Simulation apparatus and simulation method

Abstract

To provide a simulation device that can perform simulation including muscle synergy with respect to a motion of organs of the craniocervical part at the time of deglutition or manducation, and a simulation method.SOLUTION: A deglutition simulation device 1 can perform simulation of a motion of organs of the craniocervical part at the time of deglutition on the basis of the particle method by using a dynamic three-dimensional craniocervical model in which a muscle activity rate is set to a muscle particle for each muscle type, and analyzes muscle synergy with respect to the motion of the organs of the craniocervical part at the time of deglutition on the basis of an analysis result of a muscular activity rate at the time of the simulation.SELECTED DRAWING: Figure 1

Description

Technical Field

[0001] The present invention relates to a simulation apparatus and a simulation method.

Background Art

[0002] The relationship between food physical properties and the movement of head and neck organs during swallowing or mastication is complex, and it is extremely difficult to accurately grasp the movement of the head and neck organs. Here, swallowing refers to the movement of conveying food (including beverages) taken into the oral cavity to the stomach through the pharynx and esophagus. During swallowing, the respective muscles of the oral cavity, pharynx, larynx, and esophagus activate in a predetermined sequence within a short period of time, and perform complex movements. In addition, mastication is the movement of grinding food that has been taken into the oral cavity while moving the mandible, the genioglossus muscle and the like. Even during mastication, the muscles of the oral cavity, pharynx, and larynx, as well as the genioglossus muscle and other muscles activate in a predetermined sequence to perform complex movements.

[0003] By the way, in recent years, as for such muscle activity analysis, it has been considered to attach electrodes and electromyographs to a plurality of sites such as the biceps femoris and adductor muscles of a subject respectively, measure the myoelectric potential during maximum voluntary contraction (MVC), and analyze muscle synergy during walking based on the muscle synergy hypothesis that multiple muscle types are controlled by coordinated movement and the measurement results (see, for example, Non-Patent Document 1).

[0004] In addition, as an analysis of muscle activity during eating and swallowing, an eating and swallowing function evaluation method has been proposed, which is characterized by detecting biological signals using a myoelectric sensor for suprahyoid muscle groups or a myoelectric sensor for infrahyoid muscle groups (array electrodes in which multi-channel electrodes are arranged), and using feature quantities converted based on the muscle synergy hypothesis (see, for example, Patent Document 1).

[0005] On the other hand, as an analysis of muscle activity during eating and swallowing, a swallowing simulation apparatus that sets a muscle activity rate value for each muscle type has been proposed (see, for example, Patent Document 2).

Prior Art Literature

[0006] [Patent Document 1] Japanese Patent Publication No. 2021-142087 [Patent Document 2] Japanese Patent Publication No. 2021-112557 [Non-patent literature]

[0007] [Non-Patent Document 1] Naoto Matsunaga, Koji Kanaoka, "Waseda University Doctoral Dissertation, PhD (Sports Science): Muscle Activity Analysis of Sports Movements Using Synergy Analysis" [Retrieved November 18, 2022], Internet <URL:https: / / waseda.repo.nii.ac.jp / ?action=repository_action_common_download&item_id=41273&item_no=1&attribute_id=20&file_no=1> [Overview of the Initiative] [Problems that the invention aims to solve]

[0008] However, Non-Patent Literature 1 only allows for measuring electromyography (EMG) by directly attaching electromyographs to the subject's skin. Therefore, it is not possible to measure EMG in the head and neck organs, such as the oral cavity, pharynx, and larynx, which are located inside the human body. This presents a problem in that it is difficult to analyze muscle synergies related to the movement of head and neck organs during swallowing and chewing.

[0009] Patent Document 1 states that even when using an electromyogram sensor, it is not possible to acquire the electromyographic potentials of the tongue, soft palate, and pharynx. Furthermore, due to the movement of the hyoid bone and larynx, and crosstalk issues, it is considered that the specific electromyographic potentials of the suprahyoid and infrahyoid muscle groups cannot be precisely identified, making it difficult to analyze muscle synergies related to the movement of head and neck organs during swallowing and chewing.

[0010] Patent Document 2 sets muscle activity rate values ​​for each type of muscle, and does not consider muscle synergy at all.

[0011] The present invention has been made in consideration of the above points, and aims to provide a simulation device and simulation method that can perform simulations, including muscle synergies, regarding the movement of head and neck organs during swallowing or chewing. [Means for solving the problem]

[0012] The simulation device according to the present invention includes: a storage unit that stores a dynamic three-dimensional head and neck model in which a plurality of head and neck organs are modeled by a plurality of particulate or mesh-like myoforms, each of which identifies the direction of muscle fibers in a three-dimensional image for each muscle type, and in which contractile stress based on the direction of muscle fibers during swallowing or mastication is applied to the myoforms by the muscle activity rate; a motion analysis unit that uses the dynamic three-dimensional head and neck model in which the muscle activity rate can be changed to simulate the movement of the head and neck organs during swallowing or mastication in the three-dimensional image based on the particle method or the grid method; and a muscle synergy analysis processing unit that analyzes the number of muscle synergies in which the muscle types move in coordination, a spatial pattern showing the degree of muscle activity of the constituent muscle types for each muscle synergy during swallowing or mastication, and a time pattern showing the degree of activity for each muscle synergy as it changes over time, from the muscle activity rate of each muscle type that changes in accordance with the change in the movement of the head and neck organs over time.

[0013] Furthermore, the simulation method according to the present invention includes a storage step of storing in a memory unit a dynamic three-dimensional head and neck model in which multiple head and neck organs are modeled by multiple particulate or mesh-like myoforms, each of which identifies the direction of muscle fibers in a three-dimensional image for each muscle type, and in which contractile stress based on the direction of muscle fibers during swallowing or chewing is applied to the myoforms by the muscle activity rate, and a motion analysis unit uses the dynamic three-dimensional head and neck model in which the muscle activity rate can be changed to analyze the head and neck organs during swallowing or chewing. The method includes: a motion analysis step of simulating motion in the three-dimensional image based on a particle method or a grid method; and a muscle synergy analysis processing step of analyzing, using a muscle synergy analysis processing unit, the number of muscle synergies in which the muscle types perform coordinated movements, a spatial pattern showing the degree of muscle activity of the constituent muscle types for each muscle synergy during swallowing or chewing, and a temporal pattern showing the degree of activity for each muscle synergy as it changes over time, based on the muscle activity rate of each muscle type that changes in accordance with the time course of motion of the head and neck organs. [Effects of the Invention]

[0014] According to the present invention, a dynamic three-dimensional head and neck model is available in which the muscle activity rate set for each myosofactory is adjustable, allowing for simulations of the movement of head and neck organs during swallowing or chewing, including muscle synergies. [Brief explanation of the drawing]

[0015] [Figure 1] This is a block diagram showing the circuit configuration of a swallowing simulation device. [Figure 2] This is a schematic diagram showing the configuration of a dynamic three-dimensional head and neck model. [Figure 3] This is a schematic diagram illustrating the particles that make up the head and neck organs in a dynamic three-dimensional head and neck model. [Figure 4] Figure 3 is a cross-sectional view showing the cross-sectional configuration in the midline of the dynamic three-dimensional head and neck model. [Figure 5] Figure 5A is a schematic diagram showing the structure around the genioglossus muscle in a dynamic three-dimensional head and neck model, and Figure 5B is a schematic diagram showing the structure of the transverse tongue muscle in a dynamic three-dimensional head and neck model. [Figure 6] Schematic diagram showing the configuration of the geniohyoid muscle, hyoid bone and mandible in a dynamic three-dimensional head and neck model. [Figure 7] Schematic diagram showing the configuration around the palatopharyngeus muscle in a dynamic three-dimensional head and neck model. [Figure 8] 8A is a schematic diagram showing the configuration around the hyoid bone, thyroid cartilage, cricoid cartilage and inferior pharyngeal constrictor thyropharyngeal part in a dynamic three-dimensional head and neck model; 8B is a schematic diagram showing the configuration of the great cornu pharyngeal part of the middle pharyngeal constrictor in a dynamic three-dimensional head and neck model; and 8C is a schematic diagram showing the configuration of the lesser cornu pharyngeal part of the middle pharyngeal constrictor in a dynamic three-dimensional head and neck model. [Figure 9] Schematic diagram showing the state change of the dynamic three-dimensional head and neck model during swallowing. [Figure 10] Schematic diagram for explaining muscle particles. [Figure 11] Schematic diagram for explaining the muscle fiber direction of muscle particles. [Figure 12A] Graph (1) showing the temporal change of muscle activity rate of each head and neck organ in the dynamic three-dimensional head and neck model during swallowing. [Figure 12B] Graph (2) showing the temporal change of muscle activity rate of each head and neck organ in the dynamic three-dimensional head and neck model during swallowing. [Figure 13] Schematic diagram for explaining non-negative matrix factorization. [Figure 14] Schematic diagram showing the constituent muscle types of the first muscle synergy, the second muscle synergy and the third muscle synergy, and the calculated weight for each muscle type. [Figure 15] Schematic diagram showing the constituent muscle types of the fourth muscle synergy, the fifth muscle synergy and the sixth muscle synergy, and the calculated weight for each muscle type. [Figure 16] Schematic diagram showing the constituent muscle types of the seventh muscle synergy, the eighth muscle synergy and the ninth muscle synergy, and the calculated weight for each muscle type. [Figure 17] Graph showing the temporal pattern, which is the temporal change of the activity level of muscle synergies. [Figure 18]Figure 18A is a schematic diagram illustrating the muscle activity matrix X and temporal pattern matrix H divided into multiple time domains, and Figure 18B is a schematic diagram illustrating the configuration of the spatial pattern matrix W and temporal pattern matrix H of muscle synergy. [Figure 19] This graph shows the relationship between the movement of the hyoid bone and its displacement in a dynamic three-dimensional head and neck model, as well as the time domains of the segmented muscle activity matrices X1 to X6 obtained by dividing the muscle activity matrix X. [Figure 20] 20A is a graph showing the time course of muscle activity for the muscles constituting the tongue, such as the genioglossus muscle, during swallowing in a dynamic three-dimensional head and neck model; 20B is a graph showing the time course of muscle activity for the muscles constituting the soft palate, such as the levator veli palatini muscle, during swallowing in a dynamic three-dimensional head and neck model; 20C is a graph showing the time course of muscle activity for the muscles constituting the suprahyoid muscles, such as the anterior belly of the digastric muscle, during swallowing in a dynamic three-dimensional head and neck model; and 20D is a graph showing the time course of muscle activity for the muscles constituting the pharynx, such as the longitudinal portion of the palatopharyngeal muscle, during swallowing in a dynamic three-dimensional head and neck model. [Figure 21] Figure 21A is a schematic diagram illustrating the relationship between the time-divided muscle activity matrix X1 obtained by dividing the muscle activity matrix X into time segments, the spatial pattern matrix W1, and the time-divided time pattern matrix H1. Figure 21B is a schematic diagram illustrating the relationship between the time-divided muscle activity matrix X2 obtained by dividing the muscle activity matrix X into time segments, the spatial pattern matrix W2, and the time-divided time pattern matrix H2. Figure 21C is a schematic diagram illustrating the relationship between the time-divided muscle activity matrix X4 obtained by dividing the muscle activity matrix X into time segments, the spatial pattern matrices W4~7, and the time-divided time pattern matrix H4. [Figure 22] This table explains the interpolation process used to interpolate muscle activity rates. [Modes for carrying out the invention]

[0016] An embodiment of the present invention will be described in detail below with reference to the drawings. In the following description, the same reference numerals are used for identical components, and redundant descriptions are omitted.

