local activation driver classification mapping in atrial fibrillation

Abstract

The invention is entitled "Classification of local activation driver programs in atrial fibrillation." The invention provides a method that includes acquiring respective signals from a plurality of electrodes in contact with cardiac tissue undergoing atrial fibrillation, and computing a mutual information metric between pairs of the plurality of electrodes from the signals. A graph is generated with the electrodes as nodes, and edges as connections between the nodes that exceed a selected mutual information metric threshold. A respective local efficiency metric is computed for each node, and the metrics are averaged to formulate a resulting local efficiency for the selected mutual information metric threshold. The resulting local efficiency and the selected mutual information metric threshold are analyzed to classify the atrial fibrillation.

Description

[0001] Cross-references to related applications

[0002] This application claims the benefit of U.S. Provisional Patent Application 63 / 133,723, filed January 4, 2021, which is incorporated herein by reference. Technical Field

[0003] This invention relates generally to cardiology, and more specifically to cardiac arrhythmias. Background Technology

[0004] Atrial fibrillation (AF) is the most common cardiac arrhythmia, and it is projected to affect 600-1200 million people in the United States by 2050 and 17.9 million people in Europe by 2060. Radiofrequency (RF) ablation, or irreversible electroporation (IRE) or pulsed field (PF) ablation, are treatment options for AF, which work by altering the path of electrical waves in the heart. However, for ablation to be applied effectively, it is important to locate the source or driver of the AF, and methods for achieving this are known. Summary of the Invention

[0005] One embodiment of the present invention provides a method comprising the following:

[0006] Signals are collected from multiple electrodes that are in contact with heart tissue undergoing atrial fibrillation;

[0007] Calculate the corresponding mutual information metric between multiple electrode pairs based on the signal;

[0008] Generate a graph with the electrodes as nodes and the edges as connections between the nodes, wherein the corresponding mutual information metric for the connection exceeds a selected mutual information metric threshold;

[0009] Based on the path length between the connected nodes, a corresponding local efficiency metric is calculated for each node, which indicates the efficiency of information exchange between the node and other nodes connected to the node.

[0010] The corresponding local efficiency metrics of the nodes are averaged to determine the resulting local efficiency for the selected mutual information metric threshold; and

[0011] The obtained local efficiency and the selected mutual information metric threshold are analyzed to classify the atrial fibrillation.

[0012] The signal can be a unipolar or bipolar voltage or a signal relating the action potential voltage to time.

[0013] Typically, calculations based on signals include estimating the local activation time (LAT) based on the signals.

[0014] In the disclosed implementation, cardiac tissue is part of the atria, and classifying atrial fibrillation includes estimating the percentage of atrial remodeling.

[0015] In another disclosed implementation, the method includes presenting a classification of atrial fibrillation to the user of the method.

[0016] The plurality of electrodes may be located on a conduit having multiple ridges.

[0017] Typically, the nearest neighbor electrode spacing among multiple electrodes is less than 3 mm.

[0018] In another implementation, averaging the corresponding local efficiency metric for the node consists of: generating a subgraph of nodes directly connected to the given node, calculating a local efficiency metric for each subgraph in the subgraph, and averaging the calculated local efficiency metrics.

[0019] In yet another implementation, the method includes the step of repeatedly generating graphs, and averaging the corresponding local efficiency metrics while incrementing the selected mutual information metric threshold, so as to produce an ordered set of resulting local efficiencies and mutual information thresholds.

[0020] The set of ordered pairs can be analyzed to classify atrial fibrillation. Typically, analyzing the set of ordered pairs involves fitting a polynomial to the set and classifying the atrial fibrillation in response to the first derivative of the polynomial.

[0021] In yet another implementation, the calculation based on the signal includes: calculating the corresponding mutual information metric between all electrode pairs based on the signal.

