Temporal mapping of thermal ablation areas

Abstract

Various embodiments of the present disclosure include a thermal ablation probability controller (30) that uses a trained ablation probability model (32) to render a pixel ablation probability for each pixel of ablation scan images showing static anatomical ablation. During operation, the thermal ablation probability controller (30) spatially aligns a time series of ablation scan datasets representing dynamic anatomical ablation and applies the ablation probability model (32) to the spatial alignment of the time series of ablation scan datasets to render a pixel ablation probability for each pixel of the ablation scan images showing static anatomical ablation.

Description

[Technical field]

[0001] The present disclosure relates generally to percutaneous thermal ablation and, more particularly, to imaging of a thermal ablation region for monitoring percutaneous thermal ablation. [Background technology]

[0002] Percutaneous thermal ablation is a widely used tool in the interventional radiologist's arsenal for anti-cancer treatment, with approximately 200,000 procedures performed worldwide in 2016, with a growth rate of over 8%. Precise ablation of tumors requires monitoring of the ablation procedure (i.e., visualization of the growing thermal ablation zone) while sparing nearby healthy tissue. Although ultrasound imaging is commonly used to guide needle placement for ablation, current ultrasound-based monitoring methods provide insufficient monitoring and intra-procedure evaluation of the ablation, which may result in incomplete tumor ablation, tumor recurrence, or potential damage to healthy vital structures near the ablation zone.

[0003] More specifically, ultrasound-based methods for ablation monitoring have been investigated and these methods have shown promise for delineating the ablation boundary (e.g., shear wave elastography). However, challenging imaging conditions during ablation can lead to significant artifacts and / or low signal-to-noise (SNR) ratios, such as the presence of the ablation electrode, tumors, and temperature gradients, in addition to general tissue heterogeneity and physiological motion that can lead to significant artifacts and / or low signal-to-noise (SNR) ratios.

[0004] Importantly, the relevant margins observed in ablation monitoring are the contours of areas with irreversible and permanent thermal tissue damage due to coagulation necrosis. The coagulation process is characterized by a significant increase in tissue hardness. During an ablation procedure, the coagulation zone grows from the area immediately surrounding the ablation electrode tip to a distance of several centimeters from the tip. Because ablation monitoring images can be subject to artifacts and low SNR, it can be difficult to distinguish the relevant borders of the ablation in individual images at individual time points. Thus, when viewing such individual images, it is difficult to identify the contours of the ablation area, which can lead to incomplete ablation, tumor recurrence, or damage to healthy vital structures. EP 2 640 292 A1 describes a system and method for ablation planning, which includes prescribing the shape and size of one or more ablation volumes and determining a target volume to be treated based on the probability of treatment. A treatment plan is provided by determining the number and location of planned ablations in the target volume using one or more ablation volumes. A joint probability distribution is determined for at least two planned ablations in the target volume. The final configuration is visualized and it is determined whether the planning goal is achieved based on the probability of treatment of the target volume. Summary of the Invention

[0005] Although known ultrasound-based methods for ablation monitoring (e.g., shear wave elastography) have proven beneficial, there remains a need for improved techniques to provide accurate and reliable monitoring of the ablation region via scan imaging of the ablated anatomical region. The present disclosure teaches a probabilistic approach for identifying the ablation region based on observation and tracking of individual ablation monitoring images.

[0006] More specifically, for purposes of the detailed description and claims of this disclosure, a scan image of an unchanging anatomical ablation at a discrete time is considered a static anatomical ablation, and a time series of scan images of an anatomical ablation that expands over a period of time is considered a dynamic anatomical ablation. This disclosure teaches a probabilistic approach for identifying ablation regions for a static anatomical ablation based on observing and tracking the dynamic anatomical ablation.

[0007] One embodiment of the present disclosure is a thermal ablation probabilistic controller that uses a memory containing an ablation probability model trained to render a pixel ablation probability for each pixel of an ablation scan image showing static anatomical ablation (i.e., unchanging anatomical ablation at discrete time).

[0008] The thermal ablation probability controller further uses one or more processors to spatially align a time series of ablation scan data sets representing dynamic anatomical ablation (i.e., anatomical ablation growing over a period of time) and apply the ablation probability model to the spatial alignment of the time series of ablation scan data sets to render a pixel ablation probability for each pixel of the ablation scan images showing static anatomical ablation.

[0009] A second embodiment of the present disclosure is a non-transitory machine-readable storage medium encoded with instructions for an ablation probability model executed by one or more processors to render a pixel ablation probability for each pixel of an ablation scan dataset representing static anatomical ablation (i.e., unchanging anatomical ablation at discrete time).

[0010] The non-transitory machine-readable storage medium includes instructions for spatially aligning a time series of ablation scan data sets representing dynamic anatomical ablation (i.e., an anatomical ablation that grows over a period of time) and applying an ablation probability model to the spatial alignment of the time series of ablation scan data sets to render a pixel ablation probability for each pixel of an ablation scan image showing a static anatomical ablation.

[0011] A third embodiment of the present disclosure is a thermal ablation probability method executable by a thermal ablation probability controller that includes an ablation probability model trained to render a pixel ablation probability for each pixel of an ablation scan dataset representing a static anatomical ablation (i.e., an unchanging anatomical ablation at discrete time).

[0012] The thermal ablation probability method involves spatial alignment, by a thermal ablation probability controller, of a time series of ablation scan data sets representing dynamic anatomical ablation (i.e., an anatomical ablation that grows over a period of time).

[0013] The thermal ablation probability method further includes applying, by the thermal ablation probability controller, the ablation probability model to the spatial alignment of the time series of ablation scan data sets to render a pixel ablation probability for each pixel of the ablation scan images showing the static anatomical ablation.

[0014] For various embodiments of the present disclosure, applying the ablation probability model to the spatial alignment of the time series of ablation scan data sets may include temporal pixel intensities of pixels of the ablation scan images indicative of static anatomical ablation derived from the spatial alignment of the time series of ablation scan image data data.

