Delivery of Therapeutic Neuromodulation

Abstract

Embodiments of the present disclosure relate to techniques for neuromodulation delivery. To maintain desirable characteristics of neuromodulation energy, control parameters governing the application of neuromodulation energy can be dynamically altered during the course of delivery based on image data acquired from the patient. For example, the beam of neuromodulation energy may be dynamically adjusted to account for organ movement during respiration. In another embodiment, a desired region of interest within the patient is identified based on a trained neural network and acquired image data.

Description

[Technical field]

[0001] The subject matter disclosed herein relates to the application of neuromodulatory energy to identify, target, and / or deliver to a patient's region of interest for the purpose of inducing a desired physiological outcome. In particular, the disclosed techniques may be part of a personalized treatment protocol. [Background technology]

[0002] Neuromodulation has been used to treat a variety of clinical conditions. For example, electrical stimulation has been employed at various locations along the spinal cord to treat chronic low back pain. However, it is difficult to place electrodes at or near the target nerve. For example, such techniques may involve surgical procedures to place the electrodes that deliver the energy. Furthermore, it is difficult to target specific tissues with neuromodulation. Neuromodulation is mediated by an electrode placed at or near a specific target nerve, which induces action potentials in nerve fibers, resulting in neurotransmitter release from the neural synapse and synaptic communication with the next nerve. Current implementations of implanted electrodes stimulate many nerves or axons at once, and such propagation may result in physiological effects that are relatively broad or diffused more than desired. Because neural pathways are complex and interconnected, more selective targeting of the modulatory effects may be clinically useful. However, the effectiveness of selective targeting of specific nerves depends on how precisely the energy application device can be placed. How precisely the neuromodulation energy can be focused may vary depending on the anatomy of the individual patient. For example, a particular patient's height, weight, age, sex, clinical condition, etc. may affect the size and location of organs compared to other patients. Additionally, patients may undergo anatomical changes over time that may complicate the accuracy of energy delivery. Summary of the Invention

[0003] The disclosed embodiments are not intended to limit the scope of the claimed subject matter, but rather, these embodiments are intended only to provide a brief summary of possible embodiments. Indeed, the disclosure may include a variety of forms, which may be similar to or different from the embodiments set forth below.

[0004] In one embodiment, a neuromodulation delivery system is provided, the system including an energy application device configured to deliver neuromodulation energy to a region of interest of a patient. The system also includes a controller configured to receive image data of the patient's internal tissue; identify the region of interest from the image data; control application of the neuromodulation energy to the identified region of interest via the energy application device to deliver a dosage of the neuromodulation energy to the identified region of interest; receive updated image data of the patient's internal tissue before delivery of the dosage is completed; identify a change in position of the region of interest relative to the energy application device based on the updated image data; and adjust application of the neuromodulation energy via the energy application device based on the changed position of the region of interest to continue the delivery of the dosage of the neuromodulation energy to treat the patient.

[0005] In another embodiment, a method for delivering energy for neuromodulation is provided, comprising the steps of: delivering energy to a region of interest of a patient using control parameters, the energy being a portion of a total amount of energy of an individual dose applied to the region of interest, applying the energy using an energy application device; acquiring image data from the patient while delivering the energy and before the total amount of energy of the individual dose is applied, representing an internal tissue including the region of interest; determining a change in position of the region of interest relative to the energy application device based on the image data; adjusting one or more control parameters of a group of control parameters based on the change in position of the region of interest; delivering additional energy to the region of interest using the adjusted control parameters, and using the energy application device to deliver another portion of the total amount of energy of the individual dose.

[0006] In another embodiment, a neuromodulation delivery system is provided, the system including an energy application device configured to deliver neuromodulation energy to a region of interest of a patient. The system also includes a controller configured to control the energy application device to acquire image data representative of an internal tissue of the patient; identify the region of interest based on the image data using a neural network trained with image data of internal tissues of each patient in a population that are the same type of tissue as the internal tissue of the patient; control application of the neuromodulation energy by the energy application device to the identified region of interest to deliver a dose of the neuromodulation energy to treat the patient; acquire updated image data while delivering a previous dose; and dynamically modify one or more control parameters for controlling application of the neuromodulation energy based on the updated image data. [Brief description of the drawings]

[0007] These and other features, aspects, and advantages of the present invention will be better understood from the following detailed description taken in conjunction with the accompanying drawings, in which like reference numerals refer to like parts throughout.

[0008] [Figure 1] FIG. 1 is a schematic diagram of an autonomous neuromodulation delivery system according to an embodiment of the present disclosure.

[0009] [Diagram 2] FIG. 1 is a block diagram of an autonomous neuromodulation delivery system according to an embodiment of the present disclosure.

[0010] [Diagram 3] FIG. 1 is a schematic diagram of an autonomous neuromodulation delivery system for applying neuromodulation energy to a region of interest in tissue, including an anatomical structure, in an embodiment of the disclosure.

[0011] [Figure 4] FIG. 1 is a schematic diagram of an autonomous neuromodulation delivery system that tracks a moving region of interest, according to an embodiment of the present disclosure.

[0012] [Diagram 5] FIG. 5 is a schematic diagram of the application of adjusted energy based on the movements identified in FIG. 4.

[0013] [Figure 6] FIG. 1 is a flow diagram of an autonomous neuromodulation delivery technique according to an embodiment of the present disclosure.

[0014] [Figure 7] FIG. 2 is a schematic diagram of inputs to a neural network according to an embodiment of the present disclosure.

[0015] [Figure 8] FIG. 1 is a flow diagram of an autonomous neuromodulation delivery technique according to an embodiment of the present disclosure.

[0016] [Figure 9] FIG. 1 is a block diagram of an example of an autonomous neuromodulation delivery system including a dual imaging and therapeutic probe according to an embodiment of the present disclosure.