[0022] According to an embodiment of the present invention, an apparatus is also provided, the apparatus comprising:

[0023] A probe having multiple electrodes configured to contact cardiac tissue undergoing atrial fibrillation; and

[0024] Processor, the processor being configured to:

[0025] Receive signals from the electrodes and calculate mutual information metrics between multiple electrode pairs based on the signals;

[0026] Generate a graph with the electrodes as nodes and the edges as connections between the nodes, wherein the corresponding mutual information metric for the connection exceeds a selected mutual information metric threshold;

[0027] Based on the path length between the connected nodes, a corresponding local efficiency metric is calculated for each node, which indicates the efficiency of information exchange between the node and other nodes connected to the node.

[0028] The corresponding local efficiency metrics of the nodes are averaged to determine the resulting local efficiency for the selected mutual information metric threshold; and

[0029] The obtained local efficiency and the selected mutual information metric threshold are analyzed to classify the atrial fibrillation. Attached Figure Description

[0030] This disclosure will be more fully understood through the following detailed description of embodiments thereof, taken in conjunction with the accompanying drawings, wherein:

[0031] Figure 1 A schematic diagram of an atrial fibrillation classification system according to an embodiment of the present invention;

[0032] Figure 2 A flowchart illustrating the steps of an algorithm executed by the system's processor according to an embodiment of the present invention;

[0033] Figure 3 Two bar charts are shown according to an embodiment of the present invention;

[0034] Figure 4 This is an exemplary schematic diagram of an area according to an embodiment of the present invention;

[0035] Figure 5A and Figure 5B A schematic diagram illustrating a sub-diagram according to an embodiment of the present invention;

[0036] Figure 6A and Figure 6B This is a schematic diagram illustrating the relationship between the local efficiency and the mutual information threshold obtained according to an embodiment of the present invention;

[0037] Figure 7A , Figure 7B and Figure 7C This is another schematic diagram illustrating the relationship between the local efficiency and the mutual information threshold obtained according to an embodiment of the present invention; and

[0038] Figure 8 A schematic diagram illustrating the relationship between the local efficiency first derivative and the mutual information threshold according to an embodiment of the present invention is shown. Detailed Implementation

[0039] Overview

[0040] While locating the driver of atrial fibrillation (AF) is important, it is also crucial for classifying AF to identify the optimal strategy for RF (radiofrequency) or IRE (irreversible electroporation) / pulsed field (PF) ablation. Regions with significant remodeling (i.e., electrophysiological and / or structural changes in the atria) may be important in the maintenance mechanisms of AF. Therefore, understanding whether the atria have been remodeled leads to improvements in mapping and ablation strategies.

[0041] This disclosure presents a novel strategy for estimating how much remodeling exists in a fibrillated atrium.

[0042] In one embodiment of the invention, a probe comprising multiple electrodes is inserted into a human patient such that the electrodes contact cardiac tissue undergoing atrial fibrillation. A processor acquires signals from the electrodes and calculates a mutual information metric between each pair of electrodes on the probe. For a given electrode pair, the mutual information metric provides a numerical value for the mutual dependence of the signals in that pair. (For example, if the signals are independent of each other, the metric is close to zero.)

[0043] The processor generates a graph with electrodes as nodes and edges as the connections between them, exceeding a selected mutual information metric threshold. In one embodiment, the threshold is selected such that at least 50% of the electrodes are connected.

[0044] The processor calculates a local efficiency metric for each node, which is a measure of the number of effective nodes connected to the given node that exchange information. The processor then averages the local efficiency metrics of the graph to generate a resulting local efficiency metric at a selected mutual information metric threshold.

[0045] The processor repeats the steps of generating the graph and averaging the corresponding local efficiency measure while incrementing the selected mutual information metric threshold, to produce an ordered set of local efficiency and mutual information thresholds. The processor then analyzes this ordered set of pairs to classify atrial fibrillation.

[0046] This analysis typically includes estimating the percentage of remodeling in the heart tissue. Detailed Implementation

[0048] In the following description, similar elements in the accompanying drawings are identified by similar numbers, and similar elements may be distinguished as needed by appending letters after the identified numbers.