[0015] Additionally, applying the ablation probability model to the spatial alignment of the time series of ablation scan datasets may further include temporal pixel intensities of one or more adjacent pixels of the ablation scan images indicative of static anatomical ablation derived from the spatial alignment of the time series of ablation scan datasets.

[0016] Furthermore, applying the ablation probability model to the spatial alignment of the time series of ablation scan datasets may further include applying an ablation probability rule to the temporal pixel intensities of the pixel and adjacent pixels to render a pixel ablation probability for the pixel.

[0017] For various embodiments of the present disclosure, applying the ablation probability model to the spatial alignment of the time series of ablation scan data sets may include temporal pixel intensity evaluation of the time series of each pixel of the spatial alignment of the time series of ablation scan image data by the ablation probability model.

[0018] Further, applying the ablation probability model to the spatial alignment of the time series of ablation scan data sets may further include spatial pixel intensity estimation by the ablation probability model of temporal pixel intensity estimation of one or more adjacent pixels of the spatial alignment of the time series of ablation scan image data.

[0019] Additionally, applying the ablation probability model to the spatial alignment of the time series of ablation scan datasets may further include applying one or more ablation probability rules to the temporal pixel intensity estimates and the spatial pixel intensity estimates, the ablation probability rules being based on a comparison of the temporal pixel intensity estimates of the pixel and adjacent pixels to an ablation threshold.

[0020] In embodiments of the present disclosure, an ablation probability image derived from pixel ablation probability for each pixel of an ablation scan image showing static anatomical ablation may be generated and communicated (e.g., displaying and / or printing a black and white or grayscale or color-coded image of potential ablation areas of an ablated anatomical region).

[0021] Alternatively or simultaneously, an ablation probability anatomical scan image derived from the anatomical image of the static anatomical ablation and further derived from pixel ablation probabilities for each pixel of the ablation scan image showing the static anatomical ablation may be generated and communicated (e.g., displaying and / or printing an overlay of potential ablation regions on the anatomical scan image of the ablated anatomical region).

[0022] Alternatively or simultaneously, an ablation probability ablation scan image derived from the ablation scan image showing the static anatomical ablation and further derived from the pixel ablation probability for each pixel of the ablation scan image showing the static anatomical ablation may be generated and communicated (e.g., displaying and / or printing an overlay of potential ablation areas on a displayed and / or printed ablation ultrasound scan image).

[0023] For purposes of the detailed description and claims of this disclosure: (1) Technical terms including, but not limited to, “scan,” “ablation,” “static anatomical ablation,” “dynamic anatomical ablation,” “pixels,” “dataset,” “spatial alignment,” and “probabilistic model” should be construed as known in the art of this disclosure and as exemplary described in this disclosure. (2) More specifically, the term "pixel" encompasses both planar pixels and volumetric pixels (also known as voxels). (3) The term "controller" broadly encompasses all structural configurations of a main circuit board or integrated circuit for controlling the application of the various inventive principles of the present disclosure as understood in the art and as illustrative examples described herein. The structural configuration of a controller may include, but is not limited to, a processor, a computer usable / computer readable storage medium, an operating system, application modules, peripheral controllers, slots, and ports. The controller may be housed within a workstation or linked to a workstation. Examples of a "workstation" include, but are not limited to, an assembly of one or more computing devices, a display / monitor, and one or more input devices (e.g., keyboard, joystick, and mouse) in the form of a standalone computing system, a client computer of a server system, a desktop, or a tablet. (4) The term “application module” broadly encompasses an application embedded within or accessible by a controller that is comprised of electronic circuitry (e.g., electronic components and / or hardware) and / or executable programs (e.g., executable software stored on non-transitory computer-readable media and / or firmware) for executing a particular application. (5) The term "data" broadly encompasses all forms of detectable physical quantities or impulses (e.g., voltage, current, or magnetic field strength) as understood in the art of this disclosure and illustratively described herein for transmitting information and / or instructions that aid in applying the various inventive principles of this disclosure, as described later in this disclosure. The various components of data communication in this disclosure may include any communication method known in the art of this disclosure, including but not limited to data transmission / reception over any type of wired or wireless data link, and reading data uploaded to a computer usable / computer readable storage medium.

[0024] The foregoing and other embodiments of the presently disclosed invention, as well as various structures and advantages of the presently disclosed invention, will become more apparent from the following detailed description of various embodiments of the presently disclosed invention, read in connection with the accompanying drawings, in which: The detailed description and drawings are merely illustrative of the presently disclosed invention, rather than limiting, the scope of the presently disclosed invention being defined by the appended claims and their equivalents. [Brief description of the drawings]

[0025] [Figure 1] FIG. 1 illustrates an exemplary embodiment of a thermal ablation probability system according to various aspects of the present disclosure. [Diagram 2] FIG. 1 illustrates an exemplary embodiment of a flow chart depicting a thermal ablation probability method according to various aspects of the present disclosure. [Diagram 3] FIG. 1 illustrates an example spatial alignment of a time series of ablation scan data sets in accordance with various aspects of the present disclosure. [Figure 4] FIG. 1 illustrates an exemplary embodiment of a flow chart depicting a method for applying an ablation probability model in accordance with various aspects of the present disclosure. [Diagram 5] FIG. 1 illustrates an exemplary embodiment of a probabilistic decision tree according to various aspects of the present disclosure. [Figure 6A] 1A-1C illustrate an exemplary generation of a time series of ablation scan images and ablation probability images according to various aspects of the present disclosure. [Figure 6B] 1A-1C illustrate an exemplary generation of a time series of ablation scan images and ablation probability images according to various aspects of the present disclosure. [Figure 6C] 1A-1C illustrate an exemplary generation of a time series of ablation scan images and ablation probability images according to various aspects of the present disclosure. [Figure 6D] 1A-1C illustrate an exemplary generation of a time series of ablation scan images and ablation probability images according to various aspects of the present disclosure. [Figure 7] 1A-1C illustrate exemplary generation of an anatomical scan image and an ablation probability anatomical scan image in accordance with various aspects of the present disclosure. [Figure 8] 1A-1C illustrate exemplary generation of an ablation scan image and an ablation probability ablation scan image in accordance with various aspects of the present disclosure. [Figure 9] FIG. 1 illustrates an exemplary embodiment of an image scanner and thermal ablation probability controller in accordance with various aspects of the present disclosure. DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS

[0026] To facilitate understanding of the various inventions of the present disclosure, the following description of Figures 1 and 2 teaches exemplary thermal ablation probability systems and methods of the present disclosure. From this description, those skilled in the art will understand how to apply the inventive principles of the present disclosure to form and use additional embodiments of the thermal ablation probability systems and methods of the present disclosure.