[0017] (1) <Outline of the present invention> Figure 1 is a block diagram showing the circuit configuration of the swallowing simulation device 1 according to this embodiment. The swallowing simulation device 1 comprises a personal computer (also referred to as a PC) 2, an input unit 81, a display unit 4, and a storage unit 83. The input unit 81 is an input device such as a mouse or keyboard, which inputs operation commands from the developer to the personal computer 2, causing the personal computer 2 to execute various calculation processes in accordance with the operation commands. The storage unit 83 stores a dynamic three-dimensional head and neck model of particle display formed by the personal computer 2 (described later in Figure 2, etc.), various setting conditions, analysis results, muscle activity rates (described later), etc.

[0018] Personal computer 2, for example, forms a dynamic three-dimensional head and neck model consisting of head and neck organs (described later in Figure 2, etc.) using three-dimensional images, and models the orally ingested product as a simulated orally ingested product within the three-dimensional image. Personal computer 2 analyzes the movement of each head and neck organ in the dynamic three-dimensional head and neck model and the behavior of the simulated orally ingested product during swallowing within the three-dimensional image using a particle method.

[0019] In this embodiment, the behavior of the head and neck organs during swallowing of a simulated orally ingested product is reproduced using a dynamic three-dimensional head and neck model to analyze muscle synergies during swallowing. However, the present invention is not limited to this, and the behavior of the head and neck organs during chewing of a simulated orally ingested product may also be reproduced using a dynamic three-dimensional head and neck model to analyze muscle synergies during chewing.

[0020] Furthermore, personal computer 2 can analyze only the movement of each head and neck organ during swallowing in a dynamic three-dimensional head and neck model within the three-dimensional image using the particle method, without placing a simulated oral intake product within the three-dimensional image.

[0021] Such dynamic three-dimensional head and neck models are created by mimicking, for example, the head and neck of a person with dysphagia who is prone to aspiration, or the head and neck of a healthy person who is not prone to aspiration. For example, a dynamic three-dimensional head and neck model that models the head and neck of a person with dysphagia who is prone to aspiration can analyze the movement of each head and neck organ during swallowing in a person with dysphagia, as well as the behavior of a simulated oral ingestion product during swallowing, and can perform swallowing simulations including muscle synergies during swallowing for people with dysphagia. On the other hand, a dynamic three-dimensional head and neck model that models the head and neck of a healthy person who is not prone to aspiration can analyze the movement of each head and neck organ during swallowing in a healthy person, as well as the behavior of a simulated oral ingestion product during swallowing, and can perform swallowing simulations including muscle synergies during swallowing for healthy people.

[0022] The analysis results obtained by personal computer 2 are output to display unit 4 and displayed on the display screen of display unit 4. Display unit 4 is, for example, a display, and displays on its screen the three-dimensional image of the dynamic three-dimensional head and neck model output from personal computer 2, simulated oral intake products, analysis results, muscle activity rates for each muscle type, and the spatial and temporal patterns of multiple muscle synergies identified based on the muscle synergy hypothesis that multiple muscle types are controlled by coordinated movement. Spatial and temporal patterns will be described later.

[0023] As a result, the display unit 4 allows developers to visually observe the movement of each head and neck organ in the dynamic three-dimensional head and neck model, the behavior of simulated oral ingestion products during swallowing, muscle activity rates for each muscle type, and various analysis results related to muscle synergy. In this embodiment, the three-dimensional image refers to a moving image representing the dynamic three-dimensional head and neck model in a virtual three-dimensional space, and the frame images that constitute the moving image.

[0024] The swallowing simulation device 1 can perform swallowing simulations using a dynamic three-dimensional head and neck model by changing the physical properties of the simulated oral intake food, such as the amount of food bolus, viscosity, and specific gravity. The dynamic three-dimensional head and neck model can be used to check for aspiration and analyze muscle synergies. Furthermore, even if the simulated oral intake food is not set in the three-dimensional image, the swallowing simulation device 1 can perform swallowing simulations using the dynamic three-dimensional head and neck model, and based on the simulation results, it can analyze muscle synergies regarding the movement of each head and neck organ in the dynamic three-dimensional head and neck model during swallowing.

[0025] In this embodiment, as an analytical method that easily represents deformation of the liquid surface and splashes, a particle method is used, which treats the liquid or solid being analyzed as particles. Using this particle method, the movement (action) of head and neck organs in a dynamic three-dimensional head and neck model, and the behavior of orally ingested substances are represented in three-dimensional images to perform swallowing simulation. As a particle method, it is particularly desirable to apply the MPS (Moving Particle Semi-implicit) method (Koshizuka et al, Comput. Fluid Dynamics J, 4, 29-46, 1995). When analyzing the behavior of simulated orally ingested substances during swallowing using swallowing simulation, it is desirable to apply the MPS (Moving Particle Semi-implicit) method, the Hamiltonian particle method (HMPS method), or the Discrete Element Method.

[0026] Furthermore, when analyzing the motion of each particle in a dynamic three-dimensional head and neck model using swallowing simulation, it is desirable to apply the Hamiltonian particle method (Hamiltonian MPS method) as the particle method. In this embodiment, the case in which the Hamiltonian particle method (Hamiltonian MPS method) is applied as the particle method for analyzing the motion of each particle in a dynamic three-dimensional head and neck model using swallowing simulation will be described below.

[0027] In this embodiment, the particle method not only replaces the simulated oral ingestion product with particles, but also replaces the head and neck organs in the dynamic three-dimensional head and neck model with particles, and calculates physical quantities for each particle. As a result, it becomes possible to analyze the head and neck organs in the dynamic three-dimensional head and neck model, as well as the subtle changes that occur during swallowing of the simulated oral ingestion product. In this embodiment, the computational processing performed using such a dynamic three-dimensional head and neck model with the particle method is also referred to as swallowing simulation.

[0028] In this embodiment, the swallowing simulation device 1 not only replaces the head and neck organs in the dynamic three-dimensional head and neck model with particles, but also, based on medical knowledge, the direction of muscle fibers is identified in the three-dimensional image for each type of muscle of the head and neck organs, and contractile stress based on the direction of muscle fibers is applied to the particles that form the head and neck organs such as the oral cavity, pharynx, and larynx.

[0029] In this embodiment, the direction of muscle fibers is identified in a three-dimensional image for each type of muscle of the head and neck organs, and particles to which contractile stress based on the direction of muscle fibers are applied are called muscle particles (particulate myocomites are called muscle particles). The swallowing simulation device 1 reproduces the behavior of the muscle particles set for each type of muscle during swallowing (the temporal change in contractile stress applied in the direction of muscle fibers) by adjusting the value of the muscle activity rate which changes over time.

[0030] Here, muscle activity rate refers to the temporal change in the contractile stress of muscle particles during swallowing, which is set for each type of muscle, with respect to muscle particles that are subjected to contractile stress based on the direction of muscle fibers during swallowing. In the swallowing simulation device 1, the behavior of muscle particles during swallowing can be changed for each type of muscle by setting the value of the muscle activity rate that changes over time during swallowing for each type of muscle. Therefore, by adjusting the value of the muscle activity rate for each type of muscle to the optimal value, the ideal movement of each head and neck organ during swallowing can be realized.

[0031] The swallowing simulation device 1 performs a swallowing simulation (particle method calculation) using a dynamic three-dimensional head and neck model in which predetermined muscle activity rates are set for each muscle type, and obtains analysis results (position and velocity of muscle particles) regarding the behavior of muscle particles during swallowing. Here, the dynamic three-dimensional head and neck model used in the swallowing simulation device 1 according to this embodiment can be, for example, a dynamic three-dimensional head and neck model created according to the method disclosed in Japanese Patent Application Publication No. 2021-029980, in which muscle activity rates are set for each muscle type for the particles forming the dynamic three-dimensional head and neck model. Alternatively, the dynamic three-dimensional head and neck model used in the swallowing simulation device 1 according to this embodiment can be a dynamic three-dimensional head and neck model in which muscle activity rates are set according to the method disclosed in Japanese Patent Application Publication No. 2021-112557.

[0032] In this embodiment, the muscle activity rate set for each muscle type in the particles forming the dynamic three-dimensional head and neck model may be manually set by the developer for each muscle type so that the dynamic three-dimensional head and neck model reproduces the movements during swallowing during swallowing simulation, or the muscle activity rate may be set to the optimal value for each muscle type based on the analysis results of the evaluation function, as disclosed in the above-mentioned Japanese Patent Application Publication No. 2021-112557.

[0033] The swallowing simulation device 1 can achieve ideal movement of each head and neck organ during swallowing by appropriately adjusting the muscle activity rate for each muscle type. Furthermore, by accurately reproducing the movement of the head and neck organs during swallowing and the behavior of simulated oral intake products, the optimal number of muscle synergies and the spatial and temporal patterns of each muscle synergy can be identified from the muscle activity rates for each muscle type.

[0034] In the swallowing simulation device 1, the muscle activity rate of each muscle type is corrected based on the obtained spatial and temporal patterns of muscle synergies, and the dynamic three-dimensional head and neck model is operated again. As a result, if the movement of the dynamic three-dimensional head and neck model is the desired swallowing motion, it can be inferred that the number of muscle synergies is optimal, and the number of combinations of muscle types controlled by coordinated movement during swallowing can be identified from the number of muscle synergies.

[0035] (2) <Circuit configuration of the personal computer in the muscle synergy analysis device> Next, the personal computer 2 of the swallowing simulation device 1 will be described below. As shown in Figure 1, the personal computer 2 includes a head and neck modeling unit 10, an organ movement setting unit 30, an oral intake product physical property setting unit 40, a motion analysis unit 50, a physical property identification unit 70, a control unit 90 and a muscle activity rate changing unit 80, a muscle activity rate analysis unit 92, a muscle synergy analysis processing unit 93 and a simulation determination unit 94. The head and neck modeling unit 10, for example, as shown in Figure 2, forms a dynamic three-dimensional head and neck model 10c that faithfully reproduces the head and neck organs using three-dimensional images.

[0036] Here, the dynamic three-dimensional head and neck model 10c shown in Figure 2 has a configuration created, for example, from a particle-based dynamic three-dimensional head and neck model 10c using the marching cube method or the like, as shown in Figure 3. In such a dynamic three-dimensional head and neck model 10c, the setting information for each particle set in the dynamic three-dimensional head and neck model 10c remains the same, but instead of displaying each individual particle, it simply displays the surface of the head and neck organs for a simplified representation.

[0037] Furthermore, as shown in Figure 2, when displaying the simulated oral ingestion product 100 formed in the three-dimensional image on the display unit 4, the individual particles making up the simulated oral ingestion product 100 are not displayed. Instead, only the surface shape of the simulated oral ingestion product 100, generated by the marching cube method or the like, is displayed.

[0038] In this embodiment, we will explain using an example in which a dynamic three-dimensional head and neck model 10c and a simulated oral ingestion product 100 are applied, which are simply represented by showing the surface without displaying each individual particle. However, the present invention is not limited to this, and for example, as shown in Figure 3, a dynamic three-dimensional head and neck model 10c and a simulated oral ingestion product 100 that display each individual particle may also be applied.

[0039] Here, the organ movement setting unit 30 shown in Figure 1 sets the movement of each head and neck organ in the dynamic three-dimensional head and neck model 10c. The organ movement setting unit 30 in this embodiment comprises a forced movement setting unit 31 and a muscle contraction movement setting unit 32. The forced movement setting unit 31 sets particles that are forcibly moved in the head and neck organs when the simulated oral intake product 100 is swallowed by the dynamic three-dimensional head and neck model 10c as forced-moving particles, and sets the movement of these multiple forced-moving particles. The muscle contraction movement setting unit 32 sets particles as muscle particles in which the direction of muscle fibers is identified in the three-dimensional image for each type of muscle of the head and neck organs based on medical knowledge, and to which contractile stress is applied based on the direction of muscle fibers, and sets the movement of the muscle particles based on the contractile stress when the simulated oral intake product 100 is swallowed.

