Pseudo sensation provision device and motion inducing device

Abstract

[Problem] To provide a pseudo sensation provision device and the like. [Solution] A pseudo sensation provision device for providing a pseudo sensation to a subject comprises: a brain stimulation unit for stimulating the brain of the subject; and a stimulation information control unit for obtaining stimulation-unit control information for controlling the brain stimulation unit. The stimulation information control unit: has a pseudo sensation model that is a trained model constructed by using, as training data, stimulation information to the brain of a subject organism and sensation information of the subject organism; and outputs output stimulation information when sensation information for providing a pseudo sensation to the subject has been inputted to the pseudo sensation model. The stimulation information control unit obtains the stimulation-unit control information by using the output stimulation information.

Description

Apparent sensation providing device and movement inducing device

[0001] The present invention relates to an apparent sensation providing device and a movement inducing device. More specifically, the present invention relates to an apparent sensation providing device and a movement inducing device that can induce desired apparent sensations and movements by learning brain stimulation and the apparent sensations and movements caused thereby through machine learning and using the learned model obtained as a result.

[0002] Japanese Patent No. 3718500 describes a probe and a device for measuring the hemodynamic and oxygen processing characteristics of the brain. This device uses a probe having an optical fiber. Therefore, this device can only measure brain activity at a site where the optical fiber is laid.

[0003] Japanese Patent No. 3718500

[0004] An object of the present invention is to provide an apparent sensation providing device and a movement inducing device that can induce desired apparent sensations and movements by learning brain stimulation and the apparent sensations and movements caused thereby through machine learning and using the learned model obtained as a result.

[0005] The above problems are basically based on the finding that desired apparent sensations and movements can be induced by learning brain stimulation and the apparent sensations and movements caused thereby through machine learning and using the learned model obtained as a result.

[0006] The first invention relates to a pseudo-sensory provision device for providing a subject with a pseudo-sensory experience. This pseudo-sensory provision device comprises a brain stimulation unit for stimulating the subject's brain and a stimulus information control unit for obtaining stimulus unit control information to control the brain stimulation unit. The stimulus information control unit has a pseudo-sensory model, which is a trained model constructed using stimulus information to the brain of a target organism and sensory information of the target organism as training data. When sensory information for providing a pseudo-sensory experience to the subject is input to the pseudo-sensory model, the control unit outputs output stimulus information. The stimulus information control unit is an element for obtaining stimulus unit control information using the output stimulus information. Based on the stimulus unit control information, the brain stimulation unit stimulates the subject's brain, thereby providing the subject with a pseudo-sensory experience.

[0007] The senses are preferably those that include one or more of sight, hearing, touch, taste, smell, memory, consciousness, intention, and understanding. Another example of a sense is that it includes one or more of sight, hearing, touch, taste, and smell. Another example of a sense is that it includes sight.

[0008] The second invention relates to a motor induction device. This motor induction device includes a brain stimulation unit for stimulating the brain of a subject, and a stimulus information control unit for obtaining stimulus unit control information for controlling the brain stimulation unit. The stimulus information control unit has a brain stimulation induction motor model, which is a trained model constructed using stimulation information to the brain of a target organism and motor information of the target organism as training data. When motor information for causing the subject to perform a movement is input to the brain stimulation induction motor model, it outputs output stimulus information, and the stimulus information control unit obtains stimulus unit control information using the output stimulus information.

[0009] This specification describes a nerve stimulator. Another preferred use of the nerve stimulator is to provide a subject with a simulated sensation or to induce a subject to perform a movement based on nerve stimulation.

[0010] This invention provides a device that can induce desired simulated sensations or movements by learning brain stimulation and the resulting simulated sensations or movements through machine learning, and then using the resulting trained model.

[0011] Figure 1 is a block diagram illustrating a brain activity measurement device. Figure 2 is a conceptual diagram showing an example of the placement of a brain diagnostic probe. Figure 3 is a diagram showing an example of a sheet-like probe section. Figure 4 is a conceptual diagram showing how the first detection unit detects fluorescence emitted by a fluorescent substance. Figure 5 is a conceptual diagram showing an example of the circuit design on a sheet-like probe section having first and second light sources. Figure 6 is a block diagram illustrating a nerve stimulator. Figure 7 is a conceptual diagram illustrating a stimulation model. Figure 8 is a conceptual diagram illustrating a simulated sensory provision device. Figure 9 is a diagram showing an example of how simulated vision, which is a simulated sensory experience, is formed based on an image imagined by the user. Figure 9(a) shows the outline of a cat, Figure 9(b) shows a cat, and Figure 9(c) shows a landscape including a cat. Figure 10 is a conceptual diagram illustrating a motor induction device.

[0012] The embodiments for carrying out the present invention will be described below with reference to the drawings. The present invention is not limited to the embodiments described below, but also includes modifications made to the embodiments described below to the extent that is obvious to those skilled in the art.