[0017] [Figure 10] 10 is an image of the dual imaging and therapeutic probe of FIG.

[0018] [Figure 11] 1 is an exemplary graphical user interface of an autonomous neuromodulation delivery system according to an embodiment of the present disclosure.

[0019] [Figure 12] 1 is an exemplary graphical user interface of an autonomous neuromodulation delivery system during a process of alignment with a region of interest, according to an embodiment of the present disclosure.

[0020] [Figure 13] 1 is an exemplary graphical user interface of an autonomous neuromodulation delivery system during delivery of energy for neuromodulation to a region of interest, according to an embodiment of the present disclosure.

[0021] [Figure 14] 1 is an exemplary graphical user interface of an autonomous neuromodulation delivery system after completion of delivery of energy for neuromodulation to a region of interest, according to an embodiment of the present disclosure.

[0022] [Figure 15] FIG. 1 illustrates an example of organ identification using an autonomous neuromodulation delivery system according to an embodiment of the present disclosure. DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS

[0023] One or more specific embodiments are described below. To simplify the description of these embodiments, not all features of an actual implementation are described herein. It will be appreciated that developing an actual implementation, such as an engineering or design project, will require many implementation-specific decisions to be made to achieve the developer's particular goals, such as addressing system-related and business-related constraints that may vary from implementation to implementation. It will also be appreciated that such development efforts may be complex and time-consuming, but are nevertheless routine design, fabrication, and manufacturing efforts for those skilled in the art having the benefit of this disclosure.

[0024] Any examples or illustrations described herein should not be considered as limiting, restricting, or defining any of the terms used in the examples or illustrations. Instead, such examples or illustrations are described in the context of different specific embodiments and should be considered as merely illustrative. Those skilled in the art will understand that any term or terms used in such examples or illustrations are intended to encompass other embodiments in which such example or illustration is or is not used or described elsewhere herein, and such embodiments are intended to be included within the scope of the term or terms. Words expressing such non-limiting examples and illustrations include, but are not limited to, "for example," "for example," "such as," "for example," "including," "in certain embodiments," "in some embodiments," and "in one embodiment."

[0025] Disclosed herein is a technique for neuromodulation of a targeted region of interest that allows for reproducible and reliable application of energy to a specific region or regions of interest as part of a treatment protocol during the course of the treatment protocol. The disclosed technique provides an autonomous method for neuromodulation energy delivery that dynamically adjusts one or more parameters of delivery based on changes in the location of the desired energy target (e.g., region of interest) during energy delivery, so that energy delivery does not need to be interrupted. For example, a patient may be instructed not to move during delivery of neuromodulation energy, but even slight changes in the patient's position or the patient's breathing can cause internal organs to move, which can change the relative location of the target with respect to an external or extracorporeal energy application device (e.g., an ultrasound therapy probe). Because the neuromodulation energy can be focused on a volume of tissue that contains a particular axon terminal or group of axon terminals within the tissue (or conversely, a volume of the tissue that does not contain other axon terminals within the tissue), any slight movement of the tissue can cause the focal zone of the energy application device to move away from the region of interest and into an adjacent region of tissue that does not contain the desired axon terminal, thereby potentially missing the physiological outcome of neuromodulating the desired axon terminal. Movement of the region of interest, even on the order of millimeters or centimeters, can result in the neuromodulation energy being applied in an inaccurate location, which can result in failure to achieve the desired therapeutic goal.

[0026] To account for tissue movement that may occur during the time it takes to apply one or more doses of neuromodulation energy, a larger delivery area may be created outside the volume of the region of interest, taking into account minor movements. However, this may expose other axon terminals to the neuromodulation energy depending on the particular region of interest, reducing the desired specificity of the treatment and potentially resulting in confounding physiological effects that may slow or prevent the achievement of the desired therapeutic goal. Furthermore, such an approach may limit the total dose that can be delivered in a single treatment session, as the overall tissue volume exposed to the energy may be large, and the energy limit per dose may be reached in a short time, possibly before the desired dose is delivered to the region of interest.

[0027] The technology allows for precise delivery of neuromodulation energy in a user-friendly manner that eliminates or reduces the need for trained clinicians to input anatomical guidance, thereby enabling less experienced caregivers to administer treatment in the home or outpatient setting and increasing treatment options. Additionally, while trained clinicians can identify anatomical landmarks for precise energy delivery, individual clinicians may impose their own preference biases on treatment, which may hinder delivery of accurate doses over time, especially when multiple clinicians are involved in care. The technology provides a tool for less experienced caregivers to deliver neuromodulation energy as part of a treatment protocol.

[0028] In one embodiment, the disclosed technology incorporates imaging data acquired before and / or during a treatment session to track movement of the region of interest over time, allowing real-time refocusing and redirecting of energy to maintain delivery of energy on or within the region of interest. The imaging data can be acquired using a multi-function device configured to both acquire image data and deliver neuromodulation energy. In one example, a neural network is used to identify an organ that includes a region, region of interest, anatomical structure, or combination thereof. Once identified, movement can be tracked based on ongoing or updated acquired image data.

[0029] In some embodiments, a neural network can be trained on images from a patient population to rapidly identify organs or other tissues that contain regions of interest without operator involvement. The neural network architecture may use various layers that allow for identification of structures based on morphology, pattern matching, edge detection, etc. The neural network may also be configured to identify regions of interest within organs or tissues. For example, if the region of interest is the hilar region of the liver, the neural network can be trained on gold standard images of a patient population to identify regions of interest that are likely to contain or overlap the hilar region of a particular patient.

