310 includes at least two groups of channels, namely the location channel and the RI channel. This data can be set or generated in relatively small increments. For example, the data can be set at least every 16.67 ms.

[0056] The location data 312 is further processed. For example, the location data 312 can be processed by applying a low-pass filter or a low-pass FIR filter in step 316. After this step, in step 324, the filtered location, i.e., the last filtered location, is determined.

[0057] The RI 314 can also be further processed through a series of steps or processes 318, 320, and 322. As shown in Figure 3A, the RI 314 can first be analyzed for ablation correction in step 318, which takes into account any offsets caused by the ablator. The offset from the ablator affects the ability to estimate respiratory motion. Therefore, by reducing the effect of the offset, an accurate output regarding the RI 314 is provided. In one aspect, the ablation offset correction factor is determined by measuring the RI signal before and during ablation. Whenever the ablator is turned on, the estimated offset for each channel is compensated (e.g., subtracted or added) from the RI signal. This offset is estimated based on the change in the RI signal between the RI signal measured just before the start of ablation and the RI signal measured just after the start of the ablator. The effect of the ablation offset is shown in Figure 5, which shows the ablation offset due to the shift of the RI during ablation.

[0058] Returning to FIG. 4A, the data is filtered at step 320. In one embodiment, the filter includes a low-pass filter and a derivative filter. In one embodiment, the low-pass filter and the derivative filter are configured as a FIR filter with a 151-sample kernel. In one aspect, the preprocessing includes the following steps. First, the original five RI channels are acquired. Second, ablation offset correction is performed. Third, a low-pass filter and a derivative filter are executed using a 151-sample sliding window of the FIR filter to create a filtered respiration indicator (FRI) and a derivative respiration indicator (DRI). Fourth, SDV is executed to find the most prominent weights. Fifth, the most prominent weights are applied to the FRI to obtain RV(1) and to the DRI to obtain RV(2). Those skilled in the art will understand that additional filters can be used and additional steps can be implemented.

[0059] FIG. 6 shows an example of a filter that can be used at step 320 and shows the weights for the low-pass filter and its derivative signal. As shown in FIG. 6, different weight coefficients are applied over time to the low-pass filter and the derivative filter.

[0060] The data regarding RI314 is further processed at step 322 by calculating the first eigenvector of RI314 and the derivative of the first eigenvector of RI314. SVD is used to find the first eigenvector of the filtered RI to reduce redundancy and noise. The mean-compensated (e.g., subtracted or added) filtered respiration indicator (FRI) and derivative respiration indicator (DRI) can be stored. In one embodiment, this information is stored in a one-minute buffer. Those skilled in the art will understand that the buffer duration can vary.

[0061] In one aspect, the SVD is performed only once at the start of a predetermined period when the buffer is filled with valid signals. The SVD produces five coefficients of the first eigenvector (V1), and these five coefficients are used to produce a respiratory vector (RV) in step 326. In one embodiment, the RV is equal to FRI×V1, DRI×V1.

[0062] At the end of the preprocessing stage 310, the last minute of the RV data (produced in step 326) is synchronized with the filtered location data produced in step 324 and stored in the last minute buffer.

[0063] The data from steps 324 and 326 is provided to the stability detection stage 350. During the stability detection stage 350 (shown in the lower part of FIG. 4A and the upper part of FIG. 4B), motion decomposition 352 is performed. Further details regarding the motion decomposition are provided by FIG. 7B. From the motion decomposition 352, the process includes calculating site stability and site 354. This step returns the data to a site stability module 210 configured to receive catheter motion data and provide session motion data. Information regarding the motion of the catheter and site information are stored by the site stability module 210. The site stability parameters can be recalculated based on the information provided to the site stability module 210.

[0064] A more detailed overview of the preprocessing stage 310 is shown in FIGS. 7A and 7B. As shown in process 600, the preprocessing stage begins with obtaining a data set, i.e., starting from step 602. This data set can include RI, location, status, and various other information.

[0065] As shown on the right side of FIG. 7A, an ablation correction stage 604 is provided. This stage is configured to correct the offset of the RI signal due to ablation interference. The ablation correction stage 604 includes setting a buffer in step 604a. Next, in step 604b, the offset is estimated for each channel by comparing the RI signal edges (i.e., the signal 5 and 95 percentiles) immediately before and after the ablator is turned on. Next, in step 604c, this estimated offset is compensated (e.g., subtracted or added) from the RI signal. In one aspect, the ablation offset is measured and corrected for each ablation session and is performed dynamically and continuously. In each ablation, the offset measurements are averaged over time, making the estimation more accurate.

[0066] As shown in FIG. 7A, the buffer in step 604a can include a 1.5-second buffer of the RI signal to enable fast estimation. This buffer can include at least 151 samples of the RI signal. The parameters for considering the effect of ablation on the RI signal can vary depending on specific situations such as the number of RI channel signals, ablation intensity and duration, and patient characteristics. The data from step 604c is ultimately sent to step 616, which is described in more detail below.

[0067] Following step 602, a verification step 608 is performed to determine whether the dataset is complete. In one aspect, a manager within the computing system 4 or software module checks the validity of the data. As used herein, the term manager can refer to any computing interface, such as a CARTO (registered trademark) VISITAG (trademark) processing unit or an interface within the central module 220. This component is configured to insert or input data and also receive the output of the data. If the data is invalid, the manager sends a reset or soft reset signal to the algorithm processor to empty all of the buffers and begin refilling the buffers with new data if an error is detected. The algorithm ensures that the buffers are filled with valid data before proceeding to subsequent steps.

[0068] If it is determined that the data is not valid due to a location data problem, the buffer is reset in step 610 and the location filter buffer is updated in step 614. In one embodiment, step 614 includes 151 x 3 samples.

[0069] If it is determined that the data is not valid due to an RI channel related problem, the buffer is reset in step 612 and the RI filter buffer is updated in step 616. In one embodiment, step 616 includes 151 x 5 samples.