[0013] Figure 1 is a block diagram illustrating a brain activity measurement device. As shown in Figure 1, the brain activity measurement device 1 comprises a brain diagnostic probe 3, a detection signal receiving unit 5, and a signal analysis unit 7. The brain diagnostic probe 3 is inserted into the skull (the space between the skull and the surface of the brain, the subdural space) and is an element for performing brain diagnosis. The brain diagnostic probe 3 is publicly known, as described, for example, in Japanese Patent Publication No. 3718500 and Japanese Patent Publication No. 5224482.

[0014] Figure 2 is a conceptual diagram showing an example of the placement of a brain diagnostic probe. As shown in Figure 2, the brain diagnostic probe 3 has multiple sheet-like probe sections 9. Each of the multiple sheet-like probe sections 9 is placed inside the skull 11 of the target organism through a hole 13 provided in the skull 11. There may be one hole 13 or two or more. The hole 13 may be provided on the side of the skull, on the back of the skull, or on the top of the skull. In the example in Figure 2, two holes are provided. Examples of the number of sheet-like probe sections 9 include two to 100 sheets, three to 50 sheets, or four to 20 sheets. Examples of target organisms are animals, and preferred examples of target organisms are humans or non-human mammals. Among these, humans are preferred as the target organism.

[0015] The multiple sheet-like probe portions 9 are preferably shaped to cover an area of ​​1% to 90% of the target organism's brain. Here, the brain area refers to the area between the cerebrum and the skull. The brain area covered by the multiple sheet-like probe portions 9 may be 5% to 90%, 10% to 90%, 15% to 80%, 20% to 50%, 20% to 40%, or 30% to 70%.

[0016] Figure 3 shows examples of sheet-like probe portions. Figure 3(a) shows a sheet-like probe having a rectangular shape, Figure 3(b) shows a sheet-like probe with a rounded tip, Figure 3(c) shows a sheet-like probe with chamfered corners at the tip, and Figure 3(d) shows a sheet-like probe that narrows towards the tip and has a rounded tip. As shown in Figure 3, the shape of the sheet-like probe portion 9 may be strip-shaped (rectangular), or it may have a tapered tip to facilitate insertion into the brain. The shape of the sheet-like probe portion 9 may have a rounded tip or chamfered corners at the tip. It is preferable that the sheet-like probe portion 9 is a flexible sheet. Furthermore, it is assumed that the sheet-like probe portion 9 will be laid between the skull and the brain surface, and it is particularly preferable that it be laid along the cerebral cortex. Therefore, it is preferable that the sheet-like probe portion 9 has excellent biocompatibility. The sheet-like probe portion 9 preferably has a certain degree of flexibility and rigidity, and while it has the property of adhering to moisture, it is also preferable that it does not adhere to the brain surface and can be removed from the brain surface.

[0017] Preferably, each of the multiple sheet-like probe sections has an undulation that corresponds to either or both the undulation of the brain parenchyma and / or vascular structure of the area where it is laid. The brain has various parts, including, for example, the frontal lobe, parietal lobe, temporal lobe, and occipital lobe, and sheet-like probe sections 9 may be laid in each of these areas. The sheet-like probe section 9 may also be shaped to match the area where it is laid. An example of the width W (length of the widest part) of the sheet-like probe section 9 is 0.5 cm or more and 6 cm or less, but it may also be 1 cm or more and 5 cm or less, 1 cm or more and 4 cm or less, 1 cm or more and 3 cm or less, or 2 cm or more and 4 cm or less. An example of the length L (length of the longest part) of the sheet-like probe section 9 is 2 cm or more and 20 cm or less, but it may also be 4 cm or more and 15 cm or less, or 5 cm or more and 10 cm or less.

[0018] For example, each of the multiple sheet-like probe sections 9 has a predetermined location where it is to be laid. For example, each sheet-like probe section 9 preferably has a shape and properties suitable for being laid on at least one of the following areas of the target organism: the cerebral cortex, ventricles, subdural space, subarachnoid space, ventricular wall, and sulcus. Furthermore, each of the multiple sheet-like probe sections 9 has undulations (irregularities) corresponding to either or both of the undulations and vascular structures of the brain parenchyma in the area where it is to be laid, so that it can get as close as possible to the brain parenchyma and adhere closely. For example, the undulations and vascular structures of the brain parenchyma may be examined in advance using CT or MRI, and the undulations shape may be custom-made according to that shape. In this example, the process may include a brain surface shape acquisition step, which is a step of acquiring either or both (brain surface undulation information) of the undulations and vascular structures of the brain parenchyma in the area of ​​the target organism that includes the area where the sheet-like probe section 9 is to be laid, and a surface adjustment sheet-like probe section creation step, which is a step of obtaining a sheet-like probe section 9 having a surface shape based on the brain surface undulation information. The brain surface shape acquisition process can be carried out using methods that can acquire the shape of the brain, such as CT or MRI as described above. The obtained brain surface relief information can then be stored in a memory unit as appropriate. Next, in the surface adjustment sheet-like probe unit creation process, the brain surface relief information can be read from the memory unit and the surface shape of the sheet-like probe unit can be adjusted. For example, after setting the elements and circuits necessary to function as a sheet-like probe unit on a flexible substrate, the surface can be coated with resin to create the sheet-like probe unit 9. When coating the flexible substrate with resin, the surface shape of the sheet-like probe unit 9 can be adjusted based on the brain surface relief information to obtain a sheet-like probe unit 9 whose surface shape corresponds to the relief of the brain parenchyma and vascular structure. Such processing can be achieved, for example, by using a 3D printer. Of course, after forming the resin layer on the flexible substrate, the resin layer can be processed based on the brain surface relief information to obtain a sheet-like probe unit 9 whose surface shape corresponds to the relief of the brain parenchyma and vascular structure.