[0030] To that end, the disclosed neuromodulation delivery techniques may be used in conjunction with a neuromodulation system configured for use in delivering neuromodulation energy as part of a treatment protocol. Figure 1 is a schematic diagram of a neuromodulation system 10 for applying energy to achieve a neuromodulation effect, such as releasing neurotransmitters and / or activating synaptic elements (e.g., presynaptic cells, postsynaptic cells). The illustrated system includes a pulse generator 14 coupled to an energy application device 12 (e.g., an ultrasound transducer). The energy application device 12 is configured to receive energy pulses, e.g., via a lead or wireless connection, which are directed to a region of interest in a patient's internal tissue or organ during use to produce a targeted physiological outcome.

[0031] In certain embodiments, the energy application device 12 and / or the pulse generator 14 may communicate, e.g., wirelessly, with the controller 16, which may then send instructions to the pulse generator 14. In other embodiments, the energy application device 12 may be an external device, e.g., operable to apply energy percutaneously or non-invasively from a location outside the patient's body, and in certain embodiments may be integrated with the pulse generator 14 and / or the controller 16. In embodiments in which the energy application device 12 is an external device, the energy application device 12 may be operated by a caregiver and positioned at or above the patient's skin such that energy pulses are delivered percutaneously to desired internal tissue. Once positioned to apply energy pulses to desired sites, neuromodulation of one or more neural pathways may be initiated by the system 10 to achieve a desired physiological outcome or clinical effect. In other embodiments, the pulse generator 14 and / or the energy application device 12 may be implanted in a biocompatible location (e.g., the abdomen) or may be coupled internally, e.g., via one or more leads. In some embodiments, system 10 may be implemented such that some or all of its components are in communication with one another in a wired or wireless manner.

[0032] In certain embodiments, the system 10 may include an evaluation device 20 coupled to the controller 16 and configured to evaluate a characteristic indicative of whether a targeted physiological outcome of the modulation has been achieved. In one embodiment, the targeted physiological outcome may be local. For example, modulation of one or more neural pathways may result in a local change in tissue or function, such as altering tissue structure, inducing a local change in a particular molecule concentration, displacing tissue, increasing fluid movement, etc. The targeted physiological outcome may be the goal of a treatment protocol.

[0033] Modulation of one or more neural pathways to achieve a desired physiological outcome may result in systemic or non-localized changes, and may relate to changes in circulating molecular concentrations or property changes in tissues that do not include the region of interest to which energy is directly applied. In one example, displacement may be measured as a proxy for the desired modulation, and if the measured displacement falls below the expected displacement value, modulation parameters may be modified until the expected displacement value is achieved. Thus, in some embodiments, the evaluation device 20 may be configured to evaluate concentration changes. In some embodiments, the evaluation device 20 may be an imaging device configured to evaluate changes in organ size position and / or tissue properties. In another embodiment, the evaluation device 20 may be a circulating glucose monitor. Although the components of the system 10 are shown separately in the figures, it is understood that some or all of the components may be combined with each other. In another embodiment, the evaluation device may evaluate a local temperature increase in the tissue, which may be detected using a separate temperature sensor or, if the energy application device 12 is configured to apply ultrasound energy, using ultrasound imaging data. Assessment of sound speed differences can be detected using differential imaging techniques before / during / after treatment.

[0034] Based on the evaluation, the modulation parameters of the controller 16 may be altered to deliver an effective amount of energy. For example, if the desired modulation is related to a concentration (circulating or tissue concentration of one or more molecules) change relative to a baseline value over a defined time frame (e.g., 5 minutes, 30 minutes after the start of the energy application treatment) or at the start of the treatment, it may be desirable to alter the modulation parameters, such as the pulse frequency or other parameters. In such cases, the energy application parameters or modulation parameters of the pulse generator 14 may be modified by an operator or an automated feedback loop to define or adjust the parameters or modulation parameters until an effective amount of energy is applied. As described herein, the data from the evaluation device 20 may be provided as part of a feedback loop to train a personalized neural network as part of a treatment protocol and / or to re-steer or re-focus the energy to account for movement of the region of interest during treatment. In one embodiment, an updated region of interest may be obtained by adjusting the originally defined region of interest over the course of a treatment protocol based on feedback from the evaluation device regarding the effectiveness of the neuromodulation energy. The feedback may be, for example, a change in concentration of a molecule of interest as a result of application of neuromodulation energy. Such adjustments or updates to the regions of interest may be used as part of a patient-specific network, where the network is updated to recognize the particular regions of interest that most impact the physiological parameter of interest for that particular individual based on the desired clinical outcome.

[0035] The system 10 provided herein may provide energy pulses according to various modulation parameters that are part of a treatment protocol to apply an effective amount of energy. For example, the modulation parameters may include various patterns of stimulation time, ranging from continuous to intermittent patterns. Intermittent stimulation delivers energy at a constant frequency for a period of time during which the signal is on. After the signal on time, there is a period of time during which no energy is delivered, called the signal off time. The modulation parameters may further include the frequency and duration of application of the stimulation. The frequency of application may be continuous or may be delivered for various periods of time, for example, daily or weekly cycles. Additionally, the treatment protocol may specify the time of day for application of the energy or may specify the time relative to meals or other activities. The treatment time may last for various periods of time, including, but not limited to, minutes to hours, to induce a desired physiological outcome. In certain embodiments, the duration of treatment with a specified stimulation pattern may be one hour, and may be repeated at intervals of, for example, 72 hours. In certain embodiments, energy delivery may be for a shorter duration, e.g., 30 minutes, or more frequently, e.g., every 3 hours. Application of energy can be adjustably controlled according to modulation parameters such as treatment duration, frequency, and amplitude to achieve the desired results.