[0070] If it is determined that the data is valid, the location filter buffer and the RI filter buffer are also updated as described with respect to steps 614 and 616, but there is no need to reset the buffer.

[0071] Following steps 614 and 616, both the location data and the RI signal data undergo a filtering step. The location data from step 614 is filtered at step 618. In one embodiment, the filter at step 618 is a low-pass FIR filter. The RI signal data from step 616 is also processed. In one embodiment, the filter at step 612 is a low-pass and differentiating FIR filter.

[0072] Step 622 following step 618 updates the location buffer. Step 624 follows step 620 and includes updating the RI buffer. From step 622, the location data is further processed at step 626, which includes obtaining the last minute buffer and obtaining the filtered catheter movement.

[0073] Regarding the RI signal data, after step 624, this data is further processed using the eigenvector estimation stage 606. Generally, stage 606 includes converting all samples in the 5D FRI vector to their first eigenvector using the eigenvector coefficients in vector V. In one aspect, vector V is a 5-element vector of weights that multiplies 5 RIs to obtain the most prominent respiratory vector. Vector V is obtained, in one aspect, via SVD decomposition and RI = U * S * V'.

[0074] Step 606a includes measuring a break in the differential RI (DRI) signal. Step 606b includes checking that there is no breathing gap and that the data is valid. Step 606c includes calculating V using SVD. SVD is calculated only if the FRI buffer is filled with a valid number and no breathing gap is found. In one embodiment, this step may include calculating V at 1-second intervals. When V is obtained in step 606d, the data can be further processed in steps 628 and 632. Step 606e is a validity check, and if a soft reset signal is received, the vector V is reset to 0.

[0075] Step 628 includes checking whether (V1)>0. If not, step 630 includes waiting for the next round of data. If so, step 632 defines RV = [FRI×V, DRI×V]. Calculating V (or RI2RV used in one aspect of the algorithm) includes the following steps. When the RV buffer is full, a 2-second test can be performed to determine whether there is any breathing gap (i.e., due to apnea) in the DRI buffer. If the buffer is full and no gap exists, calculations are performed to determine three different Vs by three different SVDs, namely one for the first 30 seconds, one for the last 30 seconds, and one for the entire 60-second DRI buffer. If all three Vs are correlated (i.e., within a predetermined value of each other), the 60-second V is selected as the new V to proceed. Next, the process verifies that the minimum value of the resulting respiratory signal RV(1) correlates with the maximum value of the y location of the probe. In one aspect, the minimum value of the respiratory signal (RV1) preferably represents the end of exhalation (compared to within exhalation). For this process, the corresponding Y value of the catheter location is checked, and since the Y value is maximum in humans at the end of exhalation, this process ensures that the Y value at the respiratory minimum is greater than the Y value at the respiratory maximum. In this orientation, the Y direction is oriented in the patient's heart-head direction when the patient is lying supine.

[0076] If the resulting respiration signal RV(1) does not correlate with the maximum value at the y location of the probe, the selected V is inverted to be negative (i.e., -V). Finally, step 634 includes the last fraction buffer used in step 626 to provide RV.

[0077] As shown at the top of FIG. 7B (included in FIG. 7A for providing duplicate references), the filtered motion is provided from step 626 and RV is provided from step 634. Step 636 includes providing a correlation matrix A that takes into account weight buffering and respiration gaps. Generally, FIG. 7B shows steps for motion decomposition according to one aspect. This motion decomposition assumes that the filtered catheter motion (FCM) can be modeled as the sum of two main components: (1) Respiration compensation vector (RCV) = correlation matrix A × RV, and (2) catheter-heart motion = Smooth(FCM - RCV) = Spline(FCM - RCV).

[0078] The term "smooth(x)" as used herein generally refers to a mathematical smoothing function that can be implemented according to any computer interface, electronic device, or programming tool such as MATLAB®. Various other functions such as "smooth", "exp", "pinv", "spline", "rms", "norm", etc. used herein can be implemented using any known computer interface, electronic device, or programming tool such as MATLAB®. These computer interfaces, electronic devices, or programming tools are generally configured to provide matrix, array, and data manipulation, plot functions, data implementation, and / or algorithm implementation.

[0079] As used herein, the term "smooth" with respect to a function or equation refers, in one aspect, to a method of noise reduction or a data set. In one aspect, the smoothing function uses a moving average, a filter, or a smoothing spline.

[0080] As used herein, the term "spline(x)" with respect to a function or equation refers to a function that calculates, returns, or generates a vector having interpolated values corresponding to the points of an array, matrix, or data set x. In one aspect, cubic spline interpolation can be used.

[0081] As used herein, the term "exp(x)" with respect to a function or equation refers to the exponential function e x and determines each element within an array, matrix, or data set x.

[0082] As used herein, the term "pinv(x)" with respect to a function or equation refers to a function that calculates, returns, or generates the Moore-Penrose pseudoinverse of an array, data set, or matrix x.

[0083] As used herein, the term "norm(x)" with respect to a function or equation refers to a function that determines a scalar value or magnitude value and calculates, returns, or generates a measure of the magnitude of the elements within an array, matrix, or data set x.

[0084] As used herein, the term "max(x)" with respect to a function or equation refers to a function that calculates, returns, or generates the maximum element of an array, matrix, or data set x.

[0085] As used herein, the term "rms(x)" with respect to a function or equation refers to a function that calculates, returns, or generates the root mean square of an array, matrix, or data set x.

[0086] Any one or more of the function terms used in this specification may be implemented using any computing element or software program configured to perform computational analysis and functions. One such example of a program is MATLAB®. Those skilled in the art will understand that these mathematical or programming functions, and their equivalents or variations, may be implemented to perform certain functions, calculations, or operations on the data, matrices, arrays, and information described herein. Any of the functions described herein may be implemented on computing system 4, or any other electrical or computer hardware and software.

[0087] In one embodiment, the decomposition is calculated every 100 ms. These 100 ms segments can be calculated over the entire last minute of data to enable slow adaptation of the correlation matrix A over the last minute.