[0019] A flexible printed circuit board may be used for the sheet-shaped probe. From the viewpoint of biocompatibility, a flexible printed circuit board with copper foil attached to its surface is preferred. Examples of such flexible printed circuit boards are described in Japanese Patent No. 7194857 and Japanese Patent No. 7164752. With a flexible printed circuit board, the light source, detection unit, control unit (and memory unit), etc., described later can be installed on the board while maintaining flexibility. However, it is preferable that the entire flexible printed circuit board is coated and has a coating layer. This is to prevent the light source or individual elements from being left behind in the brain or damaging the brain surface. The coating layer is preferably transparent or semi-transparent in order to transmit light.

[0020] A preferred example of the brain activity measurement device 1 is that at least one of the plurality of sheet-like probe sections 9 (preferably all of the sheet-like probe sections 9) has a first detection section 17. The sheet-like probe section 9 may have a first light source 15, a first detection section 17, and a first light control section 19 for controlling the light output from the first light source 15. The first detection section 17 is used to detect signals related to brain activity. The first light source 15 may be used to excite light-emitting substances contained in the brain. In this example, for example, light-emitting substances have been introduced into the brain of the target organism. It is preferable that the nerve cells of the target organism emit light in response to a predetermined light stimulus by gene transfer or the introduction of light-emitting substances. An example of a light-emitting substance is a fluorescent protein. Examples of fluorescent proteins are green fluorescent protein (e.g., GFP), red fluorescent protein (e.g., DsRed), and calcium-sensitive fluorescent protein (GECI). A calcium-sensitive fluorescent protein is a fluorescent protein that is sensitive to calcium. For example, the nerve cells of the target organism are introduced with a luminescent substance (GECI). Examples of GECIs include GCaMP proteins (GCaMP3 protein, GCaMP6s protein, GCaMP7 protein, RCaMP1) and aequorin. However, GECIs discovered in the future may also be used. Methods for introducing GECIs are publicly known, for example, as described in Japanese Patent Publication No. 5854686. When GECIs are introduced, the brain is activated, and the GECIs emit light in response to changes in calcium levels. For example, the GECIs contained in the nerve cells of the brain emit light. In this way, the method of measuring brain function using luminescent substances can measure brain function more accurately and quickly than methods such as functional near-infrared spectroscopy and optical topography.

[0021] Each sheet-like probe section 9 may have a plurality of first light sources 15. The first optical control unit 19 controls the light intensity and ON / OFF status of the first light sources 15. The first optical control unit 19 may receive control commands from the control unit 20 and control the first light sources 15 according to the control commands. The control unit 20 may be implemented, for example, by a computer or processor. The signal analysis unit 7 may output information for controlling the first light sources 15 based on the measurement information it obtains to the first optical control unit 15. The expressions "control unit 19" and "control unit 20" are functional, and they may be physically the same element. The first optical control unit 19 may also store a program and control the first light sources 15 based on the commands of that program. The first optical control unit 19 may control a plurality of first light sources 15. The plurality of first light sources 15 may be installed at equal intervals on the sheet-like probe section 9, for example. Examples of the first light sources 15 are light-emitting diodes and LEDs. The output of the first light sources 15 is preferably at an intensity that does not damage the brain. The wavelength of the light output by the first light source 15 is preferably the wavelength corresponding to the light-emitting material. By using the first light source 15, the fluorescent material can be excited. An example of the first light source 15 is visible light, and it is preferably a light source that emits light with a wavelength of 400 nm to 600 nm. The light from the first light source 15 may be pulsed light or continuous light, for example.

[0022] An example of the first detection unit 17 is a light sensor. The first detection unit 17 may measure physical quantities other than light that are related to brain activity. The first detection unit 17 may detect the emission of light from a light-emitting material. The first detection unit 17 may detect reflected light after the first light source 15 has irradiated it, or it may detect light emitted by the light-emitting material in reaction to the first light source 15 irradiating it. An example of the first detection unit 17 is a photodiode (PD). Preferably, each sheet-like probe unit 9 has a plurality of first detection units 17.

[0023] Figure 4 is a conceptual diagram showing how the first detection unit detects fluorescence emitted by a fluorescent substance. In this example, for instance, light emitted from the first light source 15 excites GECI introduced into nerve cells. The first detection unit 17 then detects the light (arrow) emitted by the excited GECI.