[0049] See now Figure 1 This is a schematic diagram of an atrial fibrillation (AF) classification system 20 according to an embodiment of the present invention. Figure 1A physician 22 is depicted using catheter 24 (also referred to herein as probe 24) to acquire unipolar or bipolar intracardiac ECG (electrocardiogram) signals from tissue of the heart 26 of a patient 28. Catheter 24 includes multiple ridges 30 at its distal end 31, these ridges being mechanically flexible, and two or more electrodes 32 are present on each ridge. The electrodes 32 are located on the ridges 30 such that the spacing between the nearest adjacent electrodes is less than 3 mm. Although... Figure 1 A catheter with five ridges is depicted, but embodiments of the invention include catheters with other numbers of ridges, as well as other catheters with multiple electrodes, wherein the nearest neighbor electrode is separated by less than 3 mm.

[0050] Electrode 32 is coupled to processor 36 via conductors and interface 34 in conduit 24. Processor 36 includes processing unit 42, typically a central processing unit (CPU), and referred to herein as CPU 42, which is coupled to memory 46. Memory 46 includes several modules: ECG module 50, tracking module 58, and AF analysis module 54. The functions of these modules are described below.

[0051] CPU 42 typically includes a general-purpose processor with software programmed to perform the functions described herein. For example, the software may be downloaded electronically via a network, or alternatively or otherwise, the software may be provided and / or stored on a non-transitory tangible medium such as magnetic storage, optical storage, or electronic storage.

[0052] The physician 22 communicates with the processor 36 via an input device 70 (such as a keypad or pointing unit) and a screen 74, and the processor 36 can display the results of the program executed by the processor on the screen.

[0053] In system 20, CPU 42 can use tracking module 58 to track the position of electrode 32. In one embodiment, the CPU and module are configured to implement an advanced current positioning (ACL) system in system 20. ACL systems are described in U.S. Patent 8,456,182. In an ACL system, the processor estimates the corresponding position of the distal electrode based on impedance or current measured between each distal electrode in distal electrode 32 and a plurality of surface electrodes 66 coupled to the skin of patient 28. For ease of illustration, Figure 1 Only one surface electrode 66 is shown. The processor can then correlate any electrophysiological signals received from the distal electrode 32 with the location of the acquired signal.

[0054] Alternatively or additionally, the CPU and module 58 are configured to use an electromagnetic tracking system to track the position of electrode 32, for example, for a device manufactured by Biosense Webster (Irvine, California). In this system, and as described in U.S. Patent Nos. 5,391,199, 5,433,489, and 6,198,963, one or more magnetic sensors 38 (typically uniaxial, biaxial, or triaxial coils) are attached to the distal end 31. Additionally, an assembly 60 of alternating magnetic field radiators is located near, and typically below, the patient 28. In response to the field from the assembly 60, a current generated in the sensor 38 is recorded by module 58, and the module and CPU analyze the current to determine the location of the distal end and the position of the electrode 32, since the geometry of the distal end is known or can be estimated.

[0055] Figure 2 This is a flowchart of the steps of an algorithm executed by processor 36 under the overall control of physician 22 according to one embodiment of the invention. In the initial step 100 of the flowchart, when the patient has atrial fibrillation, physician 22 inserts probe 24 into patient 28. Processor 36 uses module 58 to track the distal end 31 of the probe, and physician observes the tracking of the probe on screen 74 and manipulates the distal end until the probe electrode 32 contacts the desired tissue segment within the atrium (hereinafter assumed to include the left atrium of heart 26).

[0056] Typically, the tracking within the atrium is presented to the physician by processor 36, which overlays the probe icon onto a pre-acquired mapping 78 of the atrium displayed on screen 74. In addition to using the tracking provided by module 58, the physician can use processor 36 to measure the impedance between electrode 32 and return electrode 80 attached to the skin of patient 28, and to confirm contact between the electrode and atrial tissue by impedance measurement.