[0036] FIG. 2 is a block diagram of certain components of the system 10. As described herein, the neuromodulation system 10 may include a pulse generator 14 adapted to generate a plurality of energy pulses for application to tissue of a patient. The pulse generator 14 may be separate or integrated with an external device, such as a controller 16. The controller 16 includes a processor 30 for controlling the device. A memory 32 of the controller 16 stores software code or instructions executed by the processor 30 to control the various components of the device. The controller 16 and / or the pulse generator 14 may be connected to the energy application device 12 via one or more leads 33 or wirelessly. The processor 30 may be configured to access software from the memory 32 for operating a neural network that has been pre-trained on images of other patients (e.g., not necessarily the patient being treated). Additionally, the processor may be configured to update the neural network based on image data of the patient of interest.

[0037] The controller 16 may include a user interface having input / output circuitry 34 and a display 36 adapted to allow a clinician to input selections for modulation programs and set modulation parameters. However, certain embodiments of the system 10 may include implementations that do not include a display 36 that provides feedback using sound or light. For example, a relatively simple at-home system 10 may be configured with or without a display 36 to avoid displaying information that is not useful to an inexperienced user in achieving treatment goals. The processor 30 may be configured to operate the neural network and identify regions of interest using a relatively simple interface while still providing guidance to move the energy application device 12 to the correct treatment location.

[0038] The system may include a beam controller 37 capable of controlling the position of the focus of the energy beam of the energy application device 12 by controlling one or both of the steering and / or focusing of the energy application device 12. The beam controller 37 may further control one or more articulated parts of the energy application device 12 to change the position of the transducer. The beam controller may receive instructions from the processor 30 to change the focusing and / or steering of the energy beam. The system 10 may be responsive to one or more position sensors 38 and / or contact sensors 39 that provide feedback about the energy application device 12. The beam controller 37 may include a motor that allows steering of one or more articulated parts of the energy application device 12. In one embodiment, the motor is internal to the probe housing, with a fixed surface (the lens of the ultrasound probe) contacting the body. The motor may move internally with one to six degrees of freedom while the probe is stationary on the body. In additional or other embodiments, the probe is shaped like a conventional imaging probe, held in a motorized fixture, and moved along the skin with up to six degrees of freedom, similar to freehand scanning. Angle change corresponds to three degrees of freedom and corresponds to steering of the beam in three-dimensional space. Position change corresponds to the other three degrees of freedom and consists of XY motion for gliding along the surface of the body, or Z motion for adjusting focal depth or contact force. It is contemplated that system 10 may include features that allow for adjustment of position, steering, and / or focus to enable the techniques disclosed herein.

[0039] Each modulation program stored in memory 32 may include one or more sets of modulation parameters, including pulse amplitude, pulse duration, pulse frequency, pulse repetition rate, etc. In response to a control signal from controller 16, pulse generator 14 modifies its internal parameters, thereby changing the stimulation characteristics of the energy pulses delivered via lead 33 to the patient to whom energy application device 12 is applied. Any suitable type of pulse generating circuit may be used, including but not limited to constant current, constant voltage, multiple independent current or voltage sources, etc. The applied energy is a function of the current amplitude and pulse duration. Controller 16 may adjustably control the energy by modifying modulation parameters and / or initiating energy application at specific times or canceling / reducing energy application at specific times. In one embodiment, the adjustable control of the energy application device that applies the energy is based on information regarding the concentration of one or more molecules (e.g., circulating molecules) in the patient. If the information is obtained from evaluation device 20, a feedback loop may drive the adjustable control. For example, a diagnosis may be made based on circulating glucose concentrations measured by evaluation device 20 in response to neuromodulation. If the concentration is above a predetermined threshold or range, the controller 16 can initiate a therapeutic protocol with energy application to the region of interest (e.g., the liver) using modulation parameters associated with reducing circulating glucose. The therapeutic protocol can use different modulation parameters (e.g., higher energy levels and application frequency) than those used for the diagnostic protocol.

[0040] In one embodiment, different operating modes are stored in the memory 32 that are selectable by the operator. The stored operating modes may include, for example, separate models or neural networks for identifying a particular region of interest and implementing a set of modulation parameters associated with a particular treatment site, such as a region of interest in the liver, pancreas, gastrointestinal tract, or spleen. Each organ or site may be associated with a different model. Additionally, different modulation parameters may be associated with different sites based on the depth of the associated organ, the size of the region of interest, the desired physiological outcome, etc. The controller 16 may be configured to execute appropriate instructions when a particular organ is selected, instead of having the operator manually enter the mode. In another embodiment, operating modes for different types of treatments are stored in the memory 32. For example, activation may be associated with a different stimulation pressure or frequency range than inhibiting or blocking tissue function.

[0041] In a particular example, when the energy application device is an ultrasound transducer, the effective amount of energy may include a predetermined time-averaged intensity applied to the region of interest. For example, the effective amount of energy may include a time-averaged power (time-averaged intensity) and a peak positive pressure, each of which may be between 1 mW / cm and 30,000 mW / cm. 2 (time-average intensity) and in the range of 0.1 MPa to 7 MPa (peak pressure). In one example, the time-average intensity is 35 mW / cm in the region of interest. 2 Less than 500mW / cm 2 Less than 720mW / cm 2In one example, the time-averaged intensity is associated with a level lower than that associated with thermal damage and cauterization / cavitation phenomena. In another example, if the energy application device is a mechanical actuator, the vibration amplitude is in the range of 0.1-10 mm. The selected frequency may depend on the mode of application of the energy, e.g., ultrasonic or mechanical actuator. The controller 16 may be operable in a verification mode to obtain a predefined treatment position, which may be implemented as part of a treatment mode of operation configured to execute a treatment protocol when the energy application device 12 is placed at the predefined treatment position.