[0088] In one aspect, two versions of catheter-heart motion are used by controlling the curvature of the spline. In one version, the intermediate catheter motion is calculated using a single curvature or smoothing parameter. Step 638 includes estimating the intermediate catheter motion using the following calculation: Smooth(FCM-RCV). This smoothing parameter limits the speed of spline direction changes, which is equivalent to assuming slow acceleration. The intermediate catheter motion is used in the estimation of the correlation matrix A. Specifically, the correlation matrix A is adjusted using segments with a slow intermediate catheter motion speed (i.e., slow segments). In these slow segments, the catheter acceleration is small with respect to respiratory motion, and the correlation matrix (A) is estimated more accurately. During step 638, the intermediate catheter motion is estimated using a smoothing function, limiting the resulting curvature.

[0089] In one aspect, the intermediate catheter motion (ICM) = Smooth(T0, (FCM - RCV), SmoothP). In this equation, T0 is a vector of reference times (from 0 to 60 seconds in 0.1 - second increments). The inaccuracy is estimated by the norm function of the estimation error. Using step 638, the ICM has an error component (ICMerr), where ICMerr = Norm(FCM - RCV - ICM), SmoothV = Exp(-2 * ICMerr), and SmoothP = 0.005. This process corrects or updates the ICM to allow for a greater curvature if the ICMerr is unacceptably large. When the ICM is equal to FCM - RCV, the ICMerr is 0.

[0090] SmoothP is a smoothing parameter that limits the second - derivative (i.e., acceleration) of the spline curve and is generally defined as a smoothing parameter for controlling the curvature through a segment. SmoothV is generally defined as a smoothing vector for controlling the curvature per data point. During step 638, it is assumed that slow acceleration helps converge the correlation matrix A. The slow acceleration, i.e., the intermediate catheter motion (ICM), is used in combination with the faster acceleration, i.e., the catheter motion (CM), to provide a more robust and accurate reconstruction. The ICMerr or error of the ICM is the estimation error, which is equal to Norm(FCM - RCV - ICM). The ICMerr is used to estimate the accuracy of the reconstruction of the FCM. Based on this, the spline is corrected if the error of the slow motion is large. In one aspect, the smoothing function smooths the data using a moving - average filter.

[0091] In another aspect shown in step 648, the catheter motion (CM) is a corrected ICM using a correction vector (SmoothV), which allows for high-speed acceleration when it is determined that the intermediate catheter motion is inaccurate. This correction vector (SmoothV) allows for different curvatures (i.e., accelerations) at different spots along the last minute of the data. CM is equal to Smooth(T0, (FCM - RCV), SmoothP, smoothV). This process is a way to make the smoothing function more flexible and allows for greater curvature at points where the reconstructed error is larger.

[0092] Step 650 includes estimating the catheter motion based on the following equation, i.e., Smooth(FCM - RCV, SmoothV). In one aspect, a function or process is executed to smooth the catheter motion using a spline model by restricting the spline curvature (i.e., the second derivative of the curvature).

[0093] When estimating CM, the high-speed segments are allowed to have increased curvature in order to provide a more accurate and better reconstruction. The smoothing spline function f is given by the following equation.

[0094]

Equation

[0095] In this equation, |z|2 represents the sum of the squares of all entries of z, n is the number of entries of x, and the integral is over the smallest interval containing all entries of x. The default value of the weight vector W in the error measure is 1(size(x)). The default value of the piecewise constant weight function smoothV in the roughness measure is the constant function 1.

[0096] In Equation 3, D 2f represents the second derivative of the function f. The parameter smoothP controls the smoothing intensity. When smoothP = 1, the spline passes through all the original data points without curvature limitation. When smoothP = 0, the spline cannot contain curvature, so the result is a straight line. In this situation, smoothing can be used to control the catheter allowable acceleration.

[0097] For intermediate catheter motion (ICM), only non-zero smoothP parameters are used, and sharp motions are smoothed or filtered out. To model fast or sharp motions, a smoothV vector with values less than 1 is used. SmoothV values less than 1 add additional weight or emphasis to certain aspects, and thus can include higher curvature in those locations.

[0098] The overall concept of smoothing in this aspect is to balance two opposing elements. The first element attempts to remain true or accurate to the input vector even when there is noise (i.e., many inconsistent directional changes). The second element should be as smooth as possible with minimal directional and speed changes. SmoothV is a vector of values that indicates in different parts of the input vector whether the smoothing function should prioritize the first element over the other, and vice versa. Thus, as the SmoothV value approaches 0, the smoothing function should be more accurate and less smooth. In one aspect, smoothV allows smoothP of Equation 3 to be localized.

[0099] From step 650, the catheter movement is used for site segment detection in step 652. In step 654, the catheter movement and site segment detection data are sent to a central module 655 configured to store data regarding the site segment and catheter movement. Next, an iterative process is executed that analyzes the session movement and recalculates the ablation site. The ablation is divided into sites and recalculated for each step and can vary significantly according to the compensated catheter-heart movement. This compensated location of the catheter can be stored in the central module or memory unit for site recalculation after the user or surgeon adjusts various control parameters.

[0100] Step 640 includes calculating the RCV for the last minute of information and data. In one embodiment, RCV = A × RV. In one aspect, the correlation matrix A is also identified as RV2RCV. Estimating the correlation matrix A is generally performed using weighted mean square and matrix inversion according to one embodiment. The weights (W) used in these calculations reflect how each data point in the last minute is constructed to estimate the correlation matrix A according to at least one of the following criteria: (i) weight decay over time, (ii) preference for slow intermediate catheter movement speed, (iii) RV validity, and (iv) duration from ablation.

[0101] For factor (i), older samples generally receive lower weights than more recent samples. For factor (ii), slower speeds on longer segments receive higher weights. For factor (iii), during apnea, the RV is reduced to very near 0 and cannot be used in estimating the correlation matrix A. Thus, apnea segments receive a weight of 0. Apnea detection is performed at step 642. Apnea is also commonly known as depth or shallowness of breathing. Factor (iv) results in more recent ablation segments receiving higher weights. The weight vector is identified as the stability weight, and the step of estimating the stability weight is generally shown as step 644 in FIG. 7B. In other words, the intermediate catheter movement from step 638 is used to estimate the stability weight at step 644.