[0057] Once in the appropriate position, physician 22 operates probe 24 to acquire and record a set of unipolar signals from each electrode 32, i.e., a set of voltages measured relative to electrode 80. Typically, a set of signals is acquired over a period of approximately 2 minutes, i.e., for approximately 150 heartbeats; however, embodiments of the invention may acquire signals over shorter or longer periods. Typically, at least 30 seconds of atrial fibrillation (AF) should be recorded. CPU 42 stores the acquired signals in memory 46, and for each signal, CPU uses ECG module 50 to calculate the Local Activation Time (LAT). It should be understood that, because patient 28 has atrial fibrillation, unlike a heart with sinus rhythm, the LAT value of an electrode at a fixed location in the heart has a varying value; i.e., for 150 heartbeats, there may be 150 different LAT values.

[0058] In the first analysis step 104, CPU 42 uses AF analysis module 54 to classify a set of LAT values ​​for each electrode into bins of a histogram. Typically, to simplify the calculations described below, the histogram has bins of equal width. Embodiments of the invention can use histograms to represent the distribution of LAT values; in one embodiment, the number of bins is approximately [number missing]. Where L is the number of LAT values ​​in the set.

[0059] Figure 3 Two equal-width bin histograms according to an embodiment of the invention are shown, for electrodes 32X and 32Y, also referred to herein as electrodes X and Y. In the calculation performed by the CPU, it is assumed that the number of values ​​in a given bin gives the probability of the average LAT occurring in that bin. .

[0060] Return to Figure 2 In step 104 of the flowchart, module 54 estimates the mutual information metric of multiple electrode pairs 32 according to equation (1):

[0061]

[0062] Where N represents the number of bins in the histogram.

[0063] X is the first given electrode.

[0064] Y is the second given electrode.

[0065] and x i It represents the number of LAT values ​​in the i-th bin of the histogram of electrode X.

[0066] and y i It represents the number of LAT values ​​in the i-th bin of the histogram of electrode X.

[0067] L is the total number of values ​​in each histogram.

[0068]

[0069] In the following description, unless otherwise stated, equation (1) is used to estimate the mutual information metric of all electrode pairs 32. However, those skilled in the art will be able to adjust the description, making the necessary changes, for embodiments where the metric is estimated to be less than that of all electrode pairs 32, for example if some of the electrodes are not coupled to the processor 36, and it is assumed that all such embodiments are included within the scope of the invention.

[0070] Mutual information is a known metric in the art and is a measure of the dependence between two signals: that is, the amount of information about one signal obtained based on observations of the other. Therefore, if one signal is a deterministic function of the other, their mutual information is maximized; however, if the two signals are completely independent of each other, their mutual information approaches zero.

[0071] In pattern preparation step 108, CPU 42 uses module 54 to set a mutual information threshold such that a set percentage of electrode pairs have mutual information values ​​higher than this threshold, and then prepares a pattern, also referred to herein as a region information map, where nodes of the pattern represent electrodes 32, and edges between nodes represent connections equal to or exceeding the threshold. Typically, the threshold is initially set to include 50% of the electrode pairs.

[0072] Figure 4 Figure 82 is an exemplary schematic region information diagram prepared by CPU 42 according to an embodiment of the present invention. For example, in Figure 82, there are 48 nodes 84, corresponding to the number of electrodes 32 in the distal end 31 of the catheter. Edges 86 connect to nodes corresponding to connections equal to or greater than a mutual information threshold.

[0073] In subgraph step 112, a subgraph is generated for each given node in the region information graphics module 54. This subgraph includes a set of nodes connected to the given node under consideration. The CPU 42 calculates the shortest path between each pair of nodes in the set as the minimum number of connections between that pair, and for the calculation, the CPU assumes that any given connection has a unit length. The shortest path length is stored in memory 46.

[0074] Figure 5A and Figure 5B This is a schematic diagram of a region information diagram 82 according to an embodiment of the present invention, which shows a subgraph 88 of node 84A and a subgraph 90 of node 84B, respectively.