[0024] The optical signal detected by the first detection unit 17 is converted into a detection signal and output to the detection signal receiving unit 5. The process of converting the optical signal into a detection signal and the process of outputting the detection signal to the detection signal receiving unit 5 can be performed by the first detection unit 17, for example. A normal photodiode can perform these processes. The detection signal receiving unit 5 is connected to receive signals from the brain diagnostic probe 3. The detection signal receiving unit 5 may be installed inside the brain together with the brain diagnostic probe 3. Alternatively, the detection signal receiving unit 5 may be located outside the brain. For example, the brain activity measurement device 1 may further have a wireless output unit 25 for wirelessly outputting the detection signal to the outside of the target organism. The wireless output unit 25 may wirelessly output the detection signal, and the detection signal receiving unit 5 may receive the detection signal.

[0025] The detection signal receiving unit 5 is an element for receiving detection signals from the brain diagnostic probe 3. If the detection signal receiving unit 5 and the brain diagnostic probe 3 are connected by a wire, the detection signal receiving unit 5 should receive the detection signal output from the brain diagnostic probe 3 via the wire, such as an electrical or optical signal. If the detection signal is output as a wireless signal, the detection signal receiving unit 5 should have an element for receiving wireless signals, such as an antenna, and should receive the detection signal, which is a wireless signal.

[0026] The signal analysis unit 7 is an element that analyzes the detection signal received by the detection signal receiving unit 5 and obtains measurement information for measuring brain activity. The signal analysis unit 7 can be implemented using a computer or processor.

[0027] A computer has an input unit, an output unit, a control unit, an arithmetic unit, and a memory unit, and each element is connected by a bus or the like to enable the exchange of information. For example, the memory unit may store a program or various kinds of information. When predetermined information is input from the input unit, the control unit reads the program stored in the memory unit. The control unit then reads the information stored in the memory unit as appropriate and transmits it to the arithmetic unit. The control unit also transmits the input information to the arithmetic unit as appropriate. The arithmetic unit performs calculations using the received information and stores it in the memory unit. The control unit reads the calculation results stored in the memory unit and outputs them from the output unit. In this way, various processes and steps are executed. Each unit and each means is responsible for executing these various processes. A computer may have a processor, and the processor may implement various functions and steps. A computer may be standalone. A computer may have some of its functions distributed between a server and terminals. In that case, it is preferable that the server and terminals can exchange information via a network such as the internet or an intranet. A computer may include a processor and memory connected to the processor. The memory may store instructions, and when executed by the processor, these instructions may cause the computer to perform various processes or to function as various components. The computer may build a learning model by providing various training data and perform various calculations through machine learning. In this case, the computer may perform various analyses and interpretations using the learning model created by AI (artificial intelligence) machine learning and deep learning. Doing so will improve the accuracy of machine learning.

[0028] For example, the signal analysis unit 7 has a learning model built on past detected signals and brain activity. The signal analysis unit 7 can then use this learning model to analyze the detected signals and obtain real-time measurement information on brain activity. Alternatively, the signal analysis unit 7 may store past measurement information of the target organism and obtain real-time measurement information by comparing it with past measurement information. Examples of measurement information include the degree and progression of brain diseases. Brain diseases refer to diseases caused by the death of brain nerve cells that are most important for information transmission in the brain nervous system, problems with the formation and function of synapses that transmit information between brain nerve cells, and abnormal symptoms or reductions in the electrical activity of brain nerves. An example of a brain disease is a degenerative brain disease. A degenerative brain disease is an aging-related disease defined as the gradual loss of specific populations of nerve cells and associated with protein aggregates. Examples of degenerative brain diseases include stroke, paralysis, dementia, Alzheimer's disease, Parkinson's disease, Huntington's disease, multiple sclerosis, and amyotrophic lateral sclerosis. A learning model can be constructed using the degree and progression of each of these diseases, along with past detection signals related to each, as training data. Furthermore, by using the detection signals obtained in this study and information about brain activity, the accuracy of the learning model for the target organism can be improved. For example, examples of the progression of Alzheimer's disease include mild cognitive impairment (MCI), early stage, middle stage, and late stage.

[0029] The measurement information obtained by the signal analysis unit 7 is output as appropriate. Examples of output include display on a monitor, printing on paper, or outputting as electronic information to a doctor's or the target organism's terminal. A preferred example of the brain activity measurement device 1 is one which further includes a wireless output unit 25 for wirelessly outputting the measurement information to the outside of the target organism. In particular, if the signal analysis unit 7 is expected to be located inside the brain of the target organism or is physically connected to the target organism, having such a wireless output unit 25 can make life easier for the target organism.

[0030] A preferred example of the brain activity measurement device 1 is one in which at least one of the multiple sheet-like probe units 9 has a second light source 21 having a wavelength different from that of the first light source 15. The second light source 21 is used to treat the brain of a target organism. This example may include, for example, a second optical control unit for controlling the output of the second light source 21. The second optical control unit may be installed in the sheet-like probe unit 9. The second optical control unit may have a computer or processor and control the output of the second light source 21. Alternatively, the second optical control unit may receive control commands from the control unit 20 and control the second light source 21 according to the control commands. The control unit 20 may, for example, output information to the second optical control unit for controlling the second light source 21 based on measurement information obtained by the signal analysis unit 7. The expressions "second control unit" and "control unit 20" are functional, and they may be physically the same element. This brain activity measurement device 1 also functions as a device for treating brain diseases with light. Furthermore, if the first light source 15 is not present, the sheet-shaped probe 9 will have a second light source 21.