[0040] The particles forming the dynamic three-dimensional head and neck model 10c according to this embodiment are defined as one of three types: forced-movement particles, muscle particles, and particles other than these forced-movement particles and muscle particles. Based on the settings set by the organ movement setting unit 30, the dynamic three-dimensional head and neck model 10c can perform swallowing simulations by moving each head and neck organ. When there is no need to distinguish between forced-movement particles, muscle particles, and particles other than these forced-movement particles and muscle particles, these three types of particles will be collectively referred to simply as particles.

[0041] The Oral Ingestion Product Physical Property Setting Unit 40 sets the physical property values ​​of an oral ingestion product, such as food, beverages, pharmaceuticals, or quasi-drugs, to be analyzed, and forms a simulated oral ingestion product 100 modeled after the oral ingestion product in a three-dimensional image. The Oral Ingestion Product Physical Property Setting Unit 40 can set multiple simulated oral ingestion products 100 with different physical properties, such as liquids, semi-solids, or solids, to be analyzed. Semi-solids include, for example, jelly, and solids include, for example, tablets.

[0042] In this embodiment, the oral ingestion product physical property setting unit 40 sets the physical properties of the oral ingestion product, for example, the density [g / mL] of the food bolus to be ingested, the amount of food bolus to be swallowed by the dynamic three-dimensional head and neck model 10c [mL], the surface tension [N / m], the contact angle at each head and neck organ, and the slip coefficient at each head and neck organ. Here, the slip coefficient is a parameter that controls the wettability and water repellency of the biological surface and the surface of the food bolus (oral ingestion product), and can be considered as the apparent viscosity at the contact surface. When the slip coefficient is large, the friction at the interface increases, and as a result it has the effect of braking the movement of the food bolus. When the slip coefficient is small, the friction at the interface decreases, and when it is 0, it becomes like a mirror surface. A slip coefficient of 1 means that the interface is given a frictional effect equivalent to that of the viscosity of the fluid. The slip coefficient is determined by analyzing the wettability and water repellency of the assumed oral ingestion product.

[0043] The oral ingestion product physical properties setting unit 40 sets the contact angles for each head and neck organ, specifically the contact angles for the pharynx, larynx, tongue, soft palate, etc., in the dynamic three-dimensional head and neck model 10c. Furthermore, the oral ingestion product physical properties setting unit 40 sets the slip coefficients for each head and neck organ, specifically the slip coefficients for the pharynx, larynx, tongue, soft palate, etc., in the dynamic three-dimensional head and neck model 10c.

[0044] In this embodiment, the physical properties of the simulated oral ingestion product 100 may include not only the physical properties described above, but also, for example, when the simulated oral ingestion product 100 is a liquid, physical properties such as volume, viscosity, surface tension, specific gravity, thermal conductivity, and specific heat may be set. Furthermore, when the simulated oral ingestion product 100 is a solid, physical properties such as shape, dimensions, elastic modulus, tensile strength, yield point, yield stress, viscosity shear rate dependence, dynamic viscoelasticity, static viscoelasticity, compressive stress, fracture stress, fracture strain, hardness, adhesion, cohesiveness, thermal conductivity, and specific heat may be set. Moreover, when the simulated oral ingestion product 100 is semi-solid (plastic but not fluid), physical properties such as volume, viscosity, specific gravity, yield point, yield stress, viscosity shear rate dependence, dynamic viscoelasticity, static viscoelasticity, compressive stress, adhesion, and cohesiveness may be set.

[0045] The motion analysis unit 50 performs swallowing simulations using the particle method to analyze the movement of the head and neck organs during swallowing of the dynamic three-dimensional head and neck model 10c, and the behavior of the simulated oral intake product 100 during swallowing in conjunction with the movement of the head and neck organs.

[0046] Here, Figure 3 is a schematic diagram showing an example of the dynamic three-dimensional head and neck model 10c shown in Figure 2, in which the head and neck organs constituting the dynamic three-dimensional head and neck model 10c are displayed as individual particles without processing using the marching cube method, and Figure 4 is a cross-sectional view showing the cross-sectional configuration of the dynamic three-dimensional head and neck model 10c shown in Figure 3 in the midline plane. In the dynamic three-dimensional head and neck model 10c shown in Figures 2, 3, and 4, the movement of the tongue 12 (traveling wave-like motion), the rotational motion of the epiglottis 15a, the reciprocating motion of the larynx 15, and the muscle contraction motion of the pharynx 14 are reproduced by particle analysis, and move a simulated oral ingestion product 100 (not shown in Figures 2 and 3) introduced into the head and neck from the oral cavity. The movement of the simulated oral ingestion product 100 is also analyzed by particle analysis. The simulated oral ingestion product 100 is treated as a particle whether it is solid, semi-solid, or liquid.

[0047] The motion analysis unit 50 performs a swallowing simulation using the particle method after, for example, the physical properties of the simulated oral ingestion product 100 are changed by the oral ingestion product physical property setting unit 40, or the muscle activity rate is changed by the muscle activity rate changing unit 80. When the physical properties of the simulated oral ingestion product 100 are changed, the motion analysis unit 50 can obtain analysis results showing that the path through which the simulated oral ingestion product 100 is swallowed has changed through the swallowing simulation using the particle method.

[0048] Furthermore, the motion analysis unit 50 can, for example, obtain analysis results showing changes in the behavior of each muscle type, such as the traveling wave motion of the tongue 12, the elevation of the soft palate 13b, the inversion of the epiglottis 15a, the elevation of the larynx 15, the adduction of the vocal cords 15c, the forward movement of the arytenoid region 15b, and the contraction and elevation of the pharynx 14, through swallowing simulation using the particle method.

[0049] The physical property identification unit 70 estimates the volume, viscosity, and shear rate of a simulated oral ingestion product that can prevent aspiration, based on the analysis results of the motion analysis unit 50. Of these, the volume and viscosity of the simulated oral ingestion product are physical property values ​​set by the oral ingestion product physical property setting unit 40.

[0050] The control unit 90 controls various parts of the personal computer 2 to execute the functions of the swallowing simulation device 1. The control unit 90 stores the swallowing simulator (analysis software) in its internal memory.

[0051] The muscle activity rate modification unit 80 changes the muscle activity rate values ​​set for each muscle type in the dynamic three-dimensional head and neck model 10c according to the operation commands from the input unit 81. As a result, the motion analysis unit 50 can perform a swallowing simulation using the particle method with the dynamic three-dimensional head and neck model 10c, in which the muscle activity rate values ​​have been changed for each muscle type.

[0052] The muscle activity rate analysis unit 92 uses, for example, the analysis results obtained from a swallowing simulation of a dynamic three-dimensional head and neck model 10c by the motion analysis unit 50, and the muscle activity rate values ​​set at that time, to calculate an evaluation function for evaluating the muscle activity rate values ​​set for each muscle type during the swallowing simulation. For each swallowing simulation, it determines whether a steady state has been reached where the values ​​of the evaluation functions remain almost unchanged. The muscle activity rate analysis unit 92 uses the obtained evaluation function and muscle activity rate to search for the optimal muscle activity rate. Note that the muscle activity rate analysis method used by the muscle activity rate analysis unit 92 is already publicly known in Japanese Patent Application Publication No. 2021-112557, so a detailed explanation is omitted here.

[0053] The muscle synergy analysis processing unit 93 obtains the muscle activity rate of each muscle type, which changes over time in accordance with the changes in the movement of each head and neck organ in the dynamic three-dimensional head and neck model 10c during swallowing, based on the simulation results of the movement of each head and neck organ by the motion analysis unit 50. The muscle synergy analysis processing unit 93 has a configuration that allows it to identify the number of muscle synergies in which muscle types perform coordinated movements, based on the muscle synergy hypothesis that multiple muscle types are controlled by coordinated movements, from the muscle activity rate obtained for each muscle type. In this case, the muscle synergy analysis processing unit 93 generates a muscle activity matrix X representing the change over time in the muscle activity rate of each muscle type during swallowing, from the simulation results of the movement of the head and neck organs in the dynamic three-dimensional head and neck model 10c during swallowing by the motion analysis unit 50.

[0054] The developer examines the swallowing simulation of the dynamic three-dimensional head and neck model 10c obtained by the motion analysis unit 50, and the muscle activity matrix X obtained by the muscle synergy analysis processing unit 93, and estimates the number of combinations of muscle types that perform coordinated movements based on the muscle synergy hypothesis that multiple muscle types are controlled by coordinated movements. The number of muscle synergies estimated by the developer is set in the muscle synergy analysis processing unit 93.

[0055] In this embodiment, the description focuses on a case where the developer checks the swallowing simulation of the dynamic three-dimensional head and neck model 10c obtained by the motion analysis unit 50 and the muscle activity matrix X obtained by the muscle synergy analysis processing unit 93 to estimate the number of muscle synergies. However, the present invention is not limited to this. For example, the patterns of the number of muscle synergies for the swallowing simulation of the dynamic three-dimensional head and neck model 10c obtained by the motion analysis unit 50 and the muscle activity matrix X obtained by the muscle synergy analysis processing unit 93 may be learned using a machine learning model, and the number of muscle synergies may be automatically determined from these swallowing simulations and muscle activity matrix X.

[0056] The muscle synergy analysis processing unit 93 converts the muscle activity matrix X into a spatial pattern matrix W and a temporal pattern matrix H by performing non-negative matrix factorization (NMF) on the muscle activity matrix X based on the set number of muscle synergies. Here, the spatial pattern matrix W is matrix-like numerical data representing the degree of muscle activity of each constituent muscle type for each muscle synergy, and the temporal pattern matrix H is matrix-like numerical data representing the time-dependent change in the degree of activity for each muscle synergy.

[0057] The muscle synergy analysis processing unit 93 performs calculations to approximate the muscle activity matrix X as the product of the spatial pattern matrix W and the temporal pattern matrix H, and displays the obtained spatial pattern matrix W and temporal pattern matrix H on the display unit 4 and presents them to the developer. Based on the presented spatial pattern matrix W and temporal pattern matrix H and the operation status of the dynamic three-dimensional head and neck model 10c obtained from the swallowing simulation, the developer can analyze the movement of the head and neck organs during swallowing based on the number of muscle synergies and the spatial pattern matrix W and temporal pattern matrix H of each muscle synergy.

[0058] Furthermore, the simulation determination unit 94 determines whether or not analysis by the muscle synergy analysis processing unit 93 is necessary based on the swallowing simulation results showing the movement of the head and neck organs of the dynamic three-dimensional head and neck model 10c during swallowing, as determined by the motion analysis unit 50. Specifically, a machine learning model is trained to determine whether or not the swallowing simulation of the dynamic three-dimensional head and neck model 10c obtained by the motion analysis unit 50 represents the optimal swallowing motion, and based on the machine learning model, the unit automatically determines whether or not it is necessary to change the number of muscle synergies from the swallowing simulation results. For example, if the swallowing simulation of the dynamic three-dimensional head and neck model 10c obtained by the motion analysis unit 50 does not represent the optimal swallowing motion, the unit can prompt the muscle synergy analysis processing unit 93 to modify the number of muscle synergies it sets, and prompt it to perform non-negative matrix factorization on the muscle activity matrix X based on the modified number.

[0059] (3) <Construction of a dynamic three-dimensional head and neck model> Before describing the muscle synergy analysis processing unit 93, which is a characteristic configuration of this embodiment, we will first describe the dynamic three-dimensional head and neck model 10c used for muscle synergy analysis.