[0075] like Figure 5A As shown, node 84A has four connections or edges 86A, 86B, 86C, and 86D with nodes 84B, 84C, 84D, and 84E, respectively. Therefore, the subgraph 88 of node 84A includes nodes 84B, 84C, 84D, and 84E. This can be derived from... Figure 5A The checks determine the shortest path lengths between these node pairs, and are given in Table I:

[0076] 84B-84C 2 84B-84D 12 84B-84E 8 84C-84D 12 84C-84E 8 84D-84E 4

[0077] Table I

[0078] like Figure 5BAs shown, node 84H has three connections or edges 86J, 86K, and 86L with nodes 84J, 84K, and 84L, respectively. Therefore, the subgraph 90 of node 84H includes nodes 84J, 84K, and 84L. Compared to node 84A, there are no paths between nodes in the subgraph of node 84H.

[0079] In step 112, the reciprocal of the shortest path length is used to calculate the local efficiency metric E for each node according to equation (2). 局部 :

[0080]

[0081] Where N Gi It is the number of nodes in the subgraph Gi, and

[0082] L ij It is the shortest path length between nodes i and j in subgraph Gi.

[0083] Local efficiency has been used to evaluate the brain. It represents fault tolerance and indicates the extent to which each subgraph exchanges information when the central node is eliminated.

[0084] CPU 42 uses Equation (2) to calculate the local efficiency metric for each node in the region information graph. Therefore, for node 84A, Equation (2) uses the path length values ​​from Table I to give E. 局部 It is 0.0972.

[0085] For node 84H, since there are no paths between nodes in the subgraph, therefore E 局部 It is 0.

[0086] In averaging step 116, the calculated local efficiencies are averaged to provide the resulting local efficiency for the graph at a set mutual information threshold. The CPU 42 stores the resulting local efficiency and the corresponding mutual information threshold in memory 46 as an ordered pair.

[0087] As shown in decision step 120 and increment step 124, CPU 42 repeats steps 108, 112, and 116, incrementing the mutual information metric threshold at each iteration until the metric reaches its maximum value, typically around 95%. At each iteration, the CPU stores the resulting local efficiency and the corresponding mutual information threshold as ordered pairs, such that when the iteration terminates, there is a set of ordered pairs available to CPU 42.

[0088] In the result step 128, based on the set of ordered pairs generated in step 116, CPU 42 generates a graph showing the relationship between the obtained local efficiency and the mutual information threshold.

[0089] Figure 6A and Figure 6BThis is a schematic diagram illustrating the relationship between the local efficiency obtained according to an embodiment of the present invention and the mutual information threshold of catheters with different electrode spacings. Figure 6A In Figure 200, the simulation results of the rotational signal acquired by a catheter with a minimum electrode spacing of 1 mm are shown. Figure 6B In Figure 204, the simulation results of the rotation signal acquired by the catheter with a minimum electrode spacing of 3 mm are shown.

[0090] As shown in Figure 200, a plateau period 202 exists, where the slope of the figure decreases compared to the slope on either side of the plateau. No such plateau period is observed in Figure 204. Figures 200 and 204 show plateau regions present in the figures when the signal rotates when the electrode spacing is less than 3 mm. This plateau period does not exist when the electrode spacing is 3 mm or greater.

[0091] Figure 7A , Figure 7B and Figure 7C This is a schematic diagram illustrating the relationship between the local efficiency and the mutual information threshold of a catheter with a minimum electrode spacing of 2 mm according to an embodiment of the present invention. The graphs are simulation results of different percentages of tissue remodeling in the heart 26. Graph 210 is for 10% remodeling, graph 214 is for 50% remodeling, and graph 218 is for 90% remodeling. It is evident from the graphs that 10% remodeling does not produce a plateau phase, while 50% and 90% remodeling produce plateau phases 216 and 220, respectively.