[0031] In this example, the second optical control unit or control unit 20 stores the detection signal or measurement information relating to brain activity and information relating to the output of the second light source 21 (which may also include information relating to which part of the second light source 21 to turn ON). The second optical control unit or control unit 20 then receives the detection signal or measurement information relating to brain activity, reads the information relating to the output of the second light source 21 from the storage unit, and controls the output of the second light source 21. Alternatively, the second optical control unit or control unit 20 may construct a learning model relating the detection signal or measurement information and the output of the second light source 21, and use this learning model to obtain information for controlling the output of the second light source 21 from the detection signal or measurement information. In this way, the brain can be appropriately treated based on the measurement information of the brain.

[0032] The brain activity measurement device 1 having a second light source 21 can be preferably used in cases where optogenetics are introduced into the target organism. Optogenetics are genes that can activate nerve cells based on light of a specific wavelength. Examples of photoactive proteins generated by the introduction of these optogenetics include channelrhodopsin 2, mutant ChR2, ChR2 / H134R, ChR2 / C128X (where X is T, A, or S) or ChR2 / D156A, ChR2 / E123T (ChETA), halorhodopsin, archaerodopsin 3, archaerodopsin T, OptoXRs, photoactivated adenylate cyclase, and melanopsin. Methods for introducing optogenetics that express these proteins are well known.

[0033] Figure 5 is a conceptual diagram showing a design example of a circuit on a sheet-like probe having a first and a second light source. In the example shown in Figure 5, a circuit board is formed on the sheet-like probe. The circuit board is equipped with a first light source 15, a first detection unit 17, and a second light source 21, which are connected by wiring.

[0034] A preferred example of the brain activity measurement device 1 is one in which the signal analysis unit 7 uses machine learning to analyze the brain activity of the target organism and obtain information about the organism's intentions. For example, this example can be preferably used when the target organism is in a state where communication is difficult, such as being unable to speak. In this example, a learning model is constructed using detection signals corresponding to the state of each intention (e.g., grateful, happy, OK, NG, unpleasant, want it to stop). The signal analysis unit 7 can then use such a learning model to obtain information about the organism's intentions based on the obtained detection signals. Furthermore, it is preferable that the brain activity measurement device 1 further includes an intention information output unit 23 that outputs intention information. Examples of the intention information output unit 23 include a monitor attached to the target organism and a monitor that exists separately from the target organism. If the intention information output unit 23 is a monitor attached to the target organism, for example, the monitor may display icons corresponding to the state of each intention (e.g., a smiling icon, an angry icon). In this way, the intentions of target organisms with whom communication is difficult (for example, those with advanced dementia or paralysis, or non-human mammals) can be expressed in a way that is easily understood.

[0035] In addition to icons, measurement data such as the release levels of various hormones including adrenaline and serotonin, changes in blood flow, changes in blood pressure, changes in body temperature, changes in brain state, and changes in blood glucose levels may also be displayed on the monitor attached to the target organism, as well as on a separate monitor. The display method on the monitor may be switched depending on who the observer is (family members providing care, medical professionals, the target organism itself, etc.) and the purpose of observation (use for rapid action, diagnosis including comparison with history, adjustment of stimulus content, etc.), and may also be customizable.

[0036] A preferred example of the brain activity measurement device 1 is one in which the signal analysis unit 7 uses machine learning to analyze the activity of nerve cells in the target organism. For example, when using the GECI described earlier, the activity of nerve cells in the target organism can be analyzed.

[0037] A preferred example of the brain activity measurement device 1 further comprises a deep brain stimulation electrode 31, a power supply unit 33 for supplying power to the deep brain stimulation electrode 31, and a power supply control unit 35 for controlling the power supplied by the power supply unit 33. Deep brain stimulation electrodes are known, for example, as described in Japanese Patent Publication No. 6603656. Furthermore, deep brain stimulation systems using deep brain stimulation electrodes are known, as described in Japanese Patent Publication No. 2008-513082. A preferred example of the brain activity measurement device 1 can be realized by appropriately applying these known technologies. For example, the power supply control unit 35 is configured to receive information from the control unit 20. The power supply control unit 35 may receive control commands from the control unit 20 and control the power supply unit 33 according to the control commands. Since the control unit 20 has received measurement information, it only needs to output control commands to the power supply unit 33 according to the measurement information. In this way, the deep brain stimulation electrode 31 can be driven in response to the brain's activity status. The terms "power supply control unit 35" and "control unit 20" are functional, and they may be physically the same element.

[0038] The brain activity measurement device 1 can be used, for example, by surgically creating a hole 13 in the skull of the target organism and laying multiple sheet-like probe parts 9 on the surface of the brain parenchyma or the like through the hole 13.