[0102] The apnea detection related to factor (iii) and step 642 is defined in more detail. Respiratory apnea is a time segment in which the respiratory signal, i.e., the RV, decreases due to apnea or other forms of shallow breathing. Due to their effects, the correlation matrix A generally should not be estimated between these segments. By using the weights associated with each sample (including RV and location data), the effects of apnea or shallow breathing are minimized. The weights during shallow breathing (i.e., small RV signals) are set to zero, which eliminates the influence of these measurements on the overall estimation and the correlation matrix A.

[0103] Figure 8 shows the effect of apnea on the RV during ablation. The respiratory gap is generally defined as a time segment during which the respiratory conductance function cannot reach a predetermined extreme value and can only reach less than 85% of the predetermined extreme value. When used in this context, these small and large parameters are defined by analyzing the 15 - 85 percentiles and ignoring the signals above and below these percentiles. The continuous samples within the 15 - 85 percentiles are counted and analyzed. If the signals within a continuous segment happen to occur within these thresholds and do not cross them, the segment of all samples within this continuous segment is shown as a respiratory gap, i.e., shallow breathing, and can thus be discarded. The percentile values are selected over the last minute of the data and define the thresholds for the respiratory differential amplitude. When the respiratory differential signal is between the thresholds over a predetermined period such as 5 seconds (corresponding to approximately one respiratory cycle), the respiratory signal is considered too weak and should thus be discarded. The percentile values can be modified and generally selected to identify shallow breathing. Those skilled in the art will understand that other filtering or analysis can be used to identify shallow breathing.

[0104] Further details regarding estimating stability weights are provided herein. Generally, the decomposition process into respiratory-related and catheter-related motion is performed more accurately when the catheter is relatively stable or sliding slowly relative to another surface, i.e., tissue. Identifying relatively stable segments is important to produce a reliable output. A stable segment is generally defined herein as a segment of time where the intermediate catheter motion speed is relatively slow (i.e., <2.5 mm / sec) over a particular period (i.e., >3 - 5 seconds). Estimating the correlation matrix A is performed using stable segments or essentially stable segments. Stable segments are emphasized by assigning higher weights to samples during the stable time. The stability weights are converted into a vector of numbers corresponding to the last minute and representing the point-wise correlation for the correlation matrix A estimation (i.e., the RV2RCV matrix). In addition to speed stability, recency (i.e., the age or newness of the segment) and a factor regarding the data segment collected during ablation are also considered. More recent samples receive higher weights than older samples, and ablation segments receive higher weights than non-ablation segments.

[0105] Figure 9 shows one aspect of the recency factor regarding stability weights. As shown in Figure 9, higher weights are assigned to more recent samples (i.e., the right side of the graph), and older samples receive lower weights. As shown in Figure 9, the sigmoid provides a smooth transition between recent and old signals.

[0106] Figure 10 shows one aspect of the speed factor regarding stability weights. As shown in Figure 10, higher weights are used when the catheter estimated speed is low, and lower weights are used when the catheter estimated speed is high. In one aspect, the weight is decreased when the low-speed duration is shorter than the average respiratory cycle. As shown in Figure 10, this duration is adjusted by artificially increasing the estimated speed.

[0107] FIG. 11 shows one aspect of the ablation coefficient regarding the stability weight. Generally, as shown in FIG. 12, when ablation occurs, higher weights are assigned to the signals. In one aspect, the three coefficients described above with respect to FIGS. 9, 10, and 11 are used to generate the total stability weight. In other words, the total stability weight vector can be the product of the samples of all three coefficients. Those skilled in the art will understand that more than three coefficients can be used. FIG. 12 shows how the stability weight coefficient can be measured over time. In one aspect, since the freshness coefficient is fixed, this coefficient is calculated only at initialization.

[0108] Returning to FIG. 7B, using the weight (W), the correlation matrix A is estimated in step 646 using the following equation: A(2×3)=Pinv(RV)×W(FCM - RCV). As used in this aspect, A(2×3) converts RV (a 2-column matrix representing the respiratory signal and its derivative) to RCV (a 3-column matrix representing the X - Y - Z of the respiratory matrix). The matrix A is the transformation matrix between the respiratory vector (RV) and the 3D respiratory ellipsoid. This equation represents a regression using the mean square, as FCM = A * RV + CM+ err. Since FCM cannot be fully decomposed into these models, there is some error (err) representing the difference between FCM or the actual movement and the sum of the two models. Essentially, these processes adjust the transformation matrix A to minimize the error. To minimize the error, the correlation matrix A is regressed such that A = ((FCM-(CM)) / RV which is equal to Pinv(RV) * (FCM-(CM)). In one aspect, when a weight W is added to give a larger weight to the time segment with a slow (CM), the equation becomes A = Pinv(W * RV) * W * (FCM(CM)).

[0109] As shown in FIG. 7B, the correlation matrix A in step 646 is then sent back to step 636 to provide a roundabout or continuous cycle for continuously estimating and updating the correlation matrix A. The process in FIG. 7B is iterative such that the correlation matrix A is continuously updated and becomes more accurate through each iteration. The correlation matrix A is generally updated based on any one or two or more of the calculated intermediate catheter motion, the step of estimating an apnea segment based on the RV, the step of estimating a stability weight based on the intermediate catheter motion, and the step of regressing the correlation matrix A based on A = (FCM - ICM) / RV.

[0110] In one aspect, it is assumed that the 3D RCV can be modeled as the product of the 2D RV (equal to the main respiratory signal + phase shift) and the correlation matrix A (2×3). The parameters of the correlation matrix A are estimated based on the RV, a rough estimate of the respiratory vector (equal to subtracting the intermediate catheter motion from the filtered catheter motion), and a stability weight vector by minimizing the mean squared error. The correlation matrix A is generated as an output by providing the inputs of (i) FCM, (ii) the intermediate catheter motion, and (iii) the stability weight.