[0092] To more clearly illustrate the existence of the plateau period, in step 128, CPU 42 fits a polynomial, typically a cubic polynomial of the generated graph, calculates the first derivative of the polynomial, and plots a graph of the relationship between the first derivative and the information threshold.

[0093] Figure 8 A schematic diagram illustrating the relationship between the first derivative of local efficiency generated from graphs 210, 214, and 218 and the mutual information threshold, according to an embodiment of the present invention, is shown. To generate Figure 8 The CPU 42 fits cubic polynomials to each of graphs 210, 214, and 218. The processing unit then calculates the first derivative of the local efficiency of each fitted polynomial and plots the relationship between the first derivative and the mutual information threshold.

[0094] Figure 224 shows the relationship between the local efficiency of the first derivative and the mutual information threshold of a 10% remodeled figure 210.

[0095] Figure 228 shows the relationship between the local efficiency of the first derivative and the mutual information threshold of a 50% reshaped figure 214; and

[0096] Figure 232 shows the relationship between the local efficiency of the first derivative and the mutual information threshold of 90% reshaped Figure 218.

[0097] As shown in the figure, the presence of a minimum value in the graph indicates a large or very large reshaping, while for a small amount of reshaping, such as 10%, there is no minimum value in the graph. Furthermore, the “clarity” of the minimum value (e.g., the radius of curvature at the minimum) indicates the degree of reshaping, and embodiments of the invention can measure the radius of curvature or some other clarity metric to assess the degree of modeling.

[0098] In step 128, a first derivative graph can be presented to the physician 22 on screen 74. Alternatively or additionally, CPU 42 can determine whether a minimum value exists in the first derivative graph, and if a minimum value exists, the CPU can measure its sharpness. From the determination, the CPU can present a conclusion such as "no significant reshaping," "intermediate reshaping," or "high reshaping" on screen 74.

[0099] It should be understood that in step 128, CPU 42 may not need to generate all the described physical graphs, and the CPU may generate data corresponding to some graphs. In other words, the CPU may only use the ordered pairs from step 116, and the above-described first derivative graphs and / or conclusions may be presented from the ordered pairs.

[0100] For simplicity, the above description assumes that the signal acquired by the electrodes is used to form the LAT. Those skilled in the art will be able to adjust the description, making the necessary changes, and use voltage, typically the unipolar or bipolar voltage of the signal or the action potential voltage, instead of the LAT.

[0101] It should be understood that the above embodiments are cited by way of example, and the invention is not limited to the specific content shown and described above. Rather, the scope of the invention includes combinations and sub-combinations of the various features described above, as well as variations and modifications thereof, which should be apparent to those skilled in the art upon reading the above description, and which are not disclosed in the prior art.

Claims

1. An information processing method, comprising: Signals are collected from multiple electrodes that are in contact with heart tissue undergoing atrial fibrillation; Calculate the corresponding mutual information metric between multiple electrode pairs based on the signal; Generate a graph with the electrodes as nodes and the edges as connections between the nodes, wherein the corresponding mutual information metric for the connection exceeds a selected mutual information metric threshold; Based on the path length between the connected nodes, a corresponding local efficiency metric is calculated for each node, which indicates the efficiency of information exchange between the node and other nodes connected to the node. The corresponding local efficiency metrics of the nodes are averaged to determine the resulting local efficiency for the selected mutual information metric threshold. as well as The obtained local efficiency and the selected mutual information metric threshold are analyzed to classify the atrial fibrillation based on an estimate of the percentage of cardiac tissue remodeling.

2. The method according to claim 1, wherein, The signal is a signal relating unipolar or bipolar voltage or action potential voltage to time.

3. The method according to claim 1, wherein, Calculating the corresponding mutual information metric between multiple electrode pairs based on the signal includes estimating the local activation time (LAT) based on the signal.

4. The method according to claim 1, wherein, The heart tissue is part of the atrium, and the classification of the atrial fibrillation includes estimating the percentage of atrial remodeling.