[0060] As described above, the display unit 4 displays, for example, a dynamic three-dimensional head and neck model 10c and a simulated oral ingestion product 100 as shown in Figure 2. The motion analysis unit 50 displays, as an analysis result, the movement of the head and neck organs when the simulated oral ingestion product 100 is swallowed in the dynamic three-dimensional head and neck model 10c, as well as the behavior of the simulated oral ingestion product 100, as moving images.

[0061] In this embodiment, as shown in Figure 2, a dynamic three-dimensional head and neck model 10c without particle display and a simulated oral ingestion product 100 are used to present the movement of the head and neck organs and the behavior of the simulated oral ingestion product 100 in a moving image. However, the present invention is not limited to this, and as shown in Figures 2 and 3, a dynamic three-dimensional head and neck model 10c with individual particles displayed and a simulated oral ingestion product 100 are used to present the movement of the head and neck organs and the behavior of the simulated oral ingestion product 100 in a moving image.

[0062] In Figures 2, 3, and 4, the X-axis represents the left-right direction of the body (direction of the surface normal to the midline plane) perpendicular to the midline plane of the dynamic three-dimensional head and neck model 10c, the Y-axis represents the front-back direction of the body perpendicular to the X-axis and parallel to the midline plane of the dynamic three-dimensional head and neck model 10c, and the Z-axis represents the up-down direction of the body perpendicular to the X-axis and Y-axis.

[0063] As shown in Figures 2, 3, and 4, the dynamic three-dimensional head and neck model 10c includes head and neck organs such as the thyroid cartilage 11b (Figure 4), tongue 12 including the geniohyoid muscle and genioglossus muscle, larynx 15, vocal cords 15c (Figure 4), arytenoid region 15b (Figure 4), epiglottis 15a, trachea 16 (Figure 4), pharyngeal region 14 (including the pharyngeal duct wall 18a and pharyngeal mucosa 18b), palate 13 (including the hard palate 13a and soft palate 13b), esophagus 17 (including the esophageal inlet 17a and esophageal duct wall 17b), and mandible 21a.

[0064] Furthermore, as shown in Figure 5A, the dynamic three-dimensional head and neck model 10c divides the genioglossus muscle into six fan-shaped regions, genioglossus (1) to (6), and as shown in Figure 5B, the transverse glossus muscle is divided into two regions, transverse glossus (1) and transverse glossus (2), in the anterior-posterior direction of the body. The muscle activity rate can be individually set for each particle constituting these muscle types in each region. In addition, the dynamic three-dimensional head and neck model 10c is also provided with the thyroid cartilage as shown in Figure 5A, and the hyoid bone and mandible as shown in Figure 6.

[0065] Furthermore, as shown in Figure 6, the dynamic three-dimensional head and neck model 10c has a configuration in which the mylohyoid muscle is divided into two regions, mylohyoid muscle (1) and mylohyoid muscle (2), and different muscle activity rates can be set for each region. In addition, as shown in Figure 7, the dynamic three-dimensional head and neck model 10c has a configuration in which the longitudinal portion of the palatopharyngeal muscle is divided into two regions, longitudinal portion of the palatopharyngeal muscle (1) and longitudinal portion of the palatopharyngeal muscle (2), and different muscle activity rates can be set for each region, and the tensor veli palatini, pterygoid, and levator veli palatini muscles are also provided.

[0066] As shown in Figures 8A, 8B, and 8C, the dynamic three-dimensional head and neck model 10c is equipped with the cricoid cartilage, the thyropharyngeal portion of the inferior pharyngeal constrictor muscle, the greater horn pharyngeal portion of the oropharynx constrictor muscle, and the lesser horn pharyngeal portion of the oropharynx constrictor muscle, in addition to the hyoid bone and thyroid cartilage. The thyropharyngeal portion of the inferior pharyngeal constrictor muscle is divided into two regions, the thyropharyngeal portion (1) and the thyropharyngeal portion (2), as shown in Figure 8A. The greater horn pharyngeal portion of the oropharynx constrictor muscle is divided into two regions, the greater horn pharyngeal portion (1) and the greater horn pharyngeal portion (2), as shown in Figure 8B. The lesser horn pharyngeal portion of the oropharynx constrictor muscle is divided into two regions, the lesser horn pharyngeal portion (1) and the lesser horn pharyngeal portion (2), as shown in Figure 8C. Each of these muscle types has a configuration that allows for setting different muscle activity rates.

[0067] In this embodiment, the following are mainly involved: tongue 12, larynx 15, vocal cords 15c, arytenoid region 15b, epiglottis 15a, trachea 16, pharynx 14, palate 13, esophagus 17, genioglossus muscle, hyoglossus muscle, inferior longitudinal lingual muscle, styloglossus muscle, superior longitudinal lingual muscle, transverse lingual muscle, vertical lingual muscle, levator veli palatini muscle, palatoglossus muscle, tensor veli palatini muscle, uvula muscle, anterior belly of digastric muscle, geniohyoid muscle, mylohyoid muscle, posterior belly of digastric muscle, stylohyoid muscle, sternohyoid muscle The sternothyroid muscle, thyrohyoid muscle, longitudinal portion of the palatinary pharyngeal muscle, transverse portion of the palatinary pharyngeal muscle, stylopharynx muscle, pterygopharyngeal and buccal pharyngeal portions of the superior pharyngeal constrictor muscle, lingual pharyngeal portion of the superior pharyngeal constrictor muscle, lesser angular pharyngeal portion of the oropharynx constrictor muscle, greater angular pharyngeal portion of the oropharynx constrictor muscle, thyropharyngeal portion of the inferior pharyngeal constrictor muscle, cricopharyngeal portion of the inferior pharyngeal constrictor muscle, and aryepiglottis muscle are collectively referred to as head and neck organs, but each of these can also be simply referred to as a head and neck organ.

[0068] In this embodiment, the simulated oral ingestion product 100 is represented by particles, and the head and neck organs in the dynamic three-dimensional head and neck model 10c are also represented by particles. However, as described above, when the analysis results of the swallowing simulation of the simulated oral ingestion product 100 by the dynamic three-dimensional head and neck model 10c are shown to developers and others as moving images, it is desirable to display the surface shapes of the head and neck organs of the dynamic three-dimensional head and neck model 10c and the simulated oral ingestion product 100 using the marching cube method or the like. This simplifies the display of the head and neck organs and the simulated oral ingestion product 100 compared to displaying each individual particle forming the head and neck organs and the simulated oral ingestion product 100, allowing developers and others to easily confirm the movement of the head and neck organs and the behavior of the simulated oral ingestion product 100.

[0069] (4) <Creation of a dynamic three-dimensional head and neck model> Such a dynamic three-dimensional head and neck model 10c can be manufactured according to the manufacturing method disclosed in Japanese Patent Publication No. 2021-112557. Specifically, the positions of the pharynx 14 and the esophageal inlet 17a are estimated from the structure of the head and neck as understood by medical knowledge, and from the morphology of the palate 13, tongue 12, and trachea 16 which can be roughly read from CT (Computed Tomography) images. The structures of head and neck organs such as the tongue 12, palate 13, pharynx 14, epiglottis 15a, and larynx 15 are then modeled using CG (Computer Graphics) software (such as Autodesk 3ds Max) to create a static initial shape model (not shown) that represents the head and neck organs involved in swallowing in three dimensions (three-dimensional structure).

[0070] The obtained static initial shape model is overlaid with contrast-enhanced images (frontal and lateral views) taken during swallowing using VF (videofluoroscopic examination of swallowing) to modify the three-dimensional structure and create a static three-dimensional head and neck model (not shown) that depicts the three-dimensional shape of the head and neck organs of the subject at rest using computer graphics. Alternatively, a static three-dimensional head and neck model can be created based on four-dimensional CT (Computed Tomography) images (4DCT images) taken during swallowing. Such static three-dimensional head and neck models are created by the head and neck modeling unit 10. Next, the surfaces of the head and neck organs in the static three-dimensional head and neck model created with computer graphics are identified, and particles are placed within the respective regions of each head and neck organ, using the surfaces of the head and neck organs as boundaries.

[0071] The particles in this embodiment are three-dimensional spherical particles (simply referred to as particles in this embodiment) that have a three-dimensional shape in a three-dimensional image. For example, when creating a life-size model of the head and neck of an infant or an average-sized adult male using a three-dimensional image, it is desirable to have a particle diameter of about 0.1 mm to 3.0 mm, and more preferably a diameter of about 0.6 mm to 1.5 mm.

[0072] Furthermore, in the static three-dimensional head and neck model created within the three-dimensional image, it is desirable that at least two particles be formed in the direction of the epiglottis thickness (for example, approximately 3.0 mm in adults and approximately 1.5 mm in infants). If the particle diameter is too small, the computational processing load on personal computer 2 becomes too large, which is undesirable. On the other hand, if the particle diameter is too large, it becomes impossible to reproduce the fine movements of the head and neck organs. Therefore, it is desirable that the particle diameter be within the above range.

[0073] In this embodiment, we describe a case where three-dimensional spherical particles are used as particles to form the head and neck organs. However, the present invention is not limited to this, and the head and neck organs may be formed using particles of various shapes, such as rectangular parallelepipeds or polygonal shapes.

[0074] When creating the dynamic three-dimensional head and neck model 10c, the surface of each head and neck organ is identified in the static three-dimensional head and neck model created by computer graphics based on CT and VF images. Then, particles are arranged without gaps so that they are in contact with each other within the region surrounded by the surface, thereby creating a particle model for each head and neck organ.

[0075] Furthermore, when performing a swallowing simulation in which a simulated oral ingestion product is swallowed using a dynamic three-dimensional head and neck model 10c, the simulated oral ingestion product 100 is also modeled using particles by first identifying the surface of the simulated oral ingestion product 100 created by computer graphics, and then arranging particles without gaps so that they are in contact with each other within the area surrounded by that surface.

[0076] In this way, the process of creating each head and neck organ of the static three-dimensional head and neck model and the simulated oral ingestion product 100 using particles is performed by the head and neck modeling unit 10.

[0077] In this embodiment, the dynamic three-dimensional head and neck model 10c has a plurality of particles, of which particles located in a predetermined region are set as forced-movement particles that are forcibly moved by the head and neck organs when the simulated oral intake product 100 is swallowed. In addition, in this embodiment, the dynamic three-dimensional head and neck model 10c has a plurality of particles, of which particles other than those designated as forced-movement particles are located in a predetermined region, and are set as muscle particles to which contractile stress is applied based on the direction of muscle fibers.

[0078] Furthermore, particles other than forced-moving particles and muscle particles are those for which the position they are forcibly moved to in the head and neck organs during swallowing (i.e., the coordinates they move to in the three-dimensional image during swallowing) is not defined, and the contractile stress in the direction of muscle fibers, like that of muscle particles, is also not defined. When performing swallowing simulations with the dynamic three-dimensional head and neck model 10c, the movement position and other properties of such particles other than forced-moving particles and muscle particles are analyzed using conventional particle methods.

[0079] Details of conventional swallowing simulations for particles other than forced-movement particles and muscle particles are disclosed in the paper "Kikuchi, T., Michiwaki, Y., Koshizuka, S., Kamiya, T., and Toyama Y., “Numerical simulation of interaction between organs and food bolus during swallowing and aspiration,” Computers in Biology and Medicine, 80, (2017), pp. 114-123.", so we will omit that explanation here and focus on forced-movement particles and muscle particles in the following explanation.