[0111] The intermediate catheter motion assumes a relatively slow acceleration of the catheter and cannot represent the true catheter motion in the high-speed segment. The intermediate catheter motion error coefficient can be estimated based on the segments of faster motion. The catheter motion is recalculated while increasing the curvature of the high-speed segment, enabling high-speed motion as well. Using this approach, ICM error (ICMerr) = Norm(FCM - ICM - RCV). In other words, the respiratory data is equal to the total catheter motion minus the intermediate catheter motion, and the total catheter motion minus respiration is equal to the intracardiac catheter motion. In one embodiment, ICMerr is the Euclidean distance for each sample in R3 between the measured location FCM and the reconstructed location ICM + RCV of the low-speed model. SmoothV, in one aspect, is used to correct the intermediate catheter motion to find the intracardiac catheter motion.

[0112] Localization, also known as location discovery or site discovery, generally involves identifying the period during which the catheter or probe is stable. This process includes performing the logic and functions that generate the local time edges and central locations based on the compensated motion. The localization in the last session is continuously updated along with the correlation matrix A and the resulting compensated motion. Stability is generally defined as a continuous time segment in which the estimated catheter speed is below the user-predefined threshold.

[0113] The catheter speed is calculated by taking the derivative of the catheter position over the last minute. As an example, the locations for the last minute are updated at least every 0.1 seconds and stored for one minute. These values can vary according to the specific requirements of a particular application. The last catheter speed value from this buffer is saved and accumulated in the last speed buffer.

[0114] For example, in one embodiment, the end-5 value in the catheter speed buffer is the speed of the catheter that uses the end-5 value and the end-7 value in the updated catheter 1-minute location buffer in the current iteration. Here, the end-5 value in the last speed buffer is the last catheter speed calculated 5 iterations ago.

[0115] The minimum value of the last 1-minute catheter speed and the last speed is defined as the minimum speed and is used to detect when the catheter is stable. The minimum speed buffer contains all the minimum speeds estimated in the last 1 minute. While trying to find the period during which the catheter was stable, segments of consecutive speeds smaller than the threshold speed are identified within this buffer.

[0116] Estimates close to the current time are not accurate enough because the speed can change dramatically near the buffer end without a reliable indication regarding the estimated speed. Therefore, any decision regarding stability can be delayed by at least 1 second, and this predetermined delay can vary.

[0117] The boundaries of the current stable segment are updated and can then be terminated as needed. If the current stable segment is not found or if it has ended previously, a search for a new stable segment in the most recent few seconds of the estimated catheter movement is performed.

[0118] If a stable segment has already been found, the boundaries are updated according to the updated minimum speed buffer. The starting point of the current stable segment can also be updated by looking backward from the last index of the stable segment to the earliest point where the speed is lower than the threshold.

[0119] The end point of the stability segment can be updated by looking forward from the last index of the stable end to the most recent point where the catheter speed is lower than the threshold.

[0120] The stable segment does not end until a predetermined period (i.e., approximately 1 second) has elapsed from the current time in order to avoid the segment ending due to edge inaccuracies. If a stable segment has not yet been found, a backward search for stability from the current time is performed to find the earliest point that exceeds a predetermined threshold.

[0121] In addition to the above process for determining when a segment is stable, the present disclosure also relates to defining a localized segment as a time segment in which all of its points (i.e., the distance from its own centroid) are below a threshold. In other words, when considering a curve in 3D space, the XYZ average of the curve can be calculated, which is the centroid. From this value, it is possible to calculate the distance from the center (DFC) for each point on the line.

[0122] This process involves localizing the site (i.e., the time segment) when the ablator is on and the catheter is stable and localized. One aspect of this process is described in more detail herein. In one aspect, the stable time segment during ablation is identified.

[0123] The processes and algorithms disclosed herein can divide, segment, or otherwise split the stability segment during ablation into localized segments (i.e., sites) where the maximum DFC is less than a predetermined threshold. In one aspect, the predetermined threshold is 3.0 mm. One of ordinary skill in the art will understand that this value can vary. The algorithm for splitting the stability segment into sites essentially obtains the minimum number of subsegments that maintain the localization criteria. In one aspect, it is a recursive function that ends or stops when the current subsegment is localized. In one embodiment, this includes determining whether the maximum DFC is less than the threshold, i.e., less than 3.0 mm in one aspect.

[0124] In one aspect, an algorithm is provided that essentially divides or segments a curve in a 3D dimensional space (i.e., cardiac motion 50, catheter-heart motion, probe-cavity motion, etc.) into localized curves. As used herein, the term curve is defined as a vector of XYZ coordinates in 3D space representing the course of an object's motion. In one aspect, this process includes determining the centroid of the curve defined by the dataset, where the centroid is defined as a point in 3D space representing the average value of each coordinate of the curve. The DFC is defined, in one aspect, as a one-dimensional vector representing the Euclidean distance between each point within the curve relative to the centroid. Finally, the localized curve is defined as a curve where the maximum DFC is less than a predetermined threshold distance. In one aspect, this distance is 3.0 mm.

[0125] In one aspect, defining a segmentation curve of 3D dimensional data is a recursive function. This definition of the segmentation curve is provided in the flowchart 1300 of FIG. 13. First, the flowchart 1300 includes, at 1310, finding a central localization curve based on the acquired data (i.e., intra - cardiac motion 50, catheter - cardiac motion, probe - chamber motion, etc.). Step 1310 can include, at 1320, determining whether a test curve (i.e., an input curve) defined by a data set localizes, i.e., whether the maximum DFC is less than a threshold. At 1310, an index of the minimum DFC for a given curve can be identified at 1330, and an index of the maximum DFC for a given curve can be identified at 1340. If the index of the maximum value is less than the index of the minimum value (i.e., if the maximum DFC is before the minimum DFC), at 1350, the first point of the test curve data can be excluded, i.e., a data point is secured. If the index of the maximum value is not less than the index of the minimum value, the last point of the given curve data can be excluded at 1360, i.e., can be secured. The flowchart 1300 includes, at 1370, recognizing or registering the central localization curve as a test curve. Basically, a test curve that is a potential candidate for the central localization curve is repeatedly cut from either side until it localizes and thus is declared or identified as the central localization curve. In certain situations, the input curve or the initial test curve localizes immediately, and thus the cutting steps can be omitted. In these situations, the input curve is the central localization curve, and thus a segmentation algorithm is not required.

