Adaptive neuromodulation system and method

Abstract

The invention relates to a neuromodulation system (1) for changing the brain state of an individual (5) to a target brain state during a single working phase, said system (1) comprising a brain monitoring module (10), a processing module (20) and a transcranial ultrasound stimulation module (30) configured to periodically stimulate at least one brain region of the individual (5) using transcranial ultrasounds according to a transcranial ultrasound stimulation configuration generated by the processing module (20).

Description

Adaptive neuromodulation system and method

[0001] The invention relates to the field of neuromodulation and, more particularly, to a system and a method for adaptively changing the brain state of a patient.

[0002] Neuromodulation technologies are increasingly used to treat mental health disorders but generally target only one brain region, varying between individuals, to achieve a reduction in symptoms. For example, a solution called Transcranial Magnetic Stimulation (TMS) uses a Magnetic Resonance Imaging scanner to scan the brain of an individual to determine a brain region to be stimulated in a first step and a stimulation module to stimulate the identified region in a second step. Patent application US20140194726A1 describes a transcranial ultrasound neuromodulation system for cognitive enhancement, using ultrasound to stimulate brain regions non-invasively. Patent application WO2014176483A1 describes a focused transcranial ultrasound system for precise brain stimulation.

[0003] In the existing methods, data from the brain are collected by a brain monitoring module in a preliminary phase then a target zone to be stimulated is identified manually by an operator in a subsequent analysis phase then the identified target zone is stimulated in a subsequent and final phase. However, these methods may fail or not be efficient as the brain state of an individual changes over time, sometimes rapidly, making the stimulation inadequate to the scanner monitoring if the brain state has changed in between.

[0004] It is therefore an object of the present invention to provide a device and method to remedy at least partly these drawbacks.

[0005] It is an object of the present invention to provide a system and method that allows to change the brain state of an individual in real time.

[0006] To this end, the present invention concerns a neuromodulation system for changing the brain state of an individual to a target brain state during a single working phase, said system comprising:

[0007] - a brain monitoring module configured to continuously collect brain activity data during said working phase,

[0008] - a processing module configured to:

[0009] - periodically receive said collected brain activity data,

[0010] - iteratively run a brain state transition model using the periodically collected data to iteratively generate a new transcranial ultrasound stimulation configuration,

[0011] - iteratively control the transcranial ultrasound stimulation module to work according to said new transcranial ultrasound stimulation configuration,

[0012] said brain state transition model using historical brain transition data and being configured to determine the current brain state of the individual based on the collected brain data or a brain state metric and to determine the new transcranial ultrasound stimulation configuration that allows the brain to move from the determined current brain state toward the target brain state based, or towards a specific value or direction on the brain state metric, using said historical brain transition data,

[0013] - a transcranial ultrasound stimulation module configured to periodically stimulate at least one brain region of the individual using transcranial ultrasounds according to the new transcranial ultrasound stimulation configuration, advantageously to generate a pressure wave customized in location, duration, intensity, duty cycle and frequency for each stimulated brain region.

[0014] By “transcranial ultrasound stimulation configuration”, it is meant the tuning of the parameters that allow the transcranial ultrasound stimulation module to stimulate the brain or a region of the brain like, for example the frequency of the transcranial ultrasounds, duty cycle, the intensity of the transcranial ultrasounds, the pulse duration, pulse target location of the transcranial ultrasounds or any other relevant parameters used to determine the state of the brain or brain region which is stimulated.

[0015] The system according to the invention creates a feedback mechanism that dynamically refines the transcranial ultrasound stimulation configuration, unlike static, predefined protocols in existing technologies. The system does not rely on binary "on / off" state changes. Instead, it transitions brain states progressively, allowing users to transition towards higher complexity or higher entropy brain states (reducing pain or suffering, enhancing cognitive performance, reaching an egoless state) in a continuous and incremental way. Because existing neuromodulation systems lack personalization and adaptivity, incorporating a brain state transition model that anticipates and fine-tunes stimulation parameters in real-time, based on individual responses and shifting brain dynamics, transforms the system into a self-optimizing, responsive technology fundamentally distinct from conventional approaches. The system uniquely caters to a broad spectrum of applications like e.g., mental health treatment, cognitive enhancement, athletic training, and Brain Computer Interfaces (BCI), providing a unified solution to diverse challenges. This multi-domain applicability distinguishes it from niche, single-purpose systems.

[0016] Preferably, once the target brain state is reached or the closest possible brain state to the target brain state, the neuromodulation system maintains that brain state, possibly using the same or a new configuration of the transcranial ultrasound stimulation module, de facto allowing the brain to learn and stabilize said brain state by itself.