[0080] Furthermore, regarding the forced-movement particles described below, simulations using the particle method are also disclosed in the document "Kikuchi, T., Michiwaki, Y., Kamiya, T. et al. Comp. Part. Mech. (2015) 2: 247. “Human swallowing simulation based on videofluorography images using Hamiltonian MPS method”".

[0081] The dynamic three-dimensional head and neck model 10c has forced-movement particles and muscle particles among the particles that make up the head and neck organs. The dynamic three-dimensional head and neck model 10c sets some of the particles of the head and neck organs as forced-movement particles, and uses these forced-movement particles to model the movement of the muscles that are considered to be the main active muscles during swallowing. In this case, the dynamic three-dimensional head and neck model 10c selects particles that give rigid forced displacement during swallowing from among the particles that make up the tongue 12, palate 13, pharynx 14, epiglottis 15a, larynx 15, and esophagus 17, etc., and sets these as forced-movement particles 19.

[0082] The forced-movement particles 19 are selected based on anatomical knowledge and findings from medical image analysis studies so as to reflect the muscle movements of the head and neck organs when a subject actually swallows an orally ingested product. In this embodiment, the head and neck organs are traced in VF images and 4DCT images obtained when a subject swallows a predetermined orally ingested product, and forced-movement particles 19 are selected to be forcibly moved within the dynamic three-dimensional head and neck model 10c. As shown in Figure 4, the dynamic three-dimensional head and neck model 10c can include a tongue forced-movement section 19a, a soft palate forced-movement section 19b, an epiglottis forced-movement section 19c, a thyroid cartilage forced-movement section 19d, a trachea forced-movement section 19e, and an esophageal anterior wall forced-movement section 19f as needed, but it is not necessary to use all forced-movement sections, and it is possible to perform the analysis without placing any forced-movement sections depending on the purpose of the analysis.

[0083] Furthermore, based on VF images and 4DCT images obtained when a subject swallows a predetermined oral intake product, the position in which each forced-moving particle 19 moves within the three-dimensional image is determined at predetermined time intervals (for example, 0.1 seconds (hereinafter, seconds will also be referred to as S)) from the start to the end of swallowing, and the time and position at the time of swallowing are set for each forced-moving particle 19.

[0084] In other words, in the dynamic three-dimensional head and neck model 10c, for example, at 0.0S, which is the start of swallowing, the forced-movement particle 19 whose coordinates on the X, Y, and Z axes in the three-dimensional image are (0.0, 0.2, -0.2) is set to move to coordinates (0.0, 0.2, 0.0) at 0.1S from the start of swallowing, and to coordinates (0.0, 0.3, 0.3) at 0.2S. The particles shown as black circles in Figures 3 and 4 represent the forced-movement particles 19 in this embodiment, and are set in parts of the tongue 12, palate 13, and larynx 15, for example.

[0085] Here, Figure 9 is a schematic diagram showing the state changes of the dynamic three-dimensional head and neck models 10c1, 10c2, 10c3, and 10c4 when the forced-movement particles 19 move in the dynamic three-dimensional head and neck model 10c shown in Figure 4, with a portion of the trajectory of the forced-movement particles 19 moving during swallowing represented by movement trajectory lines 22a, 22b, 22c, and 22d (hereinafter collectively referred to as movement trajectory line 22).

[0086] For example, 22a shows the trajectory of a forced-movement particle 19 set on the tongue 12, 22b shows the trajectory of a forced-movement particle 19 set on the soft palate of the palate 13, 22c shows the trajectory of a forced-movement particle 19 set on the larynx 15, and 22d shows the trajectory of a forced-movement particle 19 set on the tubular wall of the esophagus 17.

[0087] Figure 9 shows, as an example, a dynamic three-dimensional head and neck model 10c swallowing a predetermined simulated oral intake product in approximately 12S. The figures show the dynamic three-dimensional head and neck model 10c1 at 0S, the start of swallowing; the dynamic three-dimensional head and neck model 10c2 approximately 7S after the start of swallowing; the dynamic three-dimensional head and neck model 10c3 approximately 9S after the start of swallowing; and the dynamic three-dimensional head and neck model 10c4 approximately 11S after the start of swallowing.

[0088] In this way, the dynamic three-dimensional head and neck model 10c is configured to reproduce the basic movements of the head and neck organs during swallowing (traveling wave motion, rotational motion, vertical motion, anterior-posterior motion, contraction motion, etc.) by setting the position and time at which each forced-movement particle 19 moves during swallowing in advance. The movement of the forced-movement particles 19 in this dynamic three-dimensional head and neck model 10c is set in the forced motion setting unit 31 of the organ motion setting unit 30.

[0089] In order to accurately reproduce the behavior of the pharyngeal region 14, etc., during swallowing in the dynamic three-dimensional head and neck model 10c, it is desirable to reproduce the movement in which the length of the wall surface of the pharyngeal region 14, etc. shortens during swallowing. Therefore, in the dynamic three-dimensional head and neck model 10c of this embodiment, in addition to simply applying rigid forced displacement to particles such as the pharyngeal region 14, etc., a muscle fiber direction is set for each muscle particle in the three-dimensional image for each type of muscle such as the pharyngeal region 14, etc., and an optimal contractile stress based on the muscle fiber direction is applied to the muscle particle by the muscle activity rate, thereby accurately reproducing the behavior during swallowing in the dynamic three-dimensional head and neck model 10c. In the dynamic three-dimensional head and neck model 10c, particles to which contractile stress based on the muscle fiber direction is applied are called muscle particles.

[0090] In a dynamic three-dimensional head and neck model 10c, for example, when defining muscle particles from the particles 20a that form the tongue 12, as shown in Figure 10, the muscle regions ER1, ER2, etc., where the tongue 12 expands and contracts during swallowing are identified in the three-dimensional image based on anatomical knowledge, VF images, 4DCT images, etc., and the particles 20a present within each muscle region ER1, ER2 are searched for. For example, the particles 20a within the muscle region ER1 are designated as the muscle particles of the tongue 12, and the muscle fiber direction is defined for each muscle particle in the virtual space of the three-dimensional image.

[0091] The details of the muscle fiber direction set for muscle particles will be described later, but based on anatomical knowledge, VF images, 4DCT images, etc., multiple line segments A are set within the space of the muscle body region ER1 of the tongue 12, representing the direction in which muscle contraction occurs during swallowing. For each muscle particle, the weighted average of the directions of each nearby line segment A is used as the muscle fiber direction. In this embodiment, for example, two line segments are identified: the first line segment closest to the muscle particle and the second line segment second closest to the muscle particle. The distance from the muscle particle to the first line segment and the distance from the muscle particle to the second line segment are weighted according to their proximity to the muscle particle, and the directions of the first and second line segments are averaged to determine the muscle fiber direction. However, the method for determining the muscle fiber direction is not limited to this method. For example, a more smoothly spatially distributed muscle fiber direction can be obtained by performing radial basis function interpolation using all line segments defined within the muscle body region ER1.

[0092] Furthermore, in the dynamic three-dimensional head and neck model 10c, when a simulated oral intake is swallowed, the temporal change in the contractile stress of muscle particles that occurs for each type of muscle in the head and neck organs is set as the muscle activity rate, and the magnitude of the contractile stress during swallowing is determined by the muscle activity rate.

[0093] Here, we will first focus on the pharyngeal region 14 in the dynamic three-dimensional head and neck model 10c and explain the muscle fiber directions set for each type of muscle in the pharyngeal region 14. The left figure in Figure 11 is a schematic diagram of the muscle fiber model 10d, which shows the directions Aa, Ab, Ac, Ad, Ae, Af, Ag, Ah of the pharyngeal constrictor muscles in the static three-dimensional head and neck model 10b. The right figure in Figure 11 is a schematic diagram of the dynamic three-dimensional head and neck model 10e, which shows the direction Aa, Ab, Ac, Ad, Ae, Af, Ag, Ah of the pharyngeal constrictor muscles in the dynamic three-dimensional head and neck model 10c, based on the directions Aa, Ab, Ac, Ad, Ae, Af, Ag, Ah of the pharyngeal constrictor muscles shown in the muscle fiber model 10d on the left figure, and shows the muscle fiber direction of each individual muscle particle in the pharyngeal region 14.

[0094] In this embodiment, as shown in the muscle fiber model 10d and the dynamic three-dimensional head and neck model 10e, the pharyngeal region 14 is divided based on anatomical findings into the lingual pharyngeal region 14a of the superior pharyngeal constrictor muscle, the upper part of the lesser horn pharynx 14b of the oropharynx constrictor muscle, the lower part of the lesser horn pharynx 14c of the oropharynx constrictor muscle, the upper part of the greater horn pharynx 14d of the oropharynx constrictor muscle, the lower part of the greater horn pharynx 14e of the oropharynx constrictor muscle, the upper part of the thyropharynx 14f of the inferior pharyngeal constrictor muscle, the lower part of the thyropharynx 14g of the inferior pharyngeal constrictor muscle, and the cricopharyngeal region 14h of the inferior pharyngeal constrictor muscle.

[0095] For the lingual pharyngeal portion 14a of the superior pharyngeal constrictor muscle, the upper part of the lesser horn pharynx 14b of the oropharynx constrictor muscle, the lower part of the lesser horn pharynx 14c of the oropharynx constrictor muscle, the upper part of the greater horn pharynx 14d of the oropharynx constrictor muscle, the lower part of the greater horn pharynx 14e of the oropharynx constrictor muscle, the upper part of the thyropharynx 14f of the inferior pharyngeal constrictor muscle, the lower part of the thyropharynx 14g of the inferior pharyngeal constrictor muscle, and the cricopharynx portion 14h of the inferior pharyngeal constrictor muscle, particles that represent muscle particles are arranged without gaps within each region, and the model is performed using particles.

[0096] Furthermore, for the superior pharyngeal constrictor muscle (lingual pharyngeal portion 14a), the middle pharyngeal constrictor muscle (upper part of the lesser horn pharynx 14b), the middle pharyngeal constrictor muscle (lower part of the lesser horn pharynx 14c), the middle pharyngeal constrictor muscle (upper part of the greater horn pharynx 14d), the middle pharyngeal constrictor muscle (lower part of the greater horn pharynx 14e), the inferior pharyngeal constrictor muscle (upper part of the thyropharynx 14f), the inferior pharyngeal constrictor muscle (lower part of the thyropharynx 14g), and the inferior pharyngeal constrictor muscle (cricopharynx portion 14h), the directions in which the constrictor muscles run (Aa, Ab, Ac, Ad, Ae, Af, Ag, Ah) were determined based on anatomical knowledge and VF images, etc., by first setting the fine muscle contraction directions that occur for each part of each constrictor muscle during swallowing as line segments, and then weighting and averaging the directions of the nearby line segments for each muscle particle. Note that the directions Aa, Ab, Ac, Ad, Ae, Af, Ag, and Ah shown on the left side of Figure 11, where the contractile muscles run, represent the approximate direction of the muscle fibers for the sake of explanation. Each individual muscle particle in the pharyngeal region 14 defines its muscle fiber direction according to these directions Aa, Ab, Ac, Ad, Ae, Af, Ag, and Ah where the contractile muscles run.

[0097] Each muscle particle located in the lingual pharyngeal portion 14a of the superior pharyngeal constrictor muscle, the upper part of the lesser pharyngeal portion 14b of the oropharynx constrictor muscle, the lower part of the lesser pharyngeal portion 14c of the oropharynx constrictor muscle, the upper part of the greater pharyngeal portion 14d of the oropharynx constrictor muscle, the lower part of the greater pharyngeal portion 14e of the oropharynx constrictor muscle, the upper part of the thyropharynx constrictor muscle, the lower part of the thyropharynx constrictor muscle 14g of the inferior pharyngeal constrictor muscle, and the cricopharyngeal portion 14h of the inferior pharyngeal constrictor muscle is set to have a muscle activity rate that corresponds to the time progression during swallowing.