[0126] Flow diagram 1300 checks, at 1380, whether there are curves before the central localization curve. In one aspect, this step is performed based on the excluded or secured prior center points (disclosed above with respect to the index comparison step) by using a recursive partitioning function. This process generates, at 1390, a list of prior central localization curves for analysis based on 1380. Flow diagram 1300 checks, at 1385, whether there are curves after the central localization curve. In one aspect, this step is performed based on the excluded or secured posterior center points by using a recursive partitioning function. This process generates, at 1395, a list of posterior central localization curves for analysis based on 1385. The list of all localization curves includes, at 1399, the leading localization curve, the central localization curve, and the trailing localization curve. In other words, the algorithm essentially segments or divides the 3D curve into several localization curves for further analysis. In particular, the analysis determines whether the curve has the maximum DFC that is less than a predetermined distance threshold.

[0127] FIG. 14 shows a flow diagram illustrating various steps of a method 1400 for considering or minimizing the effect of the respiratory component in catheter movement data. As shown in FIG. 14, step 1410 includes acquiring respiratory data. This data can be obtained by sensors attached to the patient's body. In one aspect, this data is acquired by a series of patches or sensors (i.e., sensor 18 of FIG. 1) attached to the patient's body configured to detect impedance values based on the patient's lungs being filled with air. In another aspect, this data is acquired via a sensor configuration that includes a catheter integrated with the sensor and an external sensor, or based on only the sensor integrated with the sensor. Respiration is generally based on the up and down movement of the patient's diaphragm, which also displaces the patient's heart. Therefore, acquiring respiratory data is important for mapping the movement of the catheter within the patient's heart during respiration.

[0128] Step 1420 includes obtaining probe location data that can be generated via any known tracking method or system. The probe location data can also be obtained using sensors attached to the patient's body (i.e., sensor 18 in FIG. 1). In one aspect, the sensor is configured to generate a magnetic field, and the movement of a catheter inside the patient can generate a magnetic signal indicating the position of the catheter.

[0129] Step 1430 includes generating probe-cavity location data by compensating (e.g., subtracting or adding) respiratory data from the probe location data. Step 1440 includes identifying a period during which the probe is stable with respect to the cavity boundary. Step 1450 includes identifying a site stability site based on a period during which the probe speed is less than a predetermined speed over a predetermined period. In one aspect, the predetermined speed can be 2 mm / second, and the predetermined period can be at least 3 seconds. In one embodiment, method 1400 includes applying a filter to the respiratory data and the probe location data.

[0130] Step 1460 includes notifying the surgeon of the period during which the probe is stable with respect to the cavity boundary. This step can include providing the notification via a visual warning or indication (which can be displayed via monitor 3), an auditory warning or indication (which can be reproduced via computing system 4), or any other notification by which the surgeon is informed of the stable period. In one aspect, a marker, tag, or visual indicator is overlaid or inserted onto a three-dimensional mapping or image of the location data and displayed on a monitor (e.g., monitor 3 in FIG. 1).

[0131] The above-mentioned step 1410 may further include converting the respiration indicator by using singular value decomposition (SVD) to obtain a first eigenvector and a first eigenvector derivative. As understood by those skilled in the art, the first eigenvector can correspond to the primary respiration signal, and the first eigenvector derivative can correspond to the phase-shifted respiration signal. The method 1400 can include the primary respiration signal and the phase-shifted respiration signal transitioning from two dimensions to a three-dimensional ellipsoid based on a correlation matrix. The method 1400 can also include determining the correlation matrix by calculating the root mean square deviation between the probe location data and the probe-cavity motion data. The method 1400 can include determining the correlation matrix based on a weighting factor including at least one of (i) the age of the respiration data, (ii) the speed of the probe, (iii) the depth of respiration, or (iv) the data acquired during ablation. In one embodiment, the method 1400 includes generating an image including an estimated respiration motion based on respiration data, a probe motion based on probe location data, and a probe-cavity motion based on probe-cavity location data. This image can then be displayed on a monitor for a surgeon, along with an indicator, tag, or visual indicator indicating the period during which the probe is stable relative to the cavity boundary.

[0132] FIG. 15 shows a method 1500 for capturing mapping data with respiratory compensation. In the mapping configuration, as described above, the predetermined speed of the probe location data can be small or even zero. Further, in a mapping configuration where there is no ablation, the input for ablation is also zero. Thus, in such a configuration, the movement of the electrode relative to the tissue wall can be mainly the result of respiratory motion.

[0133] In step 1510, in the mapping configuration, signals from a plurality of electrodes are recorded over a period of about 2.5 seconds. As a result of the use of the 2.5 - second period, the heartbeat of the heart can be explained. As a result of the duration that explains the heartbeat, movement due to breathing is the major remaining factor that affects the mapping. This movement due to breathing can cause the obscuring of the data.

[0134] In step 1520, the movement of breathing can be compensated in order to reduce or remove the smearing of the data. As described above with reference to FIG. 2, this movement can be important. As described above, since the breathing movement is known and understood, the mapping points collected during the breathing movement can be compensated. For example, all of the points captured during a breathing cycle may be adjusted to the end of the breathing phase in order to produce a better anatomical mapping (less obscure). In such a configuration, the sharpness of the image is improved by adjusting the data points to remove the movement of breathing. As will be appreciated, the compensation of the mapping points can be positioned at a reference point between any part of the breathing cycle, including the start, mid - point, or any other part of the breathing cycle. The end of breathing is used as an example to aid in the understanding of the explanation. This data collection can be referred to as compensated movement with respect to the mapping.