5. The method of claim 1, further comprising presenting the classification of the atrial fibrillation to a user of the method.

6. The method according to claim 1, wherein, The plurality of electrodes are located on a conduit with a plurality of ridges.

7. The method according to claim 1, wherein, The nearest neighbor electrode spacing among the plurality of electrodes is less than 3 mm.

8. The method according to claim 1, wherein, Averaging the corresponding local efficiency metrics for the nodes includes: generating a subgraph of nodes directly connected to a given node, calculating a local efficiency metric for each subgraph in the subgraph, and averaging the calculated local efficiency metrics.

9. The method of claim 1, further comprising the step of repeatedly generating the graph, and averaging the corresponding local efficiency metric while incrementing the selected mutual information metric threshold to generate an ordered set of resulting local efficiencies and mutual information thresholds.

10. The method of claim 9, further comprising analyzing the set of ordered pairs to classify the atrial fibrillation.

11. The method according to claim 10, wherein, Analyzing the set of ordered pairs includes fitting a polynomial to the set and classifying the atrial fibrillation in response to the first derivative of the polynomial.

12. The method according to claim 1, wherein, Calculating the corresponding mutual information metric between multiple electrode pairs based on the signal includes: calculating the corresponding mutual information metric between all electrode pairs based on the signal.

13. An information processing device, comprising: A probe having multiple electrodes configured to contact cardiac tissue undergoing atrial fibrillation; as well as Processor, the processor being configured to: Signals are received from the electrodes and corresponding mutual information metrics between multiple electrode pairs are calculated based on the signals. Generate a graph with the electrodes as nodes and the edges as connections between the nodes, wherein the corresponding mutual information metric for the connection exceeds a selected mutual information metric threshold; Based on the path length between the connected nodes, a corresponding local efficiency metric is calculated for each node, which indicates the efficiency of information exchange between the node and other nodes connected to the node. The corresponding local efficiency metrics of the nodes are averaged to determine the resulting local efficiency for the selected mutual information metric threshold. as well as The obtained local efficiency and the selected mutual information metric threshold are analyzed to classify the atrial fibrillation based on an estimate of the percentage of cardiac tissue remodeling.

14. The device according to claim 13, wherein, The signal is a signal relating unipolar or bipolar voltage or action potential voltage to time.

15. The device according to claim 13, wherein, Calculating the corresponding mutual information metric between multiple electrode pairs based on the signal includes estimating the local activation time (LAT) based on the signal.

16. The device according to claim 13, wherein, The heart tissue is part of the atrium, and the classification of the atrial fibrillation includes estimating the percentage of atrial remodeling.

17. The device according to claim 13, wherein, The processor is configured to present the classification of atrial fibrillation to the user of the device.

18. The device according to claim 13, wherein, The plurality of electrodes are located on a conduit with a plurality of ridges.

19. The device according to claim 13, wherein, The nearest neighbor electrode spacing among the plurality of electrodes is less than 3 mm.

20. The device according to claim 13, wherein, Averaging the corresponding local efficiency metrics for the nodes includes: generating a subgraph of nodes directly connected to a given node, calculating a local efficiency metric for each subgraph in the subgraph, and averaging the calculated local efficiency metrics.

21. The device of claim 13, further comprising the step of repeatedly generating the graph, and averaging the corresponding local efficiency metric while incrementing the selected mutual information metric threshold to produce an ordered set of resulting local efficiencies and mutual information thresholds.

22. The device of claim 21, further comprising analyzing the set of ordered pairs to classify the atrial fibrillation.

23. The device according to claim 22, wherein, Analyzing the set of ordered pairs includes fitting a polynomial to the set and classifying the atrial fibrillation in response to the first derivative of the polynomial.

24. The device according to claim 13, wherein, Calculating the corresponding mutual information metric between multiple electrode pairs based on the signal includes: calculating the corresponding mutual information metric between all electrode pairs based on the signal.

Images