[0098] In this embodiment, the swallowing simulation device 1 allows the developer to repeatedly perform swallowing simulations while changing the muscle activity rate values ​​of each head and neck organ that constitutes the dynamic three-dimensional head and neck model 10c, and analyze the behavior of the dynamic three-dimensional head and neck model 10c during swallowing simulations to determine the muscle activity rate at which the dynamic three-dimensional head and neck model 10c can perform the desired movements during swallowing simulations.

[0099] When performing swallowing simulations by representing the movement of head and neck organs in a dynamic three-dimensional head and neck model 10c and the behavior of simulated oral ingestion products in a three-dimensional image, the muscle particles of the head and neck organs in the dynamic three-dimensional head and neck model 10c are analyzed as Mooney-Rivlin bodies using a particle method (e.g., Hamiltonian particle method: Hamiltonian MPS method) in the motion analysis unit 50.

[0100] Furthermore, the dynamic three-dimensional head and neck model 10c according to this embodiment applies the dynamic three-dimensional head and neck model disclosed in Japanese Patent Application Publication No. 2021-112557, for example, the governing equations when internal and external forces are incorporated, and the elastic force f applied to the muscle particle i. i,elastic , Contact force f applied to muscle particle i i,contact , the fluid force f from the pseudo-oral ingested substance applied to the muscle particle i i,interaction Since the settings for muscle activity levels and other aspects are the same, we will omit the explanation here.

[0101] (5) <Analysis method of muscle synergy using a dynamic three-dimensional head and neck model> Next, the method for analyzing muscle synergies during swallowing will be described below using the dynamic three-dimensional head and neck model 10c described above. In this embodiment, the following steps are performed: (i) a muscle activity matrix generation step to generate a muscle activity matrix X; (ii) a non-negative matrix factorization calculation process to obtain the muscle activity rate of each muscle type obtained from the swallowing simulation results of the dynamic three-dimensional head and neck model 10c and perform non-negative matrix factorization; (iii) a time division process to perform time division; (iv) a matrix analysis step to analyze the time division muscle activity matrix, spatial pattern matrix, and time division time pattern matrix; (v) an interpolation process to perform data interpolation on the time division muscle activity matrix obtained in the time division process; and (vi) an aggregation process to muscle synergies for each muscle type. These steps will be explained in order below.

[0102] In the above-described embodiment, the case in which (i) muscle activity matrix generation step, (ii) non-negative matrix factorization calculation step, (iii) time division processing step, (iv) matrix analysis step, (v) interpolation processing step, and (vi) aggregation processing step are performed will be explained, but the present invention is not limited thereto. For example, of these, (iii) time division processing step, (iv) matrix analysis step, and (v) interpolation processing step may be omitted, and only (i) muscle activity matrix generation step, (ii) non-negative matrix factorization calculation step, and (vi) aggregation processing step may be performed, or (ii) non-negative matrix factorization calculation step may be omitted, and only (i) muscle activity matrix generation step, (iii) time division processing step, (iv) matrix analysis step, (v) interpolation processing step, and (vi) aggregation processing step may be performed.

[0103] (5-1)<Muscle activity matrix generation process> In this embodiment, the dynamic three-dimensional head and neck model 10c is configured based on the technical details disclosed in Japanese Patent Publication No. 2021-112557 (see Japanese Patent Publication No. 2021-112557).

[0104] The muscle synergy analysis processing unit 93 acquires the muscle activity rate of each muscle type from the start to the end of swallowing during the swallowing simulation of the dynamic three-dimensional head and neck model 10c performed by the motion analysis unit 50. Figures 12A and 12B show the results obtained by the muscle synergy analysis processing unit 93 of the muscle activity rate of each muscle type from the start to the end of swallowing during the swallowing simulation of the dynamic three-dimensional head and neck model 10c. The horizontal axis of each graph represents time, and the vertical axis represents muscle activity rate. In this embodiment, the muscle synergy analysis processing unit 93 acquires the muscle activity rates of 40 types of muscles set in the dynamic three-dimensional head and neck model 10c.

[0105] The muscle synergy analysis processing unit 93 generates a muscle activity matrix X representing the time-dependent changes in muscle activity rates for each muscle type during swallowing, based on the swallowing simulation results of the movement of head and neck organs during swallowing by the motion analysis unit 50. The muscle activity matrix X is numerical data in which the muscle activity rates for each muscle type and the sampling times from the start to the end of swallowing are arranged in a matrix. For example, as shown in Figure 13, the muscle types for which muscle activity rates have been obtained are shown in vertical columns, and the sampling times from the start to the end of swallowing for which muscle activity rates have been obtained are shown in horizontal rows to generate a muscle activity matrix X representing the time-dependent changes in muscle activity rates for each muscle type.

[0106] (5-2) <Non-negative matrix factorization calculation process> Next, we will explain the process of performing non-negative matrix factorization on the muscle activity matrix X described above, and converting the muscle activity matrix X into a spatial pattern matrix W and a temporal pattern matrix H.

[0107] The muscle synergy analysis processing unit 93 receives a numerical value representing the number of muscle synergies arbitrarily selected by the developer based on the muscle synergy hypothesis that multiple muscle types are controlled by coordinated movement, via the input unit 81, and the number of such muscle synergies is set.

[0108] The muscle synergy analysis processing unit 93 performs non-negative matrix factorization on the muscle activity matrix X using the set number of muscle synergies, thereby converting the muscle activity matrix X into a spatial pattern matrix W representing the degree of muscle activity for each constituent muscle type for each muscle synergy, and a temporal pattern matrix H representing the time-series data (data of changes over time) of the degree of activity for each muscle synergy. In this embodiment, for the sake of simplicity, the number of muscle synergies to be set will be set to 9 for the following explanation, but the number of muscle synergies is not limited to this.

[0109] Specifically, the spatial pattern matrix W obtained by performing non-negative matrix factorization on the muscle activity matrix X is a matrix-like numerical data set in which, for example, 40 types of muscles (e.g., genioglossus (1), genioglossus (2), etc.) are shown in vertical columns and 9 muscle synergies are shown in horizontal rows, with each of the 9 muscle synergies representing the muscle activity rate for each type of muscle.

[0110] Figures 14, 15, and 16 show an example of the detailed structure of the nine muscle synergy data (Syn1-Syn9) shown in the horizontal rows of the spatial pattern matrix W. Muscle synergies Syn1-Syn9 represent 40 types of muscles on the vertical axis and the data on the horizontal axis showing the relative degree of muscle activity for each muscle type during swallowing for each muscle synergy Syn1-Syn9.

[0111] The muscle synergy analysis processing unit 93 presents the developer with a spatial pattern matrix W with a data structure as shown in Figures 14, 15, and 16, allowing the developer to confirm the differences in the degree of muscle activity of each muscle type that changes for each muscle synergy Syn1 to Syn9, and to infer the muscle type controlled by coordinated movement for each muscle synergy Syn1 to Syn9.

[0112] The time pattern matrix H obtained by performing non-negative matrix factorization on the muscle activity matrix X is, for example, a matrix-like numerical data set in which muscle synergies are shown in vertical columns and the sampling times from the start to the end of swallowing, when the muscle activity rate was measured, are shown in horizontal rows, representing the change in the degree of activity of each muscle synergy over time.

[0113] Figure 17 is a graph that combines the nine time patterns obtained for each muscle synergy Syn1 to Syn9 in the time pattern matrix H. In Figure 17, the horizontal axis shows the time from the start to the end of swallowing, and the vertical axis shows the degree of activity of muscle synergies Syn1 to Syn9.

[0114] As shown in Figure 17, the time pattern matrix H allows us to obtain the time patterns for each muscle synergy Syn1 to Syn9, enabling developers to confirm which of the muscle synergies Syn1 to Syn9 is active during which time period.

[0115] (5-3) <Time division processing process> Next, we will explain the time-division processing step, which involves analyzing the muscle activity matrix X described above and performing time-division of the muscle activity matrix X and the time pattern matrix H.

[0116] When reproducing swallowing movements with the dynamic three-dimensional head and neck model 10c, there are characteristic movements within the series of swallowing actions that require accurate reproduction of the head and neck organs. On the other hand, there are also head and neck organ movements that have little impact on the overall examination of head and neck organ movements during swallowing, even if the accuracy of the swallowing motion reproduction by the dynamic three-dimensional head and neck model 10c is low. Therefore, it can be said that there is a region ER3 in the muscle activity matrix X shown in Figure 13 that is important for accurately reproducing swallowing movements with the dynamic three-dimensional head and neck model 10c.

[0117] Here, if we simply perform non-negative matrix factorization on the muscle activity matrix X, the entire muscle activity matrix X will only be approximated by the spatial pattern matrix W and the temporal pattern matrix H with uniform accuracy. Therefore, in this embodiment, it is desirable to time-partition the muscle activity matrix X or impose constraints on the temporal pattern matrix H while considering the characteristic movements of each head and neck organ during swallowing. Here, we will first describe the time-partitioning process for time-partitioning the muscle activity matrix X, and the temporal pattern analysis process for imposing constraints on the temporal pattern matrix H will be described later in "(5-4)<Matrix Analysis Process>".

[0118] In this case, the muscle synergy analysis processing unit 93 obtains a muscle activity matrix X based on the results of a swallowing simulation that reproduces the movement of the dynamic three-dimensional head and neck model 10c during swallowing, and presents the muscle activity matrix X to the developer via the display unit 4. As a result, the swallowing simulation device 1 uses the results of the swallowing simulation (moving image) that reproduces the movement of the dynamic three-dimensional head and neck model 10c during swallowing, and the obtained muscle activity matrix X, to allow the developer to identify the region ER1 in the muscle activity matrix X, which is important for accurately reproducing the characteristic movements of the head and neck organs during swallowing, based on the segmentation method described later.

[0119] The muscle synergy analysis processing unit 93, based on the important region ER1 identified by the developer within the muscle activity matrix X, prompts the developer to specify the time period to be divided from the muscle activity matrix X, and based on the obtained division specification information, divides the muscle activity matrix X into multiple time domains along the time series, for example as shown in 18A of Figure 18, and generates multiple time-divided muscle activity matrices X from the muscle activity matrix X. n This generates the following. In this embodiment, we show an example in which the muscle activity matrix X is divided into six time domains along the time series to generate six time-divided muscle activity matrices X1 to X6.

[0120] In this embodiment, the division method is as follows: the time period when the activity of the genioglossus muscle (3), shown in 5A of Figure 5, is still small is designated as the time period X1; the time period when only the genioglossus muscle (3) is highly active and the genioglossus muscle (4) is not yet active is designated as the time period X2; the time period when the genioglossus muscle (4) is also active and the muscles that elevate the hyoid bone are not yet active is designated as the time period X3; the time period when the hyoid bone is elevated and the pharynx is gradually closed is designated as the time period X4; the time period just before the muscles that elevate the hyoid bone finish their activity and the cricopharyngeal portion of the inferior pharyngeal constrictor muscle is active is designated as the time period X5; and the time period when the pharynx is opened and the hyoid bone is also reduced is designated as the time period X6.

[0121] The genioglossus muscle (3) and the genioglossus muscle (4) are important parts for controlling the position of the hyoid bone and greatly influence the shape of the tongue. In addition, the cricopharyngeal portion of the inferior pharyngeal constrictor muscle is the part that moves the food bolus from the esophageal entrance into the esophagus. Furthermore, in addition to the genioglossus muscle and the cricopharyngeal portion of the inferior pharyngeal constrictor muscle, the activity of other muscles such as the geniohyoid muscle and mylohyoid muscle may also be considered when performing time segmentation.