[0135] In step 1530, additionally or alternatively, the stability of each electrode (or the spline of the catheter) may be monitored as described herein. When data is collected, stability is used to determine mapping values. This determination can include at least one of adjusting the mapped data points based on stability and calculating only the mapping values of electrodes considered to be stable. For example, when data collection is performed, data from stable electrodes can be recorded while unstable electrodes are excluded or their data is recollected. Alternatively, a weighting system can be used based on the level of stability during data collection. In any example, the waits or weights may be used in harmony, but the data is improved and the blurring is reduced. Using a gating function based on stability improves the quality of the collected data.

[0136] In step 1540, respiratory data can be acquired. In step 1550, respiratory data can be acquired to compensate for the respiratory data. This data can be obtained by sensors attached to the patient's body. In one aspect, this data is obtained by a series of patches or sensors (i.e., sensor 5 of FIG. 1) attached to the patient's body configured to detect impedance values based on the patient's lungs being filled with air. In another aspect, this data is obtained via a sensor configuration including a catheter integrated with the sensor and an external sensor, or based only on the sensor integrated with the sensor. Respiration is generally based on the up and down movement of the patient's diaphragm, which also displaces the patient's heart. Therefore, acquiring respiratory data is important for mapping the movement of the catheter within the patient's heart during respiration.

[0137] In step 1550, in order to obtain the first eigenvector and the first eigenvector derivative, the respiratory data can be converted into an indicator by using singular value decomposition (SVD). The first eigenvector can correspond to the primary respiratory signal, and the first eigenvector derivative can correspond to the phase-shifted respiratory signal. Method 1400 can include that the primary respiratory signal and the phase-shifted respiratory signal transition from two dimensions to a three-dimensional ellipsoid based on the correlation matrix. Method 1400 can also include determining the correlation matrix by calculating the root mean square deviation between the probe location data and the probe-cavity motion data. Method 1400 can include determining the correlation matrix based on a weighting factor including at least one of (i) the age of the respiratory data, (ii) the speed of the probe, (iii) the depth of respiration, or (iv) the data obtained during ablation. In one embodiment, the method 1400 includes generating an image including an estimated respiratory motion based on the respiratory data, a probe motion based on the probe location data, and a probe-cavity motion based on the probe-cavity location data. This image can then be displayed on a monitor for the surgeon, along with an indicator, tag, or visual indicator indicating the period during which the probe is stable with respect to the cavity boundary.

[0138] FIG. 16 shows a plot 1600 of respiratory data points with and without respiratory compensation. Plot 1600 includes data on the X-axis from -5 to +5 mm. Plot 1600 includes data on the Y-axis from -4 to +4 mm. In plot 1600, data point 1610 represents data without respiratory compensation. Data point 1620 represents the same data with compensation. Additionally, data point 1620 includes points where the data is shifted from a continuum of the respiratory cycle, for example, to the end point of the cycle. As shown, the data points 1610 without respiratory compensation have a Y-axis spread of approximately 8 mm and an X-axis spread of approximately 5 mm. The data points 1620 with compensation have a Y-axis spread of less than 0.5 mm and an X-axis spread of less than 0.5 mm.

[0139] The disclosed subject matter is not limited to use in connection with the heart. The disclosed subject matter can be used in a variety of applications for analyzing the characteristics of any type of object, such as a heart chamber.

[0140] Any of the functions and methods described herein can be implemented on a general-purpose computer, processor, or processor core. Suitable processors include, by way of example, general-purpose processors, dedicated processors, conventional processors, digital signal processors (DSPs), multiple microprocessors, one or more microprocessors associated with a DSP core, controllers, microcontrollers, application specific integrated circuits (ASICs), field programmable gate array (FPGA) circuits, any other type of integrated circuit (IC), and / or state machines. Such processors can be manufactured by configuring a manufacturing process using the results of other intermediate data, such as processed hardware description language (HDL) instructions and netlists, which can be stored on a computer-readable medium. The result of such processing can be a mask work, which can then be used in a semiconductor manufacturing process to manufacture a processor implementing the features of the present disclosure.

[0141] All of the functions and methods described in this specification can be implemented in a computer program, software, or firmware incorporated in a non-transitory computer-readable storage medium and executed by a general-purpose computer or processor. Examples of non-transitory computer-readable storage media include read only memory (ROM), random-access memory (RAM), registers, cache memory, semiconductor memory devices, magnetic media such as internal hard disks and removable disks, magneto-optical media, and optical media such as CD-ROM disks and digital versatile disks (DVDs).

[0142] It should be understood that many modifications are possible based on the disclosure of this specification. Although the features and elements have been described above in specific combinations, each feature or element can be used alone without other features and elements, or in various combinations with or without other features and elements.

[0143] [Embodiments] (1) A method comprising: acquiring respiratory data of a patient via at least one sensor; acquiring probe location data about a probe positioned within a cavity; generating compensated location data by compensating the acquired probe location data with the acquired respiratory data; and creating a mapping based on the generated compensated location data. (2) The method according to Embodiment 1, further comprising applying a filter to the respiratory data. (3) The method according to Embodiment 1, wherein acquiring the respiratory data further comprises acquiring a respiration indicator via the at least one sensor. (4) Obtaining the respiratory data further includes converting the respiratory indicator by using singular value decomposition (SVD) to obtain a first eigenvector and a first eigenvector derivative function, where the first eigenvector corresponds to a primary respiratory signal and the first eigenvector derivative function corresponds to a phase-shifted respiratory signal, according to the method described in Embodiment 3. (5) The method according to Embodiment 4, wherein the primary respiratory signal and the phase-shifted respiratory signal transition from a two-dimensional to a three-dimensional ellipsoid based on a correlation matrix.