[0042] The system may also include an imaging device to facilitate focusing of the energy application device 12. In one embodiment, the imaging device may be integrated with the energy application device 12, or both may be the same device, such that different ultrasound parameters (frequency, aperture, or energy) are applied when selecting (e.g., spatially selecting) the region of interest and when focusing the energy on the selected region of interest for targeting and subsequent neuromodulation. In another embodiment, the memory 32 stores one or more targeting or focusing modes that are used to spatially select the region of interest within the organ or tissue structure. The spatial selection may include selecting a sub-region of the organ to identify a volume of the organ that corresponds to the region of interest. The spatial selection may be according to image data provided herein. Based on the spatial selection, the energy application device 12 may focus (e.g., using the beam controller 37) on a focal position on the selected volume that corresponds to the region of interest. It is understood that the image data used to guide the focal position may be either three-dimensional data or two-dimensional data. For example, the energy application device 12 may be configured to first operate in a verification mode to obtain the predetermined treatment location by capturing image data that is used to identify the predetermined treatment location associated with capturing the region of interest. The verification mode energy may not be at a level and / or have modulation parameters applied that are appropriate for neuromodulation therapy. However, once the region of interest is identified, the controller 16 may be operated in a treatment mode according to modulation parameters associated with achieving a targeted physiological outcome.

[0043] The controller 16 may also be configured to accept inputs related to the desired physiological outcome as inputs for the selection of the modulation parameters. For example, if an imaging modality is used for tissue characterization, the controller 16 may be configured to receive a calculated index for that characteristic or a parameter for that characteristic. A diagnosis may be made based on whether the index or parameter is above or below a predefined threshold, and the diagnosis may be displayed (e.g., on a display). In one embodiment, the parameter may be a measurement of tissue displacement of the diseased tissue, or a measurement of the depth of the diseased tissue. Other parameters may include an evaluation of one or more concentrations of molecules of interest (e.g., one or more of a concentration change relative to a threshold or reference value / control, a rate of change, and a determination of whether the concentration is within a desired range). Additionally, the energy application device 12 (e.g., an ultrasound transducer) may operate under the control of the controller 16 to: a) acquire tissue image data that can be used to spatially select a region of interest in the target tissue; b) apply modulation energy to the region of interest; and c) acquire image data for determining whether the desired physiological outcome has occurred (e.g., by measuring displacement). In such an embodiment, the imaging device, the evaluation device 20, and the energy application device 12 may be the same device.

[0044] FIG. 3 illustrates energy delivery to a region of interest 44 using the energy application device 12 described above. The energy application device 12 includes an ultrasound transducer 42 (e.g., a transducer array) capable of applying energy to a target organ or tissue 43, such as the liver, spleen, pancreas, etc. The energy application device 12 may include control circuitry to control the ultrasound transducer 42. The control circuitry of the processor 30 (FIG. 2) may be part of the energy application device 12 (e.g., through an integrated controller 16) or may be a separate component. The energy application device 12 may also be configured to acquire image data to assist in selecting a desired or targeted region of interest 44 in space and focusing the applied energy to the region of interest of the target tissue or structure.

[0045] The region of interest 44 and / or the target tissue 43 may include anatomical features or structures to facilitate automatic recognition, for example, using neural networks, as described herein. For example, an organ may have a characteristic edge 50 with a particular shape, or may have capillaries or smaller blood vessels 52, as well as internal nerve structures 54 extending into the tissue 43. The tissue 43 may be within a predictable size or volume range based on the patient's size, weight, age, and / or clinical condition, or may be characterized by a dimensional range 56 along, for example, the x, y, or z axis. The tissue 43 may also be positioned relative to other organs 60 or other internal structures, such as larger blood vessels 62. These and other features may be provided as inputs in the recognition of the region of interest 44 and / or the target tissue 43. Additionally, such features may be obtained from patient populations to identify predictable features (e.g., location of larger blood vessels, location of organs or glands) that tend to appear stable across patient populations, in addition to variable features that vary over time between and within patients depending on clinical symptoms, metabolic state, patient weight, etc. For example, the size of certain organs may change after eating. Other cognitive features may be the entry / exit points of blood vessels, arteries, and veins into organs (e.g., "porta," "hilus," "hilum," "fissure," "indendation," "duct," etc.). Based on the variability of these factors, networks with different weights and filters may be used. If these factors change, appropriate models trained specifically on other patients with similar factors may be used. An example may be the aging of the patient. Different networks of models may be available for individuals of different ages. If the patient's age changes, the best model for that patient at a given time may be selected. Thus, different sets of networks may be accessed. Such networks may include more general or generic models, models tailored to specific demographic characteristics, and completely individualized models.

[0046] The desired target tissue 43 may be an internal tissue or organ that contains axon terminals and synapses of non-neuronal cells. The synapses may be stimulated by applying energy directly to axon terminals within a focal region or zone 48 of an ultrasound transducer 42 focused on a region of interest 44 of the target tissue 43, resulting in action potentials and / or release of molecules into the synaptic cleft, such as neurotransmitter release and / or ion channel activity changes that produce downstream effects. The region of interest 44 may be selected to include a particular type of axon terminal, such as axon terminals of a particular type of neuron and / or axon terminals that form synapses with a particular type of non-neuronal cell. Thus, the region of interest 44 may be selected to correspond to a portion of the target tissue 43 that contains the desired axon terminals (and associated non-neuronal cells). The application of energy may be selected to preferentially induce the release of one or more molecules, such as neurotransmitters, from the nerves in the synapse, to directly activate the non-neuronal cells themselves via direct energy transduction (i.e., mechanotransduction or voltage-activated proteins in the non-neuronal cells), or to activate both neuronal and non-neuronal cells to induce the desired physiological effect. The region of interest 44 may be selected at the site where the nerve enters the organ. In one embodiment, liver stimulation or modulation may refer to modulation of the region of interest 44 at or adjacent to the hepatic hilum. In identifying a predetermined treatment location 46 on the patient's skin (or clothing), the region of interest 44 may be selected such that the location on the patient's body where the region of interest 44 is within the focal zone 48 when the energy application device 12 is operated is the predetermined treatment location 46.