[0039] Figure 6 is a block diagram illustrating a nerve stimulator. As shown in Figure 6, the nerve stimulator 41 includes a diagnostic probe 43, a detection signal receiving unit 45, a signal analysis unit 47, a stimulation information acquisition unit 49, and a nerve stimulator 51. The nerve stimulator 41 is a device that analyzes nerve-derived information and stimulates nerves based on the analyzed nerve signals to provide a predetermined effect to a subject. The nerve stimulator 41 can be used, for example, for the treatment of brain diseases. Examples of brain diseases include any or more of the following: visual impairments, speech disorders, motor disorders, strokes, paralysis, dementia, Alzheimer's disease, Parkinson's disease, Huntington's disease, multiple sclerosis, and amyotrophic lateral sclerosis. Furthermore, the nerve stimulator 41 can be used to give a subject a simulated sensation or to induce movement based on nerve stimulation.

[0040] The diagnostic probe 43 is an element for detecting signals from nerves. These signals can be any signals originating from nerves. For example, the diagnostic probe 43 may detect weak electrical currents originating from nerves, or it may detect signals observed by sensors or the like after applying some kind of stimulus (e.g., temperature, light, electrical, magnetic, and ultrasonic stimulation) to a site containing nerves. An example of the diagnostic probe 43 is the brain diagnostic probe 3 described earlier. In this case, the nerves from which the diagnostic probe 43 detects signals are cranial nerves.

[0041] The detection signal receiving unit 45 is an element for receiving detection signals from the diagnostic probe 43. The detection signal receiving unit 45 can adopt the same configuration as the detection signal receiving unit 5 described earlier.

[0042] The signal analysis unit 47 is an element for analyzing the detection signal received by the detection signal receiving unit 45 and obtaining nerve signals. The detection signal may contain noise, or by processing the detection signal, it may be possible to obtain signals originating from nerves. The presence of the progression analysis unit 47 makes it possible to obtain nerve signals, which are signals originating from nerves, based on the detection signal. The signal analysis unit 47 can adopt the same configuration as the signal analysis unit 7 described earlier. Nerve signals refer to nerve-derived signals obtained as a result of analyzing the detection signal.

[0043] The stimulus information acquisition unit 49 is an element for obtaining stimulus information for stimulating nerves based on nerve signals obtained by the signal analysis unit 47. Preferably, the stimulus information acquisition unit 49 has a stimulus model 53, which is a trained model for obtaining stimulus information. The stimulus model 53 is a trained model that outputs stimulus information when a nerve signal is input. An example of training data for the stimulus model 53 includes nerve signals and stimulus information. Preferably, the feedback information to the stimulus model 53 includes the results of the stimulus information.

[0044] FIG. 7 is a conceptual diagram for explaining a stimulation model. In the example of FIG. 7, the stimulation model 53 is a learned model constructed by machine learning using information including nerve information and stimulation information as teacher data. The teacher data may include information regarding the actions of the subject in addition to nerve information and stimulation information. Examples of the actions of the subject are pseudo sensations, movements, and physiological changes. Examples of physiological changes are the release amounts of various hormones such as adrenaline and serotonin, changes in blood flow, changes in blood pressure, changes in body temperature, changes in the state of the brain, and changes in blood glucose levels. For example, when the brain is irradiated with light of a certain wavelength and intensity, and a diagnostic probe observes a predetermined signal, and in that case, if a certain memory is implanted in the subject, the wavelength and intensity of the light are stimulation information, the predetermined signal detected by the diagnostic probe is nerve information, and a certain memory is information regarding the action of the subject. By inputting a large number of these data as teacher data into the learned model, the accuracy of the stimulation model 53 is improved. Also, actually, by further inputting the result of applying a stimulation based on the stimulation information to the subject as teacher data into the learned model, the accuracy of the learned model can be further improved. Since the stimulation model 53 uses information based on the actual actions of the subject as one of the teacher data, there is a correlation between the input data and the output data.

[0045] For example, when the action of the subject is pain, by using the above-mentioned learned model, stimulation information for preventing the occurrence of pain can be obtained. In this case, by using the nerve stimulation device 41, there is an advantage that anesthesia may not be required or the anesthesia can be reduced.

[0046] The nerve stimulation unit 51 is an element for stimulating a target nerve based on the stimulation information acquired by the stimulation information acquisition unit 49. The nerve stimulation unit 51 may be an element that can exert some action on the nerve. Examples of the nerve stimulation unit 51 are those having a temperature control unit, those that can irradiate light, those that can irradiate waves such as electromagnetic waves, and those that can apply electric power. Another example of the nerve stimulation unit 51 is the brain diagnostic probe 3 or the power supply unit 33 described above.

[0047] The nerve stimulation device 41 has a detection signal receiving unit 45 that receives a detection signal from the diagnostic probe 43. A signal analysis unit 47 analyzes the detection signal received by the detection signal receiving unit 45 to obtain a nerve signal. In this way, the nerve stimulation device 41 can obtain a signal derived from a nerve. A stimulation information acquisition unit 49 obtains stimulation information for stimulating a nerve based on the nerve signal obtained by the signal analysis unit 47. For example, the stimulation information acquisition unit 49 can obtain stimulation information as output information by inputting the nerve signal into a learned model. A nerve stimulation unit 51 stimulates a target nerve based on the stimulation information acquired by the stimulation information acquisition unit 49. In this way, the nerve stimulation device 41 can stimulate a nerve in real time based on the information detected by the diagnostic probe 43.