[0144] (6) The method according to Embodiment 5, wherein the correlation matrix is determined based on a weighting factor including at least one of the age of the respiratory data, the speed of the probe, and the depth of respiration. (7) Further including generating an image including an estimated respiratory movement based on the respiratory data, a probe movement based on the probe location data, and a probe-cavity movement based on probe-cavity location data, according to the method described in Embodiment 1. (8) Further including generating a notification when the probe is stable with respect to the boundary of the cavity, according to the method described in Embodiment 1. (9) Notifying the period during which the probe is stable with respect to the boundary via a visual indicator displayed on a monitor, according to the method described in Embodiment 8. (10) Further including identifying a site stability site based on a period during which the probe speed is less than a predetermined speed for at least a predetermined period, according to the method described in Embodiment 1.

[0145] (11) A system, a probe configured to be inserted into a body cavity of a patient, at least one sensor configured to obtain respiratory data, wherein the probe and the at least one sensor are configured to obtain probe location data, at least one sensor, a processor, and the processor is By compensating the obtained probe location data with the obtained respiration data, compensated location data is generated. A system configured to create a mapping based on the generated compensated location data. (12) The system according to embodiment 11, wherein the processor is further configured to apply a filter to the respiration data and the probe location data. (13) The system according to embodiment 11, wherein the processor is further configured to generate a respiration indicator based on the respiration data from the at least one sensor. (14) The system according to embodiment 13, wherein the processor is further configured to transform the respiration indicator by using singular value decomposition (SVD) to obtain a first eigenvector and a first eigenvector derivative, the first eigenvector corresponding to a primary respiration signal, and the first eigenvector derivative corresponding to a phase-shifted respiration signal. (15) The system according to embodiment 14, wherein the primary respiration signal and the phase-shifted respiration signal transition from a two-dimensional to a three-dimensional ellipsoid based on a correlation matrix.

[0146] (16) The system according to embodiment 15, wherein the correlation matrix is determined based on a weighting factor including at least one of the age of the respiration data, the speed of the probe, and the depth of respiration. (17) The system according to embodiment 11, wherein the processor is further configured to generate an image including an estimated respiration motion based on the respiration data, a probe motion based on the probe location data, and a probe-cavity motion based on probe-cavity location data. (18) The system according to embodiment 11, wherein the processor is further configured to notify a period during which the probe is stable with respect to the boundary of the cavity. (19) The system according to embodiment 18, wherein notifying the period during which the probe is stable with respect to the boundary includes displaying a visual indicator on a monitor. (20) The system according to embodiment 11, wherein the processor is further configured to identify a site stability site based on a period during which the probe speed is less than 2 mm / second for at least 3 seconds.

Claims

1. 1. A system comprising: a probe configured to be inserted into a body lumen of a patient; at least one sensor configured to acquire respiratory data, the probe and the at least one sensor configured to acquire probe location data; a processor, the processor comprising: generating compensated location data by compensating the acquired probe location data with the acquired respiration data; The system is configured to generate a mapping based on the generated compensated location data.

2. The system of claim 1 , wherein the processor is further configured to apply a filter to the respiratory data and the probe location data.

3. The system of claim 1 , wherein the processor is further configured to generate a respiratory indicator based on the respiratory data from the at least one sensor.

4. 4. The system of claim 3, wherein the processor is further configured to transform the respiratory indicator by using singular value decomposition (SVD) to obtain a first eigenvector and a first eigenvector derivative, the first eigenvector corresponding to a primary respiratory signal and the first eigenvector derivative corresponding to a phase shifted respiratory signal.

5. The system of claim 4 , wherein the primary respiratory signal and the phase-shifted respiratory signal are transformed from a two-dimensional to a three-dimensional ellipsoid based on a correlation matrix.

6. The system of claim 5 , wherein the correlation matrix is ​​determined based on weighting factors including at least one of age of the respiratory data, speed of the probe, and respiratory depth.

7. 2. The system of claim 1, wherein the processor is further configured to generate an image including estimated respiratory motion based on the respiratory data, probe motion based on the probe location data, and probe-cavity motion based on the probe-cavity location data.

8. The system of claim 1 , wherein the processor is further configured to notify a period during which the probe is stable relative to a boundary of the cavity.

9. The system of claim 8 , wherein indicating a period during which the probe is stable relative to the boundary comprises displaying a visual indicator on a monitor.

10. The system of claim 1 , wherein the processor is further configured to identify a site stability site based on a period during which the probe speed is less than 2 mm / sec for at least 3 seconds.

11. 1. A method comprising: acquiring respiratory data of a patient via at least one sensor; acquiring probe location data for a probe positioned within the cavity; generating compensated location data by compensating the acquired probe location data with the acquired respiration data; generating a mapping based on the generated compensated location data.

12. The method of claim 11 , further comprising applying a filter to the respiratory data.

13. The method of claim 11 , wherein obtaining the respiratory data further comprises obtaining a respiratory indicator via the at least one sensor.

14. 14. The method of claim 13, wherein obtaining the respiratory data further comprises transforming the respiratory indicator by using singular value decomposition (SVD) to obtain a first eigenvector and a first eigenvector derivative, the first eigenvector corresponding to a primary respiratory signal and the first eigenvector derivative corresponding to a phase shifted respiratory signal.

15. 15. The method of claim 14, wherein the primary respiratory signal and the phase-shifted respiratory signal are transformed from a two-dimensional to a three-dimensional ellipsoid based on a correlation matrix.

16. The method of claim 15, wherein the correlation matrix is ​​determined based on weighting factors including at least one of age of the respiratory data, speed of the probe, and respiratory depth.

17. an estimated respiratory motion based on the respiratory data; and a probe movement based on the probe location data; and probe-cavity motion based on the probe-cavity location data.

18. The method of claim 11 , further comprising generating a notification when the probe is stable against a boundary of the cavity.

19. 20. The method of claim 18, further comprising indicating the period during which the probe is stable relative to the boundary via a visual indicator displayed on a monitor.

20. 12. The method of claim 11, further comprising identifying site stability sites based on the period during which the probe speed is below a predetermined rate for at least a predetermined period of time.

Images