[0047] The energy may be focused or substantially focused on only the region of interest 44 and a portion of the internal tissue 43, for example, less than about 50%, 25%, 10%, or 5% of the total volume of the tissue 43. That is, the region of interest 44 may be a sub-region of the internal tissue 43. In one embodiment, energy may be applied to two or more regions of interest 44 within the target tissue 43, and the sum of the volumes of the two or more regions of interest 44 may be less than about 90%, 50%, 25%, 10%, or 5% of the total volume of the tissue 43. In one embodiment, the energy is applied to only about 1%-50% of the total volume of the tissue 43, only about 1%-25% of the total volume of the tissue 43, only about 1%-10% of the total volume of the tissue 43, or only about 1%-5% of the total volume of the tissue 43. In certain embodiments, only axon terminals in the region of interest 44 of the target tissue 43 directly receive the applied energy and release neurotransmitters, while non-stimulated axon terminals outside the region of interest 44 do not receive substantial energy and are therefore not activated / stimulated in the same way. In some embodiments, axon terminals in the tissue portion directly receiving the energy release neurotransmitters in an altered manner. In this manner, subregions of tissue can be targeted for neuromodulation with finer granularity, for example, by selecting one or more subregions. In some embodiments, parameters during energy application can be selected to preferentially activate either neural or non-neural elements in the tissue directly receiving the energy to induce a desired combination of physiological effects. In certain embodiments, the energy can be applied to a tissue having a volume of approximately 25 mm. 3 The energy may be focused or concentrated within a volume of less than about 0.5 mm. 3 ~50mm 3 The energy may be focused or concentrated within a volume of interest 44. The focal volume and focal depth within the region of interest 44 to which the energy is focused or concentrated may depend on the size / configuration of the energy application device 12. The focal volume of the energy application may be defined by a focal point or focal zone of the energy application device 12.

[0048] Energy may be targeted and applied substantially only to one or more regions of interest 44 to preferentially activate synapses to achieve a desired physiological outcome. Thus, in certain embodiments, only a portion of multiple different types of axon terminals within tissue 43 are exposed to the directly applied energy.

[0049] As described herein, identifying the correct treatment location 46 on the patient may not be sufficient to target energy delivery from the energy application device 12 to the region of interest 44. As shown in FIG. 4, breathing or movement of the patient during treatment may cause the region of interest 44 to move out of the focal zone 48 (shown as initial focal zone 48a). The system 10 may be configured to acquire updated or continuous image data of the tissue 43 using an imaging transducer 68 in the energy application device 12 (although such image data may be acquired during or alternating dark or quiescent periods of treatment energy delivery, or gating controlled via the controller 16). The parameters of the energy used to acquire the image data may be different from the parameters of the treatment energy, and in one embodiment may not achieve the desired physiological outcome. If movement is detected away from the transducer 42, the modulation parameters may be adjusted to increase power and / or increase application time to achieve the desired exposure, as shown in Figure 5, to begin steering / focusing the ultrasound beam to the new location of the region of interest 44b, 44c. If movement is tracked closer to the skin, the modulation parameters may be adjusted to decrease power and / or decrease application time to steer and / or focus the ultrasound beam to the new location.

[0050] Such adjustments may be made dynamically to account for real-time movement of the region of interest 44 to achieve the desired exposure. Additionally, the system 10 takes into account global changes in modulation parameters when calculating the energy dose to apply. Such adjustments may be made even if the energy application device 12 has not moved from the treatment location 46 as shown. That is, the treatment location 46 allows energy to be delivered to the region of interest 44 that is within the treatable region 70. The treatable region 70 is based on the general operating parameters and geometry of the transducer 42 and the energy application device 12. As long as the region of interest 44 remains within the treatable region 70, even if it moves within the treatable region 70, the energy application device 12 will automatically steer or adjust to allow the delivery of the dose without operator intervention, without pausing the energy application device 12 or physically moving it away from the treatment location 46. That is, more precise steering / focusing can be achieved in real time by keeping the energy application device 12 in approximately the correct position (i.e., the treatment location 46). If the region of interest 44 moves out of the treatable zone 70, energy delivery is interrupted via the controller 16. An alarm or notification may be issued. The system 10 may be configured to wait to determine whether the region of interest 44 has returned to a position within the treatable zone 70 (based on image data acquired from the imaging transducer 68) before resuming. If the region of interest is determined not to be within the treatable zone 70 after a predetermined time has elapsed, instructions may be issued to move the energy application device 12 to another treatment location 46. In this manner, the energy application device is only moved when the region of interest 44 is determined not to be within the treatable zone 70, reducing operator effort and the likelihood of inaccurate positioning or repositioning of the energy application device 12. Furthermore, even slight inaccuracies in the position of the energy application device 12 may be corrected using neural networks or other techniques to identify the region of interest 44 within the treatable zone 70.

[0051] FIG. 6 is a flow diagram of a method 100 for neuromodulation energy delivery. In the description relating to the method 100, certain reference numbers described in FIGS. 1-5 may be referenced. The method 100 may be performed at the initiation or planning of a treatment protocol or as part of a treatment protocol validation. In certain embodiments, image data may be part of a patient validation procedure. In step 102, image data is acquired, for example, using the energy application device 12 in an imaging mode. In step 104, the image data is provided as an input for identifying a region of interest 44 from the image data. Once identified, in step 106, energy for neuromodulation is delivered to the region of interest, for example, by delivering energy from the energy application device through the patient's skin to reach the region of interest 44.