[0048] FIG. 8 is a conceptual diagram for explaining a pseudo-sensation providing device. As shown in FIG. 8, the pseudo-sensation providing device 61 includes a brain stimulation unit 63 and a stimulation information control unit 65. The pseudo-sensation providing device 61 is a device for giving a pseudo-sensation to a subject. The pseudo-sensation means a sensation brought to the brain by brain stimulation (including memory and emotion in addition to the five senses), rather than an actual sensation.

[0049] The brain stimulation unit 63 is an element for stimulating the brain of a subject. The brain stimulation unit 63 may be any device that can stimulate the brain. Examples of stimulation include temperature stimulation, light stimulation, electrical stimulation, magnetic stimulation, and ultrasonic stimulation. An example of the brain stimulation unit 63 is the brain diagnostic probe 3 or the power supply unit 33 described above.

[0050] The stimulation information control unit 65 has a pseudo-sensation model 67, which is a learned model constructed using, as teacher data, stimulation information for the brain of a target organism and sensation information of the target organism. When sensation information for giving a pseudo-sensation to a subject is input into the pseudo-sensation model 67, output stimulation information, which is stimulation information, is output. The pseudo-sensation model 67 can be constructed in the same manner as the stimulation model 53 described above.

[0051] For example, suppose that when a first predetermined light is shone on a first part of a person's brain and a second predetermined light is shone on a second part, the target organism obtains pseudo-visual information such as an image of a predetermined size (for example, an image of the sun) at a predetermined location. In this case, the first predetermined light on the first part and the second predetermined light on the second part are stimulus information to the brain, and the pseudo-visual information that an image of a predetermined size was seen at a predetermined location is the sensory information of the target organism. The pseudo-sensory providing device 61 constructs a trained model using such stimulus information and pseudo-sensory information as training data. Then, if it is desired to provide a pseudo-sensory experience of seeing an image of a predetermined size at a predetermined location, the information regarding the pseudo-sensory experience to be provided is input into the trained model, and output information on what kind of stimulus should be applied to provide that pseudo-sensory experience can be obtained.

[0052] Examples of senses include one or more of sight, hearing, touch, taste, smell, memory, consciousness, intention, and understanding. Preferred examples of senses are the five senses. For example, if the brain remembers certain information (e.g., historical dates, English vocabulary) when given a certain stimulus, then by building a trained model using that information as training data, the brain can be made to remember the desired information when given that stimulus. Consciousness means being aware of or recognizing things or states. For example, if the brain comes to like a certain food (becomes able to eat a certain food) when given a certain stimulus, then a trained model can be built using that stimulus and the consciousness as training data. For example, if you want a certain species to eat a certain food, you can input information about the food you want to feed it into the trained model and get an output on how to stimulate it. Intention means the will to do something. For example, suppose the brain comes to the intention to study when given a certain stimulus. In this case, a trained model can be built using that stimulus and the intention to study as training data. When information about the intention to study is input into a trained model, it can obtain stimulus information that (simulated, rather than being based on one's own intention) brings about the intention to study. Understanding means grasping the principles of something.

[0053] For example, regarding vision, by using records of brain neural activity when viewing a face or landscape (or when imagining a scene or drawing with eyes closed) as training data, and applying stimuli in a similar pattern, it is possible to reproduce similar simulated visuals and imagined scenes, such as "recognition of what is being seen / imagined (is it a landscape, a human face, and if it's a face, whose face it is, etc.)." Furthermore, it is conceivable to integrate multiple sensations experienced by one person (person A) at a certain time and project / simulate them as similar sensations to another person (person B). In this case, since there are individual differences in the patterns of neural activity between person A and person B, it is possible to bridge the gap between A and B by utilizing the past machine learning results of each person and bridging the neural stimulation patterns that evoke similar sensations, thereby enabling the simulated experiences described above.

[0054] Another possible use case is projecting the user's sensations or imagined images onto a digital device. In particular, for visual projections, the user could reconstruct imagined scenes or diagrams on the screen through machine learning processing, and then rapidly refine the scenes or diagrams by feeding back the brain's response to them, until they are closer to the user's intended result. This would allow the user to quickly "thought-project" their mental images or diagrams onto a digital device.

[0055] Figure 9 shows an example of how a simulated vision, a simulated sense, is formed based on an image imagined by the user. Figure 9(a) shows the outline of a cat, Figure 9(b) shows a cat, and Figure 9(c) shows a landscape including a cat. For example, as shown in Figure 9, the simulated vision may be formed stepwise (just as a picture gradually approaches completion) by providing appropriate stimuli. More specifically, it is conceivable that an AI that generates documents or images / videos could use the detection function of this device to run a high-speed (real-time) feedback loop on the generated output. For example, in the case of an image generated as shown in Figure 9, the user provides feedback (whether the impression is good or bad, whether it has been improved, what would you like added, etc.) to the AI ​​at high speed (real-time), the AI ​​then presents an improved output at high speed (real-time), and the feedback loop is further rotated at high speed (real-time), thus enabling a method to improve the output more rapidly than before.