[0027] With reference to FIG. 1 , the thermal ablation probability system of the present disclosure uses an anatomical scanner 20 to scan an ablated anatomical region 10 (e.g., a cranial region or a thoracic region) to generate a two-dimensional (2D) scan data set of pixels representing static anatomical ablation of the ablated anatomical region 10 and / or a three-dimensional ("3D") scan data set of volumetric pixels (a.k.a. voxels) representing static anatomical ablation of the ablated anatomical region 10.

[0028] More specifically, the generated 2D / 3D scan data sets represent unchanging anatomical ablation of the ablated anatomical region 10 at discrete times, which is considered a static anatomical ablation. The generated 2D / 3D scan data sets represent the anatomical ablation growing over a period of time, which is considered a dynamic anatomical ablation. In essence, the dynamic anatomical ablation is composed of a time series of static anatomical ablation.

[0029] In fact, the 2D / 3D anatomical scanner 20 can be any type of anatomical scanner for visualizing characteristics of the ablated anatomical region 10 that distinguish the ablated region from the surrounding non-ablated regions. Examples of such characteristics include, but are not limited to, tissue hardness, tissue density, and tissue temperature.

[0030] In one embodiment, the 2D / 3D anatomical scanner 20 is an ultrasound scanner for generating a 2D or 3D anatomical scan data set 21 (e.g., B-mode ultrasound imaging) representing a scanned image of the ablated anatomical region 10, excluding any type of depiction of the ablated region. Alternatively or simultaneously, the ultrasound scanner can generate a 2D or 3D ablation scan data set 22 (e.g., ultrasound shear wave elastography imaging) representing a scanned image of the ablated anatomical region 10, including a depiction of the ablated region. Examples of such anatomical scanners 20 include, but are not limited to, the ultrasound scanners incorporated within the Philips PercuNav and Philips UroNav systems.

[0031] 1, the thermal ablation probability system of the present disclosure further uses a display processor 40 to generate an anatomical scan image 41 from the 2D / 3D anatomical scan data set 21 of static anatomical ablation (e.g., B-mode ultrasound imaging) or from a time series of 2D / 3D anatomical scan data sets 22 of dynamic anatomical ablation (e.g., temporal B-mode ultrasound imaging). Alternatively or simultaneously, the display processor 40 can generate an ablation scan image 42 from the 2D / 3D ablation scan data set 22 of static anatomical ablation (e.g., ultrasound shear wave elastography imaging) or from a time series of 2D / 3D ablation scan data sets 22 of dynamic anatomical ablation (e.g., temporal ultrasound shear wave elastography imaging).

[0032] In fact, the display processor may be any type of display processor for converting pixels / voxels of a scan data set into display pixels / voxels, examples of such display processors include, but are not limited to, the display processors incorporated in the Philips PercuNav and Philips UroNav systems.

[0033] 1, the thermal ablation probability system of the present disclosure further employs the thermal ablation probability controller 30 of the present disclosure. To apply a probabilistic approach to identify ablation regions of static anatomical ablation based on observing and tracking dynamic anatomical ablation, the thermal ablation probability controller 30 includes application modules in the form of a spatial mapper 31, an ablation probability model 32, and an image data generator 33, which execute the thermal ablation probability method of the present disclosure, as represented by the flowchart 50 shown in FIG.

[0034] 2, during stage S52 of flowchart 50, spatial mapper 31 is configured to spatially align the time series of ablation scan data sets 22, thereby spatially aligning the temporal pixel intensities of each pixel of ablation scan data set 22. Spatial mapper 31 is further configured to buffer the spatially aligned time series of ablation scan data sets 37.

[0035] For example, FIG. 3 illustrates eight ablation scan data sets 22 along the Z axis of a coordinate system 60 defined by a baseline ablation scan data set. f1 ~22 f8 6 shows a spatial alignment 61 of the time series of the first ablation scan data set 22 of the time series on its Z axis. f1 Or there may be a first ablation scan data set 22 generated by the anatomical scanner 20. f1 ~22 f8 The time series of is registered to coordinate system 60 so that the temporal pixel intensity of each pixel 23 is aligned for processing by ablation probability model 32.

[0036] In practice, the spatial mapper 31 can perform spatial alignment using image registration techniques known in the art of this disclosure, which can be augmented with user input (which can be improved with mechanical controls of the anatomical scanner 20 (e.g., fixation of the ultrasound probe)) and / or with respiratory motion gating techniques known in the art of this disclosure.

[0037] In practice, spatial mapper 31 may also be augmented with a spatial tracking system (or active or passive mechanical holder) for anatomical scanner 20 (FIG. 1) to prevent or compensate for movement of anatomical scanner 20, leaving residual physiological tissue movement that is corrected using image-based methods. A spatial tracking system may also be utilized by spatial mapper 31 to maintain a constant imaging field of view by providing feedback if anatomical scanner 20 is moved from an initial baseline imaging position.

[0038] More practically, the buffer size depends on the frame rate of the anatomical scanner 20 and can be adjusted within a size range covering a time span between 1 second and 1 minute.

[0039] Returning to Figure 2, during stage 54 of flowchart 50, ablation probability model 32 is trained to render a pixel ablation probability for each pixel of ablation scan images 42 (Figure 1) showing static anatomical ablation. To this end, ablation probability model 32 retrieves the time series of spatially aligned ablation scan datasets 37 from the buffer and renders a pixel ablation probability for each pixel of ablation scan images 42 corresponding to ablation scan data that includes or is associated with the time series of spatially aligned ablation scan datasets 37.