[0052] In step 108, system 10 may obtain updated image data showing movement of tissue 43 from an initial position and transitions between various positions (see FIG. 4, tissues 43a, 43b, 43c) resulting in movement of regions of interest 44b, 44c. The movement or change in position is identified in step 110 and modulation parameters of energy application device 23 are adjusted in step 112. In one embodiment, system 10 may use the movement of tissue 43 as a surrogate or estimate of movement of region of interest 44.

[0053] In one example, the updated image data may be evaluated as characteristic of a particular type of motion. For example, rhythmic or periodic movement of tissue 43 and / or region of interest toward and away from transducer 68 over a period of time (e.g., 1-5 seconds) may be characteristic of respiration. The system may create a model that predicts the movement of region of interest 44 over time to predict future respirations and deliver energy at a particular time to one or more predicted locations of region of interest 44 during respiration. In additional or alternative embodiments, system 10 may identify pauses or ends in respiration from the acquired image data to deliver energy to coincide with periods when the patient is pausing to breathe and the region of interest is relatively stationary.

[0054] In one embodiment, the system 10 can use the image data and the determined location of the region of interest 44 relative to the focal zone 48 to calculate the delivered dose over time. For example, in one embodiment, the energy application device 12 can adjust the steering and / or focusing of the energy delivery to a minimal extent while adjusting other parameters as the region of interest moves. Based on the movement of the region of interest 44 identified outside the focal zone 48, the system 10 can calculate the total delivered dose. Thus, if movement occurs, the system 10 can extend the dose delivery time or account for the period during energy delivery when the region of interest 44 is outside the focal zone 48 so that the total dose delivered directly to the region of interest 44 is within the desired parameter range. Additionally, the system 10 can also account for the total delivery to areas outside the region of interest 44 and adjust the steering and / or focusing when a threshold value is reached for the energy applied outside the region of interest 44. It should be appreciated that steering can change the angle of the ultrasound beam, while focusing can change the focal depth and / or overall size of the beam.

[0055] In certain embodiments, the system 10 uses a neural network to locate the region of interest 44 and / or the target tissue 43 within the acquired image data. In one embodiment, the location may be in a generally autonomous manner with minimal operator intervention. FIG. 7 is a schematic diagram of an embodiment of constructing a neural network for locating the region of interest 44 and / or the target tissue 43 within the acquired image data. The neural network 122 may be based on image data 120 acquired from the imaging probes 68 (e.g., 68a, 68b, 68c) of each of the various patients 118, including patients 118a, 118b, 118c that are not the patient 118 of interest (i.e., the patient receiving the dose). The neural network 122 receives the image data of the population and is trained by the image data of the population based on certain ground truth parameters. The neural network 122 may be specialized for a particular tissue or organ, or in certain embodiments, for a particular dosing regimen. The neural network 122 may be part of the controller 16 as shown. In other embodiments, the neural network 122 may be in communication with the controller 16, but is not necessarily part of the controller 16. A treatment probe, which may be configured as the energy applicator 12, is responsive to outputs from the controller 16 and the neural network. As illustrated in FIG. 8, at 124, the imaging probe 68 acquires image data from the patient 118, which is provided as an input to the neural network 122.

[0056] FIG. 8 is a flow diagram of a method 130 that may be performed in conjunction with certain elements depicted in FIGS. 1-7. The method includes providing image data of a population at step 132 to train a neural network. Method 130 also includes receiving image data of a patient of interest at step 134. The patient of interest may or may not be included in the population image data. At step 136, the trained neural network is used to identify regions of interest from the image data. In one embodiment, method 130 may include collecting images from a patient, annotating the images (annotations for supervised learning), and then feeding back the annotated images to update the network. The updated network is then applied to all subsequent images collected from the patient.

[0057] Since data annotation can be a tedious process, the method 130 may further include selecting a subset of all images collected from a patient, annotating them, and using them to update the network. The subset of images used to update the network may be selected based on various factors, one of which may be how well an existing population-based network model performs on the images. For example, if a population-based model already performs well on a given image, there may be little benefit in using that image to update the model. However, images for which the network provides poor results may be manually detected or automatically detected using another criterion (e.g., low probability score), and expert annotations for such images may be input to the network.

[0058] As described herein, the neural network 122 may include one or more layers that allow for the identification of organs and / or structures. To train the model for a particular individual, certain layers of the neural network 122 may be reserved or frozen, while other layers of the network are trained with data from the patient of interest. Large and deep networks are powerful, but such networks operate with large amounts of data. When there is only a limited amount of data, as may be the case for an individual patient, freezing certain layers in a population-based model can reduce the number of parameters that the network must learn. Since population-based models are also trained with similar images and / or similar tasks, weights and filters learned in layers closer to the input layer are typically at a low enough level (i.e., edges, lines, etc.) that retraining provides limited benefit. The neural network may be supervised or unsupervised. The neural network 122 may be updated to accommodate patient-specific changes in the patient. Additionally, the neural network 122 may include a validation step to assess the reliability of the recognition of the region of interest (see FIG. 15). In one embodiment, a patient-specific model may be validated with an independent dataset from that patient to ensure that the overall accuracy of the model is improved for that patient before being deployed to a treatment device.

[0059] 9 is an exemplary system 150 that can be used with or as part of the system 10. The system 150 includes various features that enable image acquisition, such as a dual-function probe 152 (see FIG. 10) including an imaging transducer 156, shown as a GE 3S sector array probe (General Electric), and a therapeutic probe 154, shown as a HIFU probe, and an image probe controller 155 that controls image acquisition via the imaging transducer 156. The system 150 further includes a frame grabber 157 that operates on acquired image data. The system 10 operates on images rendered from the acquired image data in certain embodiments, although it should be understood that raw image data or unrendered image data can also be used as input. The therapeutic probe 154 may be operated under the control of a therapeutic probe controller 162 that controls a pulse generation circuit 160 and an RF power amplifier 158.