[0056] The stimulus information control unit 65 is an element for obtaining stimulus unit control information using output stimulus information. Specifically, the stimulus unit control information is information for controlling the brain stimulation unit 63 to stimulate the brain. The brain stimulation unit 63 stimulates the brain based on the stimulus unit control information.

[0057] A brain stimulation unit 63 is installed on the subject to stimulate their brain. Sensory information (for example, information to simulate the visual perception of a predetermined image) is input to the simulated sensation providing device 61 to give the subject a simulated sensation. The simulated sensation providing device 61 inputs the sensory information as input data to the simulated sensation model 67. The simulated sensation model 67 outputs output stimulation information as output data corresponding to the sensory information. The stimulation information control unit 65 obtains stimulation unit control information using the output stimulation information. The brain stimulation unit 63 stimulates the subject's brain based on the stimulation unit control information. In this way, the simulated sensation providing device 61 can give the subject a simulated sensation.

[0058] Figure 10 is a conceptual diagram illustrating the motor induction device. As shown in Figure 10, the motor induction device 71 includes a brain stimulation unit 73 and a stimulation information control unit 75. The brain stimulation unit 73 is the same as the brain stimulation unit 63 described earlier.

[0059] The stimulus information control unit 75 has a brain stimulus-induced motor model 77, which is a trained model constructed using stimulus information to the brain of the target organism and motor information of the target organism as training data. When motor information for causing the subject to perform a movement is input to the brain stimulus-induced motor model 77, the brain stimulus-induced motor model 77 (and the stimulus information control unit 75 having it) outputs output stimulus information. For example, suppose that when a stimulus is applied to the brain, the target organism makes a certain movement (for example, raises the right hand, inhales, moves the right foot forward, or turns over in bed). In that case, that stimulus is used as brain stimulus information, and that certain movement is used as motor information. Multiple trained models 77 are input with this kind of information as training data. For example, if it is desired to force a bedridden subject to turn over in bed, information regarding "turning over in bed" is input to the brain stimulus-induced motor model 77 as motor information. The brain stimulus-induced motor model 77 then outputs output stimulus information for causing the subject to turn over in bed. In this way, the subject can be made to perform the desired exercise. This exercise inducer 71 can, for example, sustain the life of a subject who is unable to perform a predetermined exercise of their own conscious will, or induce a predetermined exercise unconsciously. Furthermore, since this exercise inducer 71 can, for example, force a subject to exercise, it can be effectively used in rehabilitation and other applications.

[0060] The stimulus information control unit 75 obtains stimulus unit control information using the output stimulus information. The brain stimulation unit 73 stimulates the subject's brain based on the stimulus unit control information.

[0061] This invention can be used in the field of medical devices related to the brain, and in the field of communication tools for obtaining the intentions of target organisms.

[0062] 1 Brain activity measurement device 3 Brain diagnostic probe 5 Detection signal receiving unit 7 Signal analysis unit 11 Skull of target organism 13 Hole 15 First light source 17 First detection unit 19 First optical control unit 20 Control unit 21 Second light source 23 Will information output unit 25 Wireless output unit 31 Deep brain stimulating electrode 33 Power supply unit 35 Power supply control unit 41 Nerve stimulator 43 Diagnostic probe 45 Detection signal receiving unit 47 Signal analysis unit 49 Stimulation information acquisition unit 51 Nerve stimulating unit 53 Stimulation model

Claims

1. A simulated sensation providing device for giving a subject a simulated sensation, comprising: a brain stimulation unit for stimulating the subject's brain; and a stimulation information control unit for obtaining stimulation unit control information for controlling the brain stimulation unit, wherein the stimulation information control unit has a simulated sensation model which is a trained model constructed using stimulation information to the brain of a target organism and sensory information of the target organism as training data; when sensory information for giving a simulated sensation to the subject is input to the simulated sensation model, it outputs output stimulation information; and the stimulation information control unit obtains stimulation unit control information using the output stimulation information.

2. A simulated sensory device according to claim 1, wherein the sensory device includes one or more of the following senses: sight, hearing, touch, taste, smell, memory, consciousness, intention, and understanding.

3. A simulated sensory device according to claim 1, wherein the sensory device includes one or more of the following senses: sight, hearing, touch, taste, and smell.

4. A pseudo-sensory device according to claim 1, wherein the sensation includes vision.

5. A motor induction device comprising: a brain stimulation unit for stimulating the brain of a subject; and a stimulation information control unit for obtaining stimulation unit control information for controlling the brain stimulation unit, wherein the stimulation information control unit has a brain stimulation-induced motor model which is a trained model constructed using stimulation information to the brain of a subject organism and motor information of the subject organism as training data; when motor information for causing the subject to perform a movement is input to the brain stimulation-induced motor model, it outputs output stimulation information; and the stimulation information control unit obtains stimulation unit control information using the output stimulation information.

Images