[0017] Preferably, the brain monitoring module is configured to collect brain activity data using Magnetic Resonance Imaging (MRI).

[0018] Preferably, the brain monitoring module is configured to collect brain activity data using Functional Magnetic Resonance Imaging (fMRI).

[0019] Preferably, the brain monitoring module is configured to collect brain activity data using Functional Magnetic Resonance Imaging (fMRI) measuring the Blood Oxygen Level Dependent (BOLD) signal.

[0020] Preferably, the brain state transition model is a pretrained artificial intelligence neural network model or a predictive model (for example based on Model Predictive Control or MPC).

[0021] Advantageously, the brain state transition model is a deep learning model trained on diverse datasets of brain state transitions as such reinforcement learning allows to adapt stimulation protocols dynamically based on real-time feedback.

[0022] Advantageously, the brain state transition model integrates user-specific historical data to enhance long-term personalization.

[0023] In an embodiment, the brain monitoring module is configured to collect the brain activity data using Magnetic Resonance Imaging. Such functional imaging validates stimulation effects and provides additional precision for identifying brain state transitions.

[0024] The combination of transcranial ultrasound stimulation and MRI hardware makes the system non-invasive.

[0025] In another embodiment, the brain monitoring module is configured to collect the brain activity data using EEG (Electroencephalography), MEG (Magnetoencephalography), OPM-MEG (Optically Pumped Magnetometer MEG), NIRS (Near-Infrared Spectroscopy), PET (Positron Emission Tomography), SPECT (Single Photon Emission Computed Tomography), Magnetic Particle Imaging (MPI), Ultrasound Neuroimaging (Functional Ultrasound) or SQUID-Based MEG. Measurement systems like EEG adds portability, increasing its accessibility compared to cumbersome MRI-only systems or invasive neuromodulation methods.

[0026] Advantageously, the transcranial ultrasound stimulation module comprises wearable or portable focused ultrasound transducers that allow to deliver non-invasive, high-precision stimulation to specific brain regions.

[0027] Preferably, the brain state transition model has been trained according to various brain states. The brain state transition model is configured to identify brain state profiles and target states of the individual and store historical transition data associated with transition from an original brain state to a target brain state. When the brain state transition model is a pretrained artificial intelligence neural network model, the storage of historical data is not explicit like a database but implicit in that it is a discrete part of the AI model.

[0028] In an embodiment, the target brain state is a high complexity brain state. The complexity of the brain may be determined based on Lempel-Ziv, fractal dimension or connectome network connectivity criteria in a way known from the skilled person. The threshold of high complexity may then be defined by the skilled person in an known manner, for example based on one of these criteria.

[0029] Preferably, the brain state transition model is configured to determine target brain state of high entropy or high complexity to improve the information processing of the brain. Shifting the brain towards a high complexity or high entropy state has been shown to correlate with enhanced well-being, potentially healing mental health disorders.

[0030] In an embodiment, the transcranial ultrasound stimulation module and the brain monitoring module and the processing module are embedded into two devices or a single device.

[0031] The invention also relates to a neuromodulation method for changing the brain state of an individual to a target brain state using a system as described here before, said method comprising, in a single working phase, several iterations of the steps of:

[0032] - collecting brain activity data,

[0033] - determining the current brain state of the individual based on said collected brain activity data,

[0034] - running a brain state transition model using the periodically collected data and historical brain transition data to iteratively generate a new transcranial ultrasound stimulation configuration that allows the brain to move from the determined current brain state toward the target brain state,

[0035] - controlling the transcranial ultrasound stimulation module to work according to said new transcranial ultrasound stimulation configuration,

[0036] - stimulating at least one brain region using transcranial ultrasounds according to the determined transcranial ultrasound stimulation configuration to generate brain data for each stimulated brain region.

[0037] Advantageously, the brain state transition model is a pretrained artificial intelligence neural network model or a predictive model.

[0038] Preferably, the step of collecting the generated data is realized using Magnetic Resonance Imaging.

[0039] Preferably, the step of collecting the generated data is realized using EEG (Electroencephalography), MEG (Magnetoencephalography), OPM-MEG (Optically Pumped Magnetometer MEG), NIRS (Near-Infrared Spectroscopy), PET (Positron Emission Tomography), SPECT (Single Photon Emission Computed Tomography), Magnetic Particle Imaging (MPI), Ultrasound Neuroimaging (Functional Ultrasound) or SQUID-Based MEG.

[0040] In an embodiment, the step of stimulating is realized using wearable or portable focused ultrasound transducers.

[0041] Advantageously, the brain state transition model has been trained according to various brain states.