[0040] In practice, the ablation probability model 32 is any statistical analysis tool trained to estimate the probability that which regions of the ablation scan images 42 are ablation regions based on a time series of the ablation scan images 42. The training of the ablation probability model 32 can be trained from clinical data or simulated data to match observed tissue ablation behavior to general tissue types or to different tissue types. Alternatively or simultaneously, the ablation probability model 32 can be trained with different adaptations to the tissue type being ablated, different adaptations to the ablation electrode, and / or different adaptations to the type of anatomical scanner 20.

[0041] In one embodiment, the ablation probability model 32 defines a set of rules that describe the likelihood that a given image pixel is part of an ablation region in the ablation scan image 42, given the time history of image intensity at that pixel and in a neighborhood of N pixels (N>=1) around the pixel. In this embodiment, Figure 4 shows a flowchart 70 that represents one embodiment of a method of applying the ablation probability model of the present disclosure.

[0042] Referring to FIG. 4, stage S72 of flowchart 70 includes: (X、Y) Temporal pixel intensity of 24 (X、Y) For example, in the case of ultrasound shear wave elastography, each pixel in the ablation scan dataset 37 has an intensity range between 10 blue intensities representing non-ablated tissue stiffness, non-ablated tissue density, or non-ablated tissue temperature, and 50 red intensities representing fully ablated tissue stiffness, fully ablated tissue density, or fully ablated tissue temperature.

[0043] In one embodiment of stage S72, the temporal pixel intensity 24 (X、Y) is calculated for each pixel in the time series of a spatially aligned ablation scan dataset. (X、Y) is the intensity average of the temporal pixel intensities of

[0044] In a second embodiment of stage S72, the temporal pixel intensity 24 (X、Y) is calculated for each pixel in the time series of a spatially aligned ablation scan dataset. (X、Y) is the median intensity value of the temporal pixel intensities of

[0045] In a third embodiment of stage S72, the temporal pixel intensity 24 (X、Y) is calculated for each pixel in the time series of a spatially aligned ablation scan dataset. (X、Y) is the maximum intensity value of the temporal pixel intensities of

[0046] In a fourth embodiment of stage S72, the temporal pixel intensity 24 (X、Y) is calculated for each pixel in the time series of a spatially aligned ablation scan dataset. (X、Y) is the minimum intensity value of the temporal pixel intensities of

[0047] In a fifth embodiment of stage S72, the temporal pixel intensity 24 (X、Y) is calculated for each pixel in the time series of a spatially aligned ablation scan dataset. (X、Y) is an intensity value derived from the standard deviation of the temporal pixel intensities of

[0048] With continued reference to FIG. 4, stage S74 of flowchart 70 involves the ablation probability model 32 being implemented as an assessment 75 of the temporal pixel intensities 24 in the spatial neighborhood of each pixel 23.

[0049] In practice, the spatial neighborhood of each pixel 23 depends on the (X,Y) location of the given pixel 23 and the number of neighboring pixels 23 that constitute the pixels 23 adjacent to the given pixel 23 .

[0050] For example, given pixel 23 (X、Y) 8 pixels surrounding 23 (X-1、Y+1) , Pixel 23 (X、Y+1) , Pixel 23 (X+1、Y+1) , Pixel 23 (X-1、Y) , Pixel 23 (X+1、Y) , Pixel 23 (X-1、Y-1) , Pixel 23 (X、Y-1) , and pixel 23 (X+1、Y-1) One or more of these neighboring pixels 23 are displayed as pixels 23 (X、Y) Furthermore, the pixel 23 is adjacent to eight pixels 23. (X、Y) One or more pixels (not shown) that are not adjacent to pixel 23 (X、Y) may be contained within a spatial vicinity of

[0051] Pixel 23(X、Y) In rendering a spatial neighborhood of pixels 23, the evaluation 75 involves grouping the temporal pixel intensities 24 of the pixels 23.

[0052] Continuing to refer to FIG. 4, stage S76 of flowchart 70 involves the ablation probability model 32 calculating temporal ablation probability 77, spatial ablation probability 78, and pixel ablation probability 79 according to the trained probability rules.

[0053] The probability rule for temporal ablation probability 77 is (X、Y) is given pixel 23 during stage S72. (X、Y) provides a preliminary estimate of the probability that a pixel represents tissue to be ablated as a function of its temporal pixel intensity valuation.

[0054] The probability rule for spatial ablation probability 78 is (X、Y) is given pixel 23 during stage S74. (X、Y) provides a complementary estimate of the probability of representing tissue to be ablated as a function of the spatial pixel intensity assessment of the

[0055] The probability rule for pixel ablation probability 79 is (X、Y) is given pixel 23 (X、Y) provides a final estimate of the probability of representing tissue being ablated as a function of both the temporal ablation probability 77 and the spatial ablation probability 78 of

[0056] In one embodiment, a probabilistic decision tree 80 as shown in FIG. 5 implements the probability rules for ablation probabilities 77-79.

[0057] Referring to FIG. 5, node 81 is a node for a given pixel 23 (X、Y) Temporal pixel intensity assessment of 24 (X、Y) is compared to an ablation intensity threshold (eg, 48 for ultrasound shear wave elastography) that indicates the stiffness of the tissue to be ablated, the density of the tissue to be ablated, or the temperature of the tissue to be ablated.

[0058] At node 81, given pixel 23 (X、Y) Temporal pixel intensity assessment of 24 (X、Y) If ABT is greater than the ablation intensity threshold, then node 82a selects the ablation intensity threshold for the given pixel 23. (X、Y) Determine whether the number of neighboring pixels for is equal to a neighborhood ablation threshold NABT1. For example, the neighborhood ablation threshold NABT1 is 100% of neighboring pixels having a temporal pixel intensity estimate 24 greater than the ablation strength threshold ABT.