[0060] FIG. 11 illustrates an example of a graphical user interface that may be used with system 10 and displays images acquired using system 150. The graphical user interface displays a neuromodulation prescription, which is a term describing a treatment protocol (including treatment dates or timing information for an individual patient) and / or target tissue, and may include, in certain embodiments, a total energy dose for each administration, and observable parameters related to the treatment. For example, the prescription may describe a concentration change target for a molecule of interest relative to a baseline value before treatment begins. The user interface may display a status of the procedure and acquired ultrasound images, such as acquired patient image data, and a caption provided by the neural network indicating anatomical structures detected from the images. The neural network identified the anatomical structure of the kidney. However, the associated treatment protocol is for autonomous neuromodulation energy delivery to the liver.

[0061] Thus, as shown in the exemplary graphical user interface of FIG. 12, when starting administration, the system waits until the target anatomical structure (in this case the liver) appears in the field of view before administering the treatment. The status is shown as "adjusting" since the probe placement identified is not suitable for administering the treatment. In FIG. 13, once the location of the anatomical structure present in the field of view identified by the neural network is determined to match the location of the anatomical structure of interest, the treatment energy or dose of neuromodulation energy is delivered. The status is shown as "delivering" and the treatment beam is visually displayed on the ultrasound image. In FIG. 14, the target organ (i.e. the liver) is in focus but the system has stopped delivering the treatment and the total energy of the individual dose has been reached as shown in this frame.

[0062] 15 shows results from a neural network trained for organ detection to identify and locate the spleen, kidneys, and liver on an ultrasound image. The captured image includes a particular probability indication (e.g., 99%, 97%). In one embodiment, an organ or region of interest within an organ may be identified when a probabilistic threshold is reached based on the output of the neural network. In one example, the threshold may be at least 95%, at least 97%, or at least 99%.

[0063] The disclosed techniques allow for delivery of neuromodulation energy that takes into account movement of the desired region of interest. The movement may be during treatment, for example as a result of respiration or blood flow. The movement may also be a change in position or change in size of the organ when administration is performed at intervals. For example, the size or depth of the organ may change due to weight loss or clinical symptoms of the patient. Such changes can then be evaluated to allow for more accurate delivery of neuromodulation energy.

[0064] Examples are used herein to disclose the invention, including the best mode, and to enable those skilled in the art to practice the invention, including making and using any device or system, and performing any incorporated methods. The patentable scope of the invention is defined by the claims, and may include other examples that occur to those skilled in the art. Such other examples are intended to be within the scope of the claims if they have structural elements that do not differ from the literal language of the claims, or if they include equivalent structural elements that do not differ substantially from the literal language of the claims.

Claims

1. A pulse generator; an ultrasound energy application device configured to receive energy pulses from the pulse generator and deliver a neuromodulation energy dose to a region of interest of a patient; a controller in communication with the pulse generator; The controller: receiving image data of an internal tissue of the patient; activating a neural network trained on images of internal tissues of each patient in the population to identify the presence and location of the region of interest within the received image data; generating a probabilistic representation of the presence based on output from the neural network; controlling the pulse generator to initiate delivery of application of the neuromodulation energy to the region of interest via the energy application device when the region of interest is present in the image data; receiving updated image data of the internal tissue of the patient during delivery of the dose; identifying a change in position of the region of interest relative to the energy application device based on the updated image data using the neural network; and configured to discontinue delivery of energy based on a determination that the region of interest is not present in the updated image data; A neuromodulation system, wherein the presence of the region of interest is identified when a probabilistic threshold is reached based on the output of the neural network.

2. 2. The system of claim 1, wherein the controller is configured to predict a path of movement of the area of ​​interest using the neural network and dynamically modify the one or more control parameters based on the predicted path of movement.

3. The system of claim 1 , wherein the neural network is trained to identify a region of interest that includes the hilum of the liver.

4. The system of claim 1 , wherein the controller is configured to time delivery of the application of the neuromodulation energy at a period expected to be between breaths of the patient.

5. The system of claim 1 , wherein the energy application device is controlled such that the neuromodulation energy is not applied to areas outside the region of interest.

6. The system of claim 1 , wherein the controller is configured to track the region of interest using the neural network and automatically adjust the energy application device based on tracked movement.

7. The system of claim 6, wherein the automatic adjustment includes reducing the output of the energy pulse based on the area of ​​interest moving closer to the energy application device or increasing the output of the energy pulse based on the area of ​​interest moving away from the energy application device.

8. The system of claim 1 , wherein the controller is configured to automatically control the energy application device to steer and / or focus an ultrasound beam formed by the ultrasound energy application device to the region of interest identified by the neural network.

9. 2. The system of claim 1, wherein the ultrasound energy applicator comprises a motor configured to change a position or angle of the energy applicator and / or a transducer of the energy applicator relative to the patient in response to instructions from the controller to align the beam of neuromodulation energy to the location of the region of interest.

10. The system of claim 1 , comprising an imaging transducer configured to acquire the image data and the updated image data.

11. The system of claim 10 , wherein the imaging transducer forms part of the energy application device such that the imaging transducer is in contact with the patient during application of the neuromodulation energy.

12. The system of claim 1 , wherein the neural network is trained using image data from the patient, or both image data from the patient and the updated image data.

13. The system of claim 1 , wherein the region of interest is defined by the presence of a particular anatomical structure within the patient.

14. 14. The system of claim 13, wherein the anatomical structures include one or more of an organ, a nerve, a nerve plexus, a blood vessel, and the neural network is trained to identify the anatomical structures within or adjacent to the region of interest.

15. The system of claim 13 , wherein the anatomical structure is within the liver or spleen.

16. The system of claim 13 , wherein the neural network includes one or more layers configured to identify the anatomical structures.

Images