[0042] Advantageously, the brain state transition model identifies brain state profiles and target states of the individual and stores historical transition data associated with transition from an original brain state to a target brain state.

[0043] These and other features, aspects, and advantages of the present invention are better understood with regard to the following Detailed Description of the Preferred Embodiments, appended Claims, and accompanying Figures, where:

[0044] schematically illustrates an embodiment of the system according to the invention.

[0045] schematically illustrates an embodiment of the method according to the invention.

[0046] illustrates an example of neuromodulation system 1 according to the invention. The system 1 allows to change the brain state of an individual 5 during a unique, one and same single working phase.

[0047] For example, a change of brain state may be useful for mental health treatment to addresses psychiatric conditions like depression, PTSD, and anxiety by targeting relevant neural circuits. For example, a change of brain state may be useful for athletic brain training to enhance neuroplasticity, focus, and reaction times to improve athletic performance. For example, a change of brain state may be useful for meditation and cognitive enhancement to guide users into optimized states of relaxation, creativity, or cognitive flow. For example, a change of brain state may be useful for brain-Computer Interfaces (BCIs) to improve BCI signal clarity and user engagement by fostering ideal neural conditions.

[0048] The system 1 comprises a brain monitoring module 10, a processing module 20 and a transcranial ultrasound stimulation module 30. The brain monitoring module 10, the processing module 20 and the transcranial ultrasound stimulation module 30 may be embedded into two separate devices or into one single device.

[0049] The brain monitoring module 10 is configured to continuously collect the brain activity data of the individual 5 during the working phase.

[0050] The brain monitoring module 10 is preferably a Magnetic Resonant Imaging device as such functional imaging validates stimulation effects and provides additional precision for identifying brain state transitions.

[0051] Alternatively, other devices to collect brain data may be used such as, for example, EEG (Electroencephalography), MEG (Magnetoencephalography), OPM-MEG (Optically Pumped Magnetometer MEG), NIRS (Near-Infrared Spectroscopy), PET (Positron Emission Tomography), SPECT (Single Photon Emission Computed Tomography), Magnetic Particle Imaging (MPI), Ultrasound Neuroimaging (Functional Ultrasound), SQUID-Based MEG or any other relevant device.

[0052] The processing module 20 is preferably a computer or a portable device such as a tablet or a smartphone.

[0053] The processing module 20 is configured to periodically receive the brain activity data collected by the brain monitoring module 10.

[0054] The processing module 20 is configured to iteratively run a brain state transition model using the periodically collected data to iteratively generate a new transcranial ultrasound stimulation configuration. A transcranial ultrasound stimulation configuration is a set of transcranial ultrasound stimulation values related to the one or several stimulation parameters.

[0055] The brain state transition model uses historical brain transition data and is configured to determine the current brain state of the individual 5 based on the collected brain data or a brain state metric and to determine the new transcranial ultrasound stimulation configuration that allows the brain to move from the determined current brain state toward the target brain state based, or towards a specific value or direction on the brain state metric, using said historical brain transition data.

[0056] More particularly, the brain state transition model is configured to determine a brain state transition from the determined current brain state to or at least toward a new brain state (like a target brain state). The brain state transition model is configured to generate the new set of transcranial ultrasound stimulation parameters using said determined brain state transition.

[0057] Preferably, the brain state transition model is a pretrained artificial intelligence neural network model or a predictive model, for example a Model Predictive Control (MPC). Dynamically adjusting stimulation parameters in response to individual 5 responses enable continuous brain state optimization.

[0058] The AI model may be a deep learning model trained on diverse datasets of brain state transitions to adapt stimulation protocols dynamically based on real-time feedback.

[0059] The processing module 20 is configured to iteratively control the transcranial ultrasound stimulation module 30 to work according to a generated transcranial ultrasound stimulation configuration and generate changes of brain states towards the desired brain state during the working phase.

[0060] The system 1 may advantageously be calibrated. For example, brain activity data may be collected to calculate the attenuation of sound waves inside the skull and brain. This calibration models the pressure wave distortion dependent on the target location to adapt the pressure wave control to reach the desired target location with the desired intensity, frequency and duty cycle.

[0061] Several brain states profiles may be personalized for a given individual 5. Such user-specific historical data allows to enhance long-term personalization and improve stimulation to reach faster the desired state of the brain.

[0062] Preferably, for example in the case of mental health treatment, target brain state of high entropy and / or high complexity is desired as they improve the information processing of the brain. Shifting the brain towards a high complexity or high entropy state has been shown to correlate with enhanced well-being, potentially healing mental health disorders.