[0059] At node 82a, a given pixel 23 (X、Y) If the number of neighboring pixels of a given pixel 23 is equal to the neighborhood ablation threshold NABT1, then node 83a (X、Y) We consider,to represent 100% pixel ablation probability of the ablated tissue.

[0060] At node 82a, a given pixel 23 (X、Y) If the number of neighboring pixels of a given pixel 23 is less than the neighborhood ablation threshold NABT1, then node 83b (X、Y) is considered to represent a pixel ablation probability of M% of the tissue being ablated, depending on the temporal pixel intensity estimates 24 of all the pixels involved, where M%>=50% and M%<100%.

[0061] Conversely, at node 81, given pixel 23 (X、Y) Temporal pixel intensity assessment of 24 (X、Y) If ABT is less than or equal to the ablation intensity threshold ABT, then node 82b selects the ablation intensity threshold ABT for the given pixel 23. (X、Y) Determine whether the number of neighboring pixels having a temporal pixel intensity estimate 24 greater than the ablation strength threshold ABT is equal to the neighborhood ablation threshold NABT2. For example, the neighborhood ablation threshold NAB2 is 0% of neighboring pixels having a temporal pixel intensity estimate 24 greater than the ablation strength threshold ABT.

[0062] At node 82b, given pixel 23 (X、Y) If the number of neighboring pixels of a given pixel 23 is greater than the neighborhood ablation threshold NABT2, then node 83c (X、Y)is considered to represent the pixel ablation probability of N% of the tissue to be ablated, depending on the temporal pixel intensity evaluation 24 of all the pixels involved, where N%>=0% and N%<100%.

[0063] At node 82b, given pixel 23 (X、Y) Node 83d (then node 83a) determines whether a given pixel 23 has a neighboring pixel count equal to the neighborhood ablation threshold NABT2. (X、Y) We consider,to represent a pixel ablation probability of 0% for ablated tissue (i.e., non-ablated tissue).

[0064] To arrive at a particular percentage probability of tissue being ablated, nodes 83b and 83c may implement additional probability rules.

[0065] For example, the probability rule for node 83c is (X、Y) is adjacent to neighboring pixels with a pixel ablation probability of 100%, and a given pixel 23 (X、Y) Temporal pixel intensity assessment of 24 (X、Y) A given pixel 23 is ablated if it is greater than 80% of the ablation strength threshold ABT. (X、Y) The probability of ablation of a pixel can be considered to be 50%.

[0066] By way of further example, the probability rule for node 83c is (X、Y) is adjacent to neighboring pixels with a pixel ablation probability of 100%, and a given pixel 23 (X、Y) Temporal pixel intensity assessment of 24 (X、Y) A given pixel 23 is ablated if it is greater than 90% of the ablation strength threshold ABT. (X、Y) This may increase the pixel ablation probability.

[0067] By way of further example, the probability rule for node 83c is (X、Y) is less than 100% pixel ablation probability and is adjacent to neighboring pixels greater than 0% pixel ablation intensity, and a given pixel 23 (X、Y) Temporal pixel intensity assessment of 24(X、Y) A given pixel 23 is ablated if ABT is less than 80% of the ablation strength threshold ABT. (X、Y) This may reduce the pixel ablation probability.

[0068] In practice, the probabilistic decision trees of the present disclosure may be defined with different adaptations to the tissue type being ablated, different adaptations to the ablation electrodes, and / or different adaptations to the anatomical scanner 20.

[0069] In one embodiment, the threshold and pixel ablation probability may be calibrated based on test ablations on different tissue types.

[0070] In a second embodiment, a spatio-temporal model of the expected development of the ablation region (e.g., derived from finite element modeling) can be integrated by fitting an ablation probability model 32 to the time history of the ablation scan images. (X、Y) The probability of ablation at a given pixel 23 (X、Y) It can be defined as a combination of the probability of originating from the image combined with a quality metric (e.g., the DICE coefficient) of a model that fits closely to

[0071] In a third embodiment, a spatiotemporal model of the expected evolution of the ablation region can be used to determine the threshold of the ablation probability model 32 by matching the ablation region calculated during the early stages of ablation with image values ​​obtained during those time points.

[0072] In a fourth embodiment, the threshold and pixel ablation probability can be learned with machine learning techniques (e.g., convolutional neural networks) by training the ablation probability model 32 on well-controlled experiments (with ground truth validation of the ablation zone using a reference imaging modality) or simulations of ablations imaged with a given imaging modality.

[0073] Returning to FIG. 2, during stage S54 of flowchart 50, image data generator 33 is configured to generate image data representing pixel ablation probabilities for each pixel rendered by ablation probability model 32 during stage S54.

[0074] In practice, the image data generator 33 assigns a unique color level (e.g., white) to each pixel with a pixel ablation probability of 100% and a unique color level (e.g., black) to each pixel with a pixel ablation probability of 0%. All other pixels either have the same unique color level as all pixels with a pixel ablation probability of 0% or have a grayscale color depending on their associated pixel ablation probability.

[0075] Also in practice, the image data generator 33 can generate an ablation probability image dataset 34, which enables the display processor 40 to generate a black / white or grayscale ablation probability image 43. The ablation probability image dataset 34 represents the contours of the ablation region in a probability map at a particular probability level (e.g., 100%).

[0076] For example, Figures 6A-6D show a series of ultrasound shear wave elastography images 42a with ablation regions 90 and corresponding ablation probability images 43a with ablation regions 91 at t=5 minutes as shown in Figure 6A, t=10 minutes as shown in Figure 6B, t=15 minutes as shown in Figure 6C, and t=20 minutes as shown in Figure 6D.

[0077] With reference to Figures 6A-6D, it is important to note how the ablation probability model 32 defines the ablation 90 growing in the ultrasound shear wave elastography image 42a without including the artifact 92 that appears temporarily under the ablation region 90 at t=10 minutes and t=15 minutes (the artifact disappears again at t=20 minutes after imaging conditions improve).