[0122] Multiple time-divided muscle activity matrices X n When dividing the time into segments, it is desirable to analyze the movement of the hyoid bone from the simulation results obtained from the swallowing simulation using the dynamic three-dimensional head and neck model 10c, as shown in Figure 19, or to analyze the muscle activity rate of each muscle type measured from the simulation results obtained from the swallowing simulation using the dynamic three-dimensional head and neck model 10c, as shown in 20A to 20D of Figure 20, and to determine the characteristic movements of the head and neck organs by analyzing basic medical knowledge such as anatomy and physiology, and clinical medical knowledge such as rehabilitation science, otolaryngology, dentistry, physical therapy, and speech therapy.

[0123] As shown in 18A of Figure 18, when the muscle activity matrix X is divided into six time domains along the time series (time division), the time pattern matrix H obtained by performing non-negative matrix factorization on the muscle activity matrix X can also be divided into six time domains in the same time domains from which the muscle activity matrix X was divided.

[0124] The muscle synergy analysis processing unit 93 divides the time pattern matrix H into multiple time domains along the time series, based on the six time domains into which the muscle activity matrix X is divided, and generates multiple time-divided time pattern matrices H from the time pattern matrix H. n This generates the following. In this embodiment, we show an example in which the time pattern matrix H is divided into the same six time domains along the time series to generate six time-divided time pattern matrices H1 to H6, corresponding to the time-divided muscle activity matrices X1 to X6 obtained by dividing the muscle activity matrix X into six time domains.

[0125] (5-4)<Matrix analysis process> Next, we will explain the matrix analysis process, which analyzes the time-divided time pattern matrices H1 to H6, etc., obtained by time-dividing the time pattern matrix H in the above-mentioned "(5-3) <Time Division Processing Process>". Here, the time-divided time pattern matrices H1 to H6 are, for example, data obtained by dividing the time pattern matrix H shown in Figure 17 into six time domains along the time series.

[0126] As described above, the muscle synergy analysis processing unit 93 divides the time pattern matrix H into the same six time domains based on the six time domains obtained by dividing the muscle activity matrix X, generating six time-division time pattern matrices H1 to H6. For each time domain of the time-division time pattern matrices H1 to H6, the unit identifies the activity level of each muscle synergy Syn1 to Syn9.

[0127] For example, as shown in Figures 19 and 20A of Figure 20, the time domain of the first time-division muscle activity matrix X1 is 0.8 to 1.1 [S], and in the same time domain of the first time-division time pattern matrix H1 between 0.8 and 1.1 [S], as shown in Figure 17, only the activity level of the first muscle synergy Syn1 has a numerical value.

[0128] As a result, the muscle synergy analysis processing unit 93 defines the row for the first muscle synergy Syn1 as a valid value "1" in the first time-division time pattern matrix H1, as shown in 18B of Figure 18, and defines the other rows from the second muscle synergy Syn2 to the ninth muscle synergy Syn9 as invalid values ​​"0".

[0129] Furthermore, as shown in Figures 19 and 20A, for example, the time domain of the second time-division muscle activity matrix X2 is 1.2[S], and in the time domain of the second time-division time pattern matrix H2 at the same 1.2[S], as shown in Figure 17, only the activity level of the second muscle synergy Syn2 has a numerical value.

[0130] As a result, the muscle synergy analysis processing unit 93 defines the row for the second muscle synergy Syn2 in the second time-division time pattern matrix H2 as a valid value of "1", and the rows for the other muscle synergies, from the first muscle synergy Syn1 to the ninth muscle synergy Syn9, as invalid values ​​of "0", as shown in 18B of Figure 18. In the same way, the muscle synergy analysis processing unit 93 defines valid and invalid values ​​for the third time-division time pattern matrix H3, the fifth time-division time pattern matrix H5, and the sixth time-division time pattern matrix H6.

[0131] Here, the time domain of the fourth time-division muscle activity matrix X4 is 1.4 to 1.9 [S], as shown in Figures 19 and 20C and 20D, for example. In the time domain of the fourth time-division time pattern matrix H4, which is also between 1.4 and 1.9 [S], the activity levels of multiple muscle synergies from the fourth muscle synergy Syn4 to the seventh muscle synergy Syn7 have numerical values, as shown in Figure 17.

[0132] In Figure 17, within the time domain of the fourth time-division time pattern matrix H4, which is 1.4 to 1.9 [S], the activity level of the fourth muscle synergy Syn4 is numerically represented between 1.4 and 1.6 [S], the activity level of the fifth muscle synergy Syn5 is numerically represented between 1.5 and 1.8 [S], the activity level of the sixth muscle synergy Syn6 is numerically represented between 1.7 and 1.9 [S], and the activity level of the seventh muscle synergy Syn7 is numerically represented at 1.9 [S].

[0133] Therefore, as shown in 18B of Figure 18, the muscle synergy analysis processing unit 93 defines the fourth time-division time pattern matrix H4 as having a valid value of "1" for the time periods in the rows of the fourth muscle synergy Syn4 to the seventh muscle synergy Syn7, respectively, where the activity level has a numerical value, and defines the other time periods in which the muscle synergy activity rate does not have a numerical value, as well as the rows of the first muscle synergy Syn1 to the third muscle synergy Syn3, the eighth muscle synergy Syn8, and the ninth muscle synergy Syn9, respectively, as invalid values ​​of "0". As a result, in the fourth time-division time pattern matrix H4, the regions defined as having a valid value of "1" are defined diagonally in a stepped manner.

[0134] In this way, the muscle synergy analysis processing unit 93 identifies the activity level of each muscle synergy Syn1 to Syn9 for each time domain of the time-division time pattern matrices H1 to H6, and defines the rows and time periods of muscle synergies Syn1 to Syn9 that have an activity level as a valid value of "1" in the time-division time pattern matrices H1 to H6, and the rows of muscle synergies Syn1 to Syn9 that do not have such an activity level as an invalid value of "0".

[0135] Here, as shown in 21A of Figure 21, the first time-divided muscle activity matrix X1 is approximated by the first spatial pattern matrix W1 and the first time-divided temporal pattern matrix H1. First, the first time-divided muscle activity matrix X1 is transformed into the first spatial pattern matrix W1 and the first time-divided temporal pattern matrix H1 (i.e., X1 ≈ W1·H1) by factorization of the matrix product. Next, an initial effective value of "1" is given to the first time-divided temporal pattern matrix H1, and iterative calculations are performed using non-negative matrix factorization to update the values ​​of the first spatial pattern matrix W1 and the first time-divided temporal pattern matrix H1. The time-divided muscle activity matrix X1 is represented as the product of the updated first spatial pattern matrix W1 and the row of muscle synergy Syn1 in the updated first time-divided temporal pattern matrix H1. Similarly, as shown in 21C of Figure 21, the fourth time-division muscle activity matrix X4 is approximated by the product of the fourth to seventh spatial pattern matrices W4 to W7 and the fourth time-division time pattern matrix H4, through iterative calculation using non-negative matrix factorization, and the sixth time-division muscle activity matrix X6 is approximated by the product of the ninth spatial pattern matrix W9 and the sixth time-division time pattern matrix H6, through iterative calculation using non-negative matrix factorization.

[0136] As shown in 21B of Figure 21, the second time-division muscle activity matrix X2 is represented as the product of the second spatial pattern matrix W2 (= X2) and the row of muscle synergy Syn2 that gave a valid value of "1" in the second time-division time pattern matrix H2. Similarly, the third time-division muscle activity matrix X3 is represented as the product of the third spatial pattern matrix W3 (= X3) and the row of muscle synergy Syn3 that gave a valid value of "1" in the third time-division time pattern matrix H3, and the fifth time-division muscle activity matrix X5 is represented as the product of the eighth spatial pattern matrix W8 (= X5) and the row of muscle synergy Syn8 that gave a valid value of "1" in the fifth time-division time pattern matrix H5. Furthermore, the fourth time-division muscle activity matrix X4 is represented as the product of the corresponding spatial pattern matrices W4 to W7 (from the fourth to the seventh) and the rows of muscle synergies Syn4 to Syn7 of the fourth time-division temporal pattern matrix H4.

[0137] (5-5) Interpolation Process Next, the interpolation process for data interpolation will be described. Here, since the fourth time-division muscle activity matrix X4 and the fourth time-division time pattern matrix H4 in the time domain of 1.4 to 1.9 [S] contain multiple muscle synergies Syn4 to Syn7, it is desirable to further refine the 0.1 [S] time step in the time domain of 1.4 to 1.9 [S] and interpolate the muscle activity rates of each muscle type during swallowing simulation using the dynamic three-dimensional head and neck model 10c to smooth the operation of the dynamic three-dimensional head and neck model 10c.

[0138] Figure 22 shows data regarding the muscle activity rates of the genioglossus muscles (1) to (6) and the hyoglossus muscle, etc., during the time domain of the fourth time-division muscle activity matrix X4, from 1.4 [S] to 1.9 [S]. As shown in Figure 22, the time domain of the fourth time-division muscle activity matrix X4 has a time step of 0.1 [S], and the muscle activity rates are not specified between 1.4 [S] and 1.5 [S]. Therefore, the muscle synergy analysis processing unit 93 interpolates the muscle activity rates in the time domain of the fourth time-division muscle activity matrix X4, including the time period between 1.4 [S] and 1.5 [S], using linear interpolation.

[0139] Subsequently, the muscle synergy analysis processing unit 93 performs non-negative matrix factorization on the fourth time-divided muscle activity matrix X4, which is obtained by interpolating the muscle activity rates between time steps, using four set values, which are the number of divisions for muscle synergies Syn4 to Syn7. This converts the fourth time-divided muscle activity matrix X4 into a new spatial pattern matrix W4 to W7 and a fourth time-divided time pattern matrix H4.

[0140] The muscle synergy analysis processing unit 93 linearly interpolates the muscle activity rate of the fourth time-division muscle activity matrix X4 as described above, and integrates the newly obtained fourth time-division time pattern matrix H4 with the remaining first time-division time pattern matrices H1 to H3, the fifth time-division time pattern matrix H5, and the sixth time-division time pattern matrix H6, thereby obtaining new time patterns representing the time-dependent changes in the degree of activity for each muscle synergy Syn1 to Syn9, as shown in Figure 17.

[0141] In the embodiments described above, the spatial pattern matrices W4 to W7 and the fourth time-divided time pattern matrix H4 were obtained by performing non-negative matrix factorization on a fourth time-divided muscle activity matrix X4 obtained by linear interpolation of muscle activity rates. However, the present invention is not limited to this. For example, the spatial pattern matrix W and the time pattern matrix H may be obtained by performing non-negative matrix factorization on a muscle activity matrix X obtained by linear interpolation of muscle activity rates over the time domain of the fourth time-divided muscle activity matrix X4.

[0142] (5-6) <Integration Processing Process> As shown in Figure 17, the muscle synergy analysis processing unit 93 obtains time patterns representing the change in activity level over time for each muscle synergy Syn1 to Syn9, and then performs an aggregation process to consolidate the muscle activity rates obtained for each of the 40 muscle types into a muscle activity rate that reflects the results of muscle synergies Syn1 to Syn9 for each muscle type.

[0143] Specifically, for example, in the first muscle synergy Syn1 shown in Figure 14, we calculate the weight each muscle type occupies within the first muscle synergy Syn1. Since the palatoglossus muscle occupies the largest weight in the first muscle synergy Syn1, we will focus on the palatoglossus muscle in the following explanation.