[0063] The transcranial ultrasound stimulation module 30 is configured to periodically stimulate at least one brain region of the individual 5 using transcranial ultrasounds according to a predetermined or a generated transcranial ultrasound stimulation configuration, during the working phase, to reach or approach in the closest manner the targeted brain state.

[0064] Preferably, the transcranial ultrasound stimulation module 30 comprises wearable or portable focused ultrasound transducers to deliver non-invasive, high-precision stimulation to specific brain regions.

[0065] The stimulation is realized according to a predetermined set of transcranial ultrasound stimulation values related to predetermined parameters. For example, the parameters may be the frequency of the transcranial ultrasounds, the intensity of the transcranial ultrasounds, the pulse duration transcranial ultrasounds or any other relevant parameters used to determine the state of the brain region which is stimulated.

[0066] Example of operation

[0067] As a prerequisite, the brain state transition model is trained on a plurality of brain states with a plurality of individuals, preferably over 1 000 or 10 000 or 100 000 or 1 000 000 individuals. This training consists, for each individual, in stimulating the brain of the individual with ultrasound signals having specific parameters in a given brain state, for example during rest or during a specific activity or mood like doing a given sport, playing chess, playing the piano, feeling depressed, etc., while measuring at the same time several parameters of the brain or of brain regions which are associated with said given brain state. Such measurements allow to generate an average profile for each brain state and identify brain states that help conditioning for a given task or for mental health treatment like brain states with high entropy or high complexity. Measurements are also realized while changing from one given brain state to another to determine which parameters and which values of said parameters change and how they change. This allows to further activate the right parameters with the right values (i.e., to use adequate transcranial ultrasound stimulation signals) to realize a transition between from one current state to a target state or desired state (e.g., of more complexity).

[0068] Once the brain state transition model has been sufficiently trained. The method can be implemented on a given individual 5 during a single working phase, where the following steps S1-S5 are iterated. Preferably, a working phase lasts for a duration comprised for example between 30 and 180 min.

[0069] In a step S1, the brain monitoring module 10 collects, for example during one minute, brain activity data of the individual 5, which are provided as input to the brain state transition model of the processing module 20 to determine the brain state of the individual 5 in a step S2.

[0070] In a step S3, the brain state transition model of the processing module 20 generates, based on its preliminary training, a new transcranial ultrasound stimulation configuration to allow a transition from the current brain state of the individual 5 toward the target brain state.

[0071] In a step S4, the processing module 20 controls the transcranial ultrasound stimulation module 30 to stimulate, in a step S5, for example during one minute, the brain or selected brain regions a given brain region using the new set of transcranial ultrasound stimulation values of selected parameters (or of new parameters), i.e. the new transcranial ultrasound stimulation configuration, to move the brain state of the individual 5 toward the target brain state.

[0072] More precisely, the transcranial ultrasound stimulation module 30 is used to stimulate at least one brain region of the individual 5 using transcranial ultrasounds according to a predetermined set of transcranial ultrasound stimulation parameters and associated values to generate brain activity data for each stimulated brain region. The predetermined set of transcranial ultrasound stimulation parameters and associated values (transcranial ultrasound stimulation configuration) are chosen by the brain state transition model to move the brain toward a target brain state, which has been chosen by the operator of the system 1, like for example reaching brain state with high entropy or high complexity.

[0073] The method is iterated from step S1 as many times as needed during the working phase to reach the target brain state or the closest possible brain state to the target brain state. The neuromodulation system 1 will then maintain that brain state, de facto allowing the brain to learn and stabilize said brain state by itself.

[0074] The neural network model or predictive model analyzes the brain data and creates new stimulation parameters at each iteration. The real-time or near real-time brain data analysis enables the model to continuously and iteratively adjust stimulation parameters to transition from existing brain states towards a target brain state. The closed-loop control system allows to continuously monitor brain activity and refine stimulation parameters in real-time, stops or modifies stimulation upon achieving the desired brain state or detecting unintended responses. Adjustable parameters (e.g., frequency, intensity, pulse duration) allow fine control, enhancing both safety and efficacy.

[0075] In case where the target brain state main is not determined, the actual brain activity is measured, then the brain complexity is calculated, then the transcranial ultrasound stimulation module 30 is configured to increase brain complexity based on historical data, then new configurations (i.e. parameters) of the transcranial ultrasound stimulation module 30 are generated by iteration until the maximum of brain complexity is reached or a certain threshold / value is reached.

[0076] The use of iterations (for example one minute of brain data collection alternated with one minute of brain stimulation) during a single working phase for collecting and processing brain activity data and adapting the stimulation to new brain activity data in real-time allows to converge rapidly and efficiently toward the desired target brain state.