[0078] 2, alternatively or simultaneously, the image data generator 33 may generate an ablation probability anatomical overlay dataset 35 such that the display processor 40 may generate an overlay of probable ablation regions on an anatomical scan image 44 (e.g., a B-mode ultrasound image). The ablation probability image overlay dataset 35 represents the contours of ablation regions at a particular probability level (e.g., 100%) in the anatomical scan image.

[0079] For example, FIG. 7 shows anatomical image 41 a and anatomical image 44 a with an ablation overlay 93 rendered by ablation probability model 32 .

[0080] 2, alternatively or simultaneously, the image data generator 33 may generate an ablation probability anatomical overlay dataset 36 such that the display processor 40 may generate an overlay of potential ablation regions on the anatomical scan image 44. The ablation probability image overlay dataset 36 represents the contours of ablation regions at a particular probability level (e.g., 100%) in the ablation scan image.

[0081] For example, FIG. 7 shows ablation image 42 a and ablation image 45 a with an ablation overlay 92 rendered by the ablation probability model 32 .

[0082] Returning to FIG. 2, flow chart 50 is executed in a loop during the ablation procedure.

[0083] In practice, the spatial mapper 31 operated in a mode that controlled the update frequency of the ablation probability for each pixel.

[0084] In one embodiment, the spatial mapper 31 operates under the latest ablation scan dataset 32 ​​in an out of buffer mode / oldest ablation scan dataset 32, so that each scan of the ablated anatomical region 10 results in an updated ablation probability for each pixel.

[0085] In a second embodiment, the spatial mapper 31 operates under a clear buffer / fill buffer mode, whereby each Nth scan of the ablated anatomical region 10 results in an updated ablation probability for each pixel.

[0086] To facilitate a further understanding of the various inventions of the present disclosure, the following description of Figure 9 teaches an exemplary embodiment of a thermal ablation probability controller of the present disclosure. From this description, one of ordinary skill in the art will understand how to apply various aspects of the present disclosure to form and use additional embodiments of the thermal ablation probability controller of the present disclosure.

[0087] Referring to FIG. 9, the thermal ablation probability controller 130 includes one or more processors 131, memory 132, a user interface 133, a network interface 134, and a storage device 135 interconnected via one or more system buses 136.

[0088] Each processor 131 may be any hardware device known in the art of this disclosure or contemplated below that is capable of executing instructions or storing or otherwise processing data stored in memory 132. In non-limiting examples, processor 131 may include a microprocessor, a field programmable gate array (FPGA), an application specific integrated circuit (ASIC), or other similar device.

[0089] Memory 132 may include various memories as known in the art of this disclosure or as contemplated below, including, but not limited to, L1, L2, or L3 cache or system memory. In non-limiting examples, memory 132 may include static random access memory (SRAM), dynamic RAM (DRAM), flash memory, read only memory (ROM), or other similar memory devices.

[0090] The user interface 133 may include one or more devices known in the art of the present disclosure or contemplated below for enabling communication with a user, such as an administrator. In non-limiting examples, the user interface may include a command line interface or a graphical user interface that may be presented to a remote terminal via the network interface 134.

[0091] The network interface 134 may include one or more devices known in the art of the present disclosure or contemplated below to enable communication with other hardware devices. In a non-limiting example, the network interface 134 may include a network interface card (NIC) configured to communicate according to an Ethernet protocol. Additionally, the network interface 134 may implement a TCP / IP stack for communication according to a TCP / IP protocol. Various alternative or additional hardware or configurations for the network interface 134 will be apparent.

[0092] The storage device 135 may include one or more machine-readable storage media, including, but not limited to, read-only memory (ROM), random access memory (RAM), magnetic disk storage media, optical storage media, flash memory devices, or similar storage media, as known in the art of this disclosure or as contemplated below. In various non-limiting embodiments, the storage device 135 may store instructions for execution by the processor 311 or may store data on which the processor 131 may operate. For example, the storage device 135 may store a basic operating system for controlling various basic operations of the hardware. The storage device 135 also stores application modules in the form of executable software / firmware for implementing various functions of the controller 130a as previously described in this disclosure, including, but not limited to, the spatial mapper 137a, the ablation probability model 137b, and the image data generator 137c as previously described in this disclosure.

[0093] The controller 130 is located in a stand-alone workstation 100 using a monitor 101, a keyboard 102, a computer 103, and ultrasound probes 104 and 105. A display processor (not shown) and a scanning processor (not shown) are also located in the computer. Alternatively, the controller 130 may be located in a server whereby the controller is accessible via a client workstation or a mobile device (e.g., a tablet).

[0094] With reference to Figures 1-9, one of ordinary skill in the art to which the present disclosure pertains will appreciate the many advantages of the presently disclosed invention, including, but not limited to, a probabilistic approach for identifying ablation regions based on observing and tracking individual ablation monitoring images over time.

[0095] Furthermore, as one skilled in the art will appreciate in view of the teachings provided herein, the structures, elements, components, etc. described in this disclosure / specification and / or illustrated in the figures may be implemented in various combinations of hardware and software to provide functionality that may be combined into a single element or multiple elements. For example, the functionality of the various structures, elements, components, etc. shown / illustrated / illustrated in the figures may be provided by using dedicated hardware as well as hardware capable of executing software in association with appropriate software for additional functionality. When provided by a processor, functionality may be provided by a single dedicated processor, a single shared processor, or multiple individual processors, some of which may be shared and / or multiplexed. Furthermore, the explicit use of the terms "processor" or "controller" should not be construed to refer only to hardware capable of executing software, but may imply including, but not limited to, digital signal processor (DSP) hardware, memory (e.g., read only memory (ROM) for storing software, random access memory (RAM), non-volatile storage, etc.), and virtually any means and / or machine (hardware, software, firmware, combinations thereof, etc.) capable of (and / or configurable) executing and / or controlling a process.

[0096] Moreover, all statements herein reciting principles, aspects, and embodiments of the invention, as well as specific examples thereof, are intended to encompass both structural and functional equivalents thereof. Moreover, such equivalents are intended to include both currently known equivalents and equivalents developed in the future (e.g., elements developed that can perform the same or substantially similar functions, regardless of structure). Thus, for example, in light of the teachings provided herein, one of ordinary skill in the art will appreciate that any block diagrams presented herein may represent conceptual views of illustrative system components and / or circuits embodying the principles of the invention. Similarly, in light of the teachings provided herein, one of ordinary skill in the art should appreciate that any flow charts, flow diagrams, etc. may represent various processes that may be substantially represented on a computer-readable storage medium and executed by a computer, processor, or other device having processing capabilities, whether or not such a computer or processor is explicitly shown.

[0097] Although preferred exemplary embodiments of the various numerous inventions of the present disclosure have been described, which embodiments are intended to be illustrative and not limiting, it should be noted that modifications and variations can be made by those skilled in the art in light of the teachings provided herein, including the figures. Thus, it should be understood that changes can be made to the preferred exemplary embodiments of the present disclosure that are within the scope of the embodiments disclosed herein.

[0098] Additionally, corresponding and / or related systems that incorporate and / or implement the devices / systems or that may be used / implemented within / with the devices in accordance with the present disclosure are also contemplated and considered to be within the scope of the present disclosure. Additionally, corresponding and / or related methods for making and / or using the devices and / or systems according to the present disclosure are also contemplated and considered to be within the scope of the present disclosure.

Claims

1. A thermal ablation probability controller, the thermal ablation probability controller comprising: a memory including an ablation probability model trained to render a pixel ablation probability for each pixel of an ablation scan image showing an anatomical ablation; at least one processor in communication with the memory; The at least one processor comprises: spatially aligning a time series set of ablation scan data representing dynamic anatomical ablation; applying the ablation probability model to a spatial alignment of the time series set of ablation scan data to render the pixel ablation probability for each pixel of the ablation scan images indicative of the anatomical ablation; The time series set of scan data representing the dynamic anatomical ablation is a time series data set of images showing an expanding ablation region in a plurality of frame images intermittently captured from a specific site over a period of time from a specific time. Thermal cauterization probability controller.

2. Applying the ablation probability model to the spatial alignment of the time sequence sets of ablation scan data includes: deriving temporal pixel intensities for pixels of the ablation scan images indicative of the anatomical ablation from the spatial alignment of the time series sets of ablation scan data; deriving a temporal pixel intensity of at least one adjacent pixel of the ablation scan image indicative of the anatomical ablation from the spatial alignment of the time series set of ablation scan data; and 2. The thermal ablation probability controller of claim 1, further comprising the at least one processor configured to apply an ablation probability rule to the temporal pixel intensity of the pixel and the temporal pixel intensity of the at least one neighboring pixel to render the pixel ablation probability for the pixel.

3. Applying the ablation probability model to the spatial alignment of the time sequence sets of ablation scan data includes:

2. The thermal ablation probability controller of claim 1, further comprising the at least one processor configured to perform a time series temporal pixel intensity evaluation of each pixel of the spatial alignment of the time series set of the ablation scan data with the ablation probability model.

4. The temporal pixel intensity estimate for each pixel of the time series set of ablation scan data comprises: an intensity average of each pixel of the spatial alignment of the time series set of ablation scan data; a median intensity value of each pixel of the spatial alignment of the time series set of ablation scan data; a maximum intensity value of each pixel in the spatial alignment of a time series set of ablation scan data; a minimum intensity value of each pixel of the spatial alignment of the time series set of ablation scan data; and 10. The thermal ablation probability controller of claim 1, wherein the intensity value derived from the standard deviation of the time series of each pixel of the spatial alignment of the time series set of the ablation scan data.

5. Applying the ablation probability model to the spatial alignment of the time sequence sets of ablation scan data includes:

4. The thermal ablation probability controller of claim 3, further comprising the at least one processor configured to perform a spatial pixel intensity estimation of the or each temporal pixel intensity estimation of at least one adjacent pixel of the spatial alignment of the time series set of the ablation scan data with the ablation probability model.

6. Applying the ablation probability model to the spatial alignment of the time sequence sets of ablation scan data includes: the at least one processor is further configured to apply, with the ablation probability model, at least one ablation probability rule to the temporal pixel intensity estimates and the spatial pixel intensity estimates of the pixel ablation probability for each pixel of the ablation scan image indicative of the anatomical ablation; The thermal ablation probability controller of claim 5 , wherein the at least one ablation probability rule is based on a comparison of the temporal pixel intensity estimate of the pixel and the temporal pixel intensity estimate of the at least one neighboring pixel to an ablation threshold.

7. The at least one processor generating an ablation probability image derived from the pixel ablation probabilities for each pixel of the ablation scan image indicative of the anatomical ablation; generating an ablation probability anatomical scan derived from the anatomical image of the anatomical ablation and further derived from the pixel ablation probability for each pixel of the ablation scan indicative of the anatomical ablation; and 2. The thermal ablation probability controller of claim 1, further configured to at least one of: generate an ablation probability ablation scan image derived from the ablation scan image indicative of the anatomical ablation and further derived from the pixel ablation probability for each pixel of the ablation scan image indicative of the anatomical ablation.

8. The thermal ablation probability controller of claim 1 , wherein an intensity value of each pixel of the time series set of ablation scan data represents one of anatomical stiffness, anatomical density, or anatomical temperature.

9. 1. A non-transitory machine-readable storage medium encoded with instructions for executing, by at least one processor, a trained ablation probability model to render a pixel ablation probability for each pixel of an ablation scan dataset representing an anatomical ablation, the non-transitory machine-readable storage medium comprising: spatially aligning a time series set of ablation scan data representing dynamic anatomical ablation; instructions for applying the ablation probability model to a spatial alignment of the time sequence set of ablation scan data to render a pixel ablation probability for each pixel of an ablation scan image indicative of the anatomical ablation; The time series set of scan data representing the dynamic anatomical ablation is a time series data set of images showing an expanding ablation region in a plurality of frame images intermittently captured from a specific site over a period of time from a specific time. A non-transitory machine-readable storage medium.