[0077] The Specification, which includes the Summary of Invention, Brief Description of the Drawings and the Detailed Description of the Preferred Embodiments, and the appended Claims refer to particular features (including process or method steps) of the invention. Those of skill in the art understand that the invention includes all possible combinations and uses of particular features described in the Specification. Those of skill in the art understand that the invention is not limited to or by the description of embodiments given in the Specification but defined by the claims.

Claims

A neuromodulation system (1) for changing the brain state of an individual (5) to a target brain state during a single working phase, said system (1) comprising:- a brain monitoring module (10) configured to continuously collect brain activity data during said working phase,- a processing module (20) configured to:- periodically receive said collected brain activity data,- iteratively run a brain state transition model using the periodically collected data to iteratively generate a new transcranial ultrasound stimulation configuration,- iteratively control the transcranial ultrasound stimulation module to work according to said new transcranial ultrasound stimulation configuration,said brain state transition model using historical brain transition data and being configured to determine the current brain state of the individual (5) based on the collected brain data or a brain state metric and to determine the new transcranial ultrasound stimulation configuration that allows the brain to move from the determined current brain state toward the target brain state based, or towards a specific value or direction on the brain state metric, using said historical brain transition data,- a transcranial ultrasound stimulation module (30) configured to periodically stimulate at least one brain region of the individual (5) using transcranial ultrasounds according to the new transcranial ultrasound stimulation configuration.The system (1) according to claim 1, wherein the brain state transition model is a pretrained artificial intelligence neural network model or a predictive model.The system (1) according to any of the preceding claims, wherein the brain monitoring module (10) is configured to collect the brain activity data using Magnetic Resonance Imaging.The system (1) according to any of claims 1 to 3, wherein the brain monitoring module (10) is configured to collect the brain activity data using EEG (Electroencephalography), MEG (Magnetoencephalography), OPM-MEG (Optically Pumped Magnetometer MEG), NIRS (Near-Infrared Spectroscopy), PET (Positron Emission Tomography), SPECT (Single Photon Emission Computed Tomography), Magnetic Particle Imaging (MPI), Ultrasound Neuroimaging (Functional Ultrasound) or SQUID-Based MEG.The system (1) according to any of the preceding claims, wherein the transcranial ultrasound stimulation module (30) comprises wearable or portable focused ultrasound transducers.The system (1) according to any of the preceding claims, wherein the brain state transition model is configured to identify brain state profiles and target states of the individual (5) and store historical transition data associated with transition from an original brain state to a target brain state.The system (1) according to any of the preceding claims, wherein the target brain state is a high complexity brain state.The system (1) according to any of the preceding claims, wherein the brain monitoring module (10), the processing module (20) and the transcranial ultrasound stimulation module (30) are embedded into two devices or a single device.A neuromodulation method for changing the brain state of an individual (5) to a target brain state using a system (1) according to any of the preceding claims, said method comprising, in a single working phase, several iterations of the steps of:- collecting (S1) brain activity data,- determining (S2) the current brain state of the individual (5) based on said collected brain activity data,- running (S3) a brain state transition model using the periodically collected data and historical brain transition data to iteratively generate a new transcranial ultrasound stimulation configuration that allows the brain to move from the determined current brain state toward the target brain state,- controlling (S4) the transcranial ultrasound stimulation module to work according to said new transcranial ultrasound stimulation configuration,- stimulating (S5) at least one brain region using transcranial ultrasounds according to the determined transcranial ultrasound stimulation configuration to generate brain data for each stimulated brain region.The method according to the preceding claim, wherein the brain state transition model is a pretrained artificial intelligence neural network model or a predictive model.The method according to any of claims 9 or 10, wherein the step of collecting the generated data is realized using Magnetic Resonance Imaging.The method according to any of claims 9 to 10, wherein the step of collecting the generated data is realized using EEG (Electroencephalography), MEG (Magnetoencephalography), OPM-MEG (Optically Pumped Magnetometer MEG), NIRS (Near-Infrared Spectroscopy), PET (Positron Emission Tomography), SPECT (Single Photon Emission Computed Tomography), Magnetic Particle Imaging (MPI), Ultrasound Neuroimaging (Functional Ultrasound) or SQUID-Based MEG.The method according to any of claims 9 to 12, wherein the step of stimulating is realized using wearable or portable focused ultrasound transducers.The method according to any of claims 9 to 13, wherein the brain state transition model has been trained according to various brain states.The method according to any of claims 9 to 14, wherein the brain state transition model identifies brain state profiles and target states of the individual (5) and stores historical transition data associated with transition from an original brain state to a target brain state.

Images