10. Applying the ablation probability model to the spatial alignment of the time sequence sets of ablation scan data includes: deriving temporal pixel intensities for pixels of the ablation scan images indicative of the anatomical ablation from the spatial alignment of the time series sets of ablation scan data; deriving a temporal pixel intensity of at least one adjacent pixel of the ablation scan image indicative of the anatomical ablation from the spatial alignment of the time series set of ablation scan data; and 10. The non-transitory machine-readable storage medium of claim 9, further comprising instructions for applying an ablation probability rule to the temporal pixel intensity of the pixel and the temporal pixel intensity of the at least one neighboring pixel to render the pixel ablation probability for the pixel.

11. The instructions for applying the ablation probability model to the spatial alignment of the time sequence set of ablation scan data include:

10. The non-transitory machine-readable storage medium of claim 9, comprising instructions for performing a time series temporal pixel intensity assessment of each pixel of the spatial alignment of a time series set of the ablation scan data by the ablation probability model.

12. The temporal pixel intensity estimate for each pixel of the time series set of ablation scan data comprises: an intensity average of each pixel of the spatial alignment of the time series set of ablation scan data; a median intensity value of each pixel of the spatial alignment of the time series set of ablation scan data; a maximum intensity value of each pixel in the spatial alignment of a time series set of ablation scan data; a minimum intensity value of each pixel of the spatial alignment of the time series set of ablation scan data; 12. The non-transitory machine-readable storage medium of claim 11, wherein the intensity value is one of: an intensity value derived from a standard deviation of the time series of each pixel of the spatial alignment of a time series set of the ablation scan data.

13. The instructions for applying the ablation probability model to the spatial alignment of the time sequence set of ablation scan data include:

12. The non-transitory machine-readable storage medium of claim 11, further comprising instructions for: calculating, with the ablation probability model, a spatial pixel intensity estimate of the or each temporal pixel intensity estimate of at least one adjacent pixel of the spatial alignment of the time sequence set of the ablation scan data.

14. The instructions for applying the ablation probability model to the spatial alignment of the time sequence set of ablation scan data include: further comprising instructions for applying, by the ablation probability model, at least one ablation probability rule to the temporal pixel intensity estimates and the spatial pixel intensity estimates of the pixel ablation probability for each pixel of the ablation scan image indicative of the anatomical ablation; 14. The non-transitory machine-readable storage medium of claim 13, wherein the at least one ablation probability rule is based on a comparison of the temporal pixel intensity assessment of the pixel and the temporal pixel intensity assessment of the at least one neighboring pixel to an ablation threshold.

15. generating an ablation probability image derived from the pixel ablation probabilities for each pixel of the ablation scan image indicative of the anatomical ablation; generating an ablation probability anatomical scan derived from the anatomical image of the anatomical ablation and further derived from the pixel ablation probability for each pixel of the ablation scan indicative of the anatomical ablation; and 10. The non-transitory machine-readable storage medium of claim 9, further comprising instructions for at least one of: generating an ablation probability ablation scan image derived from the ablation scan image indicative of the anatomical ablation and further derived from the pixel ablation probability for each pixel of the ablation scan image indicative of the anatomical ablation.

16. A method of operation of a system including an ablation probability model trained to render a pixel ablation probability for each pixel of an ablation scan dataset representing an anatomical ablation, the method comprising: a thermal ablation probability controller of the system spatially aligning a time sequence set of ablation scan data representative of dynamic anatomical ablation; the thermal ablation probability controller applying the ablation probability model to a spatial alignment of the time series set of ablation scan data to render the pixel ablation probability for each pixel of an ablation scan image indicative of the anatomical ablation; The time series set of scan data representing the dynamic anatomical ablation is a time series data set of images showing an expanding ablation region in a plurality of frame images intermittently captured from a specific site over a period of time from a specific time. How it works.

17. The step of the thermal ablation probability controller applying the ablation probability model to the spatial alignment of the time sequence set of ablation scan data may further comprise the thermal ablation probability controller: deriving temporal pixel intensities of pixels of the ablation scan image indicative of the anatomical ablation from the spatial alignment of the time series set of ablation scan data; deriving a temporal pixel intensity of at least one adjacent pixel of the ablation scan image indicative of the anatomical ablation from the spatial alignment of the time series set of ablation scan data; and applying an ablation probability rule to the temporal pixel intensity of the pixel and the temporal pixel intensity of the at least one neighboring pixel to render the pixel ablation probability for the pixel.

18. The step of the thermal ablation probability controller applying the ablation probability model to the spatial alignment of the time sequence set of ablation scan data may further comprise the thermal ablation probability controller: performing a time series temporal pixel intensity assessment of each pixel of the spatial alignment of the time series set of the ablation scan data with the ablation probability model; and performing, by the ablation probability model, spatial pixel intensity estimation of one or each of at least one neighboring pixel of each pixel of the spatial alignment of the time series set of the ablation scan data.

19. The step of applying, by the thermal ablation probability controller, the ablation probability model to the spatial alignment of the time sequence set of ablation scan data comprises: applying at least one probability rule to the temporal and spatial pixel intensity estimates by the ablation probability model to render the pixel ablation probability for each pixel of the ablation scan image indicative of the anatomical ablation; The method of claim 16 , wherein at least one ablation probability rule is based on a comparison of the temporal pixel intensity estimate of the pixel and the temporal pixel intensity estimate of at least one neighboring pixel to an ablation threshold.

20. the thermal ablation probability controller: generating an ablation probability image derived from the pixel ablation probabilities for each pixel of the ablation scan image indicative of the anatomical ablation; generating an ablation probability anatomical scan derived from the anatomical image of the anatomical ablation and further derived from the pixel ablation probability for each pixel of the ablation scan showing the anatomical ablation; and generating an ablation probability ablation scan image derived from the ablation scan image indicative of the anatomical ablation and further derived from the pixel ablation probability for each pixel of the ablation scan image indicative of the anatomical ablation.

Images