[0144] The muscle synergy analysis processing unit 93 calculates the weight of the palatoglossus muscle within the first muscle synergy Syn1. As shown in Figure 17, it multiplies the time-dependent activity level of the first muscle synergy Syn1 by the weight of the palatoglossus muscle within the first muscle synergy Syn1 to obtain first data representing the muscle activity of the palatoglossus muscle from the start to the end of swallowing in terms of the activity level of the first muscle synergy Syn1.

[0145] Similarly, the muscle synergy analysis processing unit 93 calculates the weight of the palatoglossus muscle within the second muscle synergy Syn2, and, as shown in Figure 17, multiplies the time-dependent activity level of the second muscle synergy Syn2 by the weight of the palatoglossus muscle within the second muscle synergy Syn2 to obtain second data representing the muscle activity of the palatoglossus muscle from the start to the end of swallowing as the activity level of the second muscle synergy Syn2.

[0146] In this way, the muscle synergy analysis processing unit 93 obtains muscle activity rates from the 1st to the 9th, representing the muscle activity rate of the palatoglossus muscle from the start to the end of swallowing for each muscle synergy Syn1 to Syn9, calculated from the weights of each muscle synergy Syn1 to Syn9. The obtained muscle activity rates from the 1st to the 9th represent the muscle activity rates of the palatoglossus muscle that reflect the results of the muscle synergy analysis.

[0147] The muscle synergy analysis processing unit 93 calculates the first to ninth muscle activity rates for all 40 types of muscle, and calculates a muscle activity rate for each muscle type that aggregates the results of muscle synergies Syn1 to Syn9. In this way, the muscle synergy analysis processing unit 93 obtains a muscle activity rate for all 40 types of muscle that reflects the analysis results of muscle synergies Syn1 to Syn9.

[0148] The motion analysis unit 50 sets the muscle activity rate for each of the 40 types of muscles to a dynamic three-dimensional head and neck model 10c that reflects the analysis results of muscle synergy that changes over time from the start to the end of swallowing, and uses this dynamic three-dimensional head and neck model 10c to simulate the movement of the head and neck organs during swallowing using a particle method in three-dimensional images.

[0149] If the number of muscle synergies is appropriate, the simulation results will show that the desired swallowing motion is achieved in the dynamic three-dimensional head and neck model 10c. On the other hand, if the number of muscle synergies is inappropriate, such as too few, the simulation results will show that an undesirable swallowing motion is achieved in the dynamic three-dimensional head and neck model 10c, for example, that the epiglottis does not invert and does not cover the larynx.

[0150] As a verification test, using the swallowing simulation device 1 according to the above embodiment, the number of muscle synergies was set to 5, and muscle activity rates reflecting the results of the 5 muscle synergies Syn1 to Syn5 were obtained for all 40 types of muscles according to "(5) <Analysis method of muscle synergies using a dynamic three-dimensional head and neck model>" above. When the movement of the head and neck organs during swallowing was simulated in three dimensions using a particle method with a dynamic three-dimensional head and neck model 10c that had been set with muscle activity rates reflecting the analysis results of the muscle synergies, the epiglottis did not invert and did not cover the trachea, and the desired action could not be reproduced. Therefore, it was inferred that the number of muscle synergies needs to be changed in order to reproduce the desired swallowing action with the dynamic three-dimensional head and neck model 10c.

[0151] (6) <Simulation judgment process> The simulation determination unit 94 uses a dynamic three-dimensional head and neck model 10c, in which the muscle activity rates of 40 types of muscles have been changed to muscle activity rates that reflect the results of muscle synergy analysis, to perform a swallowing simulation using the motion analysis unit 50. Based on the results, it determines whether or not the number of muscle synergies needs to be modified. Specifically, the simulation determination unit 94 stores in advance a machine learning model that can determine whether or not the swallowing simulation of the dynamic three-dimensional head and neck model 10c is an optimal swallowing motion. The simulation determination unit 94 uses the machine learning model to determine whether or not the swallowing simulation of the dynamic three-dimensional head and neck model 10c is an optimal swallowing motion based on the set number of muscle synergies.

[0152] If the simulation determination unit 94 determines that the optimal swallowing motion has not been reproduced, it can prompt the muscle synergy analysis processing unit 93 to modify the number of muscle synergies set, and then prompt it to perform non-negative matrix factorization on the muscle activity matrix X again based on the modified number of muscle synergies.

[0153] In this way, the swallowing simulation device 1 can perform swallowing simulations while analyzing the number of muscle synergies in which muscle types perform coordinated movements, a spatial pattern showing the degree of muscle activity of each constituent muscle type in each muscle synergy during swallowing, and a temporal pattern showing the change in the degree of activity of each muscle synergy over time.

[0154] (7) <Mechanism of Action and Effects> In the above configuration, the swallowing simulation device 1 stores in the memory unit 83 a dynamic three-dimensional head and neck model 10c in which multiple head and neck organs are modeled by multiple muscle particles that specify the direction of muscle fibers in a three-dimensional image for each muscle type, and contractile stress based on the direction of muscle fibers during swallowing is applied to the muscle particles by the muscle activity rate. The motion analysis unit 50 uses the dynamic three-dimensional head and neck model 10c to simulate the movement of head and neck organs during swallowing in a three-dimensional image based on the particle method.

[0155] The muscle synergy analysis processing unit 93 obtains the muscle activity rate of each muscle type, which changes in accordance with the time-dependent changes in the movement of the head and neck organs, based on the simulation results of the swallowing movement of the head and neck organs during swallowing performed by the motion analysis unit 50. Based on the muscle activity rate of each muscle type obtained from the dynamic three-dimensional head and neck model 10c, the muscle synergy analysis processing unit 93 can analyze the number of muscle synergies determined based on the muscle synergy hypothesis that multiple muscle types are controlled by coordinated movement, a spatial pattern showing the degree of muscle activity of the constituent muscle types for each muscle synergy during swallowing, and a time pattern showing the degree of activity for each muscle synergy as it changes over time.

[0156] Thus, the swallowing simulation device 1 uses a dynamic three-dimensional head and neck model 10c in which muscle activity rates are set for each muscle type and these muscle activity rates can be changed. The device simulates the movement of head and neck organs during swallowing based on the particle method, and can perform muscle synergy analysis regarding the movement of head and neck organs during swallowing based on the analysis results of the muscle activity rates during the simulation. Therefore, the swallowing simulation device 1 can perform simulations including muscle synergy regarding the movement of head and neck organs during swallowing.

[0157] Furthermore, the muscle synergy analysis processing unit 93 generates a muscle activity matrix X representing the time-dependent changes in muscle activity rates for each muscle type during swallowing, based on the simulation results of head and neck organ movement during swallowing performed by the motion analysis unit 50. The muscle synergy analysis processing unit 93 then performs non-negative matrix factorization on the muscle activity matrix X using the set number of muscle synergies, converting the muscle activity matrix X into a spatial pattern matrix W and a temporal pattern matrix H, and calculates these spatial pattern matrix W and temporal pattern matrix H as analysis results of the spatial and temporal patterns. In this way, the muscle synergy analysis processing unit 93 can easily obtain the spatial pattern matrix W and temporal pattern matrix H through calculations using the muscle activity matrix X, and can analyze the spatial and temporal patterns of muscle synergies during swallowing.

[0158] Furthermore, the muscle synergy analysis processing unit 93 divides the muscle activity matrix X into time domains corresponding to those time domains along the time series, and generates multiple time-divided time pattern matrices H n This generates multiple time-resolved time pattern matrices H obtained. n From this, a time domain of the time-division time pattern matrix H4, which includes the activity levels of multiple muscle synergies Syn1 to Syn9, is identified, and the muscle activity rates of each muscle type between sampling times are interpolated within the identified time domain. In this way, by interpolating the muscle activity rates of each muscle type during swallowing simulation, the movement of the dynamic three-dimensional head and neck model 10c can be made smoother.

[0159] Furthermore, the muscle synergy analysis processing unit 93 generates a muscle activity rate for each muscle type that reflects the analysis results of muscle synergies, which are obtained by aggregating the activity levels obtained for each muscle synergy Syn1 to Syn9, and sets the muscle activity rate for each muscle type that reflects the analysis results of muscle synergies obtained for each muscle type. As a result, the motion analysis unit 50 simulates the movement of the head and neck organs during swallowing using a three-dimensional image based on the particle method, with the dynamic three-dimensional head and neck model 10c, which has been set for each muscle type by the muscle synergy analysis processing unit 93 that reflects the analysis results of muscle synergies obtained for each muscle type. This makes it possible to analyze the number of muscle synergies, etc., from the simulation results.

[0160] (8) <Other Embodiments> In the embodiments described above, in addition to forming a dynamic three-dimensional head and neck model 10c that reproduces the head and neck organs of a person with dysphagia or a healthy person using three-dimensional images, a dynamic three-dimensional head and neck model that reproduces the head and neck organs of, for example, an infant or an elderly person may also be formed using three-dimensional images. Furthermore, in the embodiments described above, in addition to the MPS method adopted in this embodiment, other particle methods such as the SPH (Smoothed Particle Hydrodynamics) method or the Discrete Element Method may also be applied.

[0161] Furthermore, in the embodiments described above, the particle method was used as a numerical analysis method to represent and simulate the movement (action) of head and neck organs in a dynamic three-dimensional head and neck model and the behavior of a simulated orally ingested product in a three-dimensional image. However, the present invention is not limited to this, and a grid method may be used to treat the head and neck organs of the dynamic three-dimensional head and neck model and the simulated orally ingested product as a mesh. In this case, instead of muscle particles, which are particulate myoplastic bodies, a mesh-like myoplastic body (hereinafter simply referred to as a muscle mesh) is used to form the dynamic three-dimensional head and neck model 10c.

[0162] Specifically, the dynamic three-dimensional head and neck model 10c is used, in which multiple head and neck organs are modeled by multiple muscle meshes, each with its muscle fiber direction identified within a three-dimensional image for each muscle type, and contractile stress based on the muscle fiber direction during swallowing is applied to the muscle mesh by the muscle activity rate. The motion analysis unit 50 may then use this dynamic three-dimensional head and neck model (dynamic three-dimensional head and neck mesh model) to perform a swallowing simulation using the grid method, analyzing the movement (action) of the neck organs during swallowing and the behavior of orally ingested food during swallowing, and displaying these results in a three-dimensional image. The muscle activity rate here refers to the temporal change in the contractile stress of the muscle mesh during swallowing, which is set for each muscle type with respect to the muscle mesh to which contractile stress based on the muscle fiber direction is applied during swallowing.

[0163] For the grid method, it is desirable to apply the finite element method, finite volume method, finite difference method, or boundary element method. It is also possible to apply interface capture methods such as the Volume of Fluid (VOF) method.

[0164] Here, the dynamic three-dimensional head and neck model 10c used in the swallowing simulation device 1 according to this embodiment can be fabricated in accordance with the description of the particle method. That is, after fabricating a static three-dimensional head and neck model in the same manner as (4) <Fabrication of the dynamic three-dimensional head and neck model>, a grid is fabricated within each region for each head and neck organ, with the surface of the head and neck organs as the boundary. The grid can be fabricated according to known methods disclosed, for example, in Japanese Patent Application Publication No. 4-181481 or Pal et al., Proc. R. Soc. Lond. B., 271, 2587 (2004).

[0165] The grid in this embodiment is a three-dimensional cubic grid or rectangular grid (simply referred to as a grid in this embodiment) that has a three-dimensional shape in the three-dimensional image. For example, when creating a life-size model of the head and neck of an infant or an average-sized adult male in a three-dimensional image, it is desirable to set the size of each side of the grid to approximately 0.1 mm to 3.0 mm, and more preferably to set the size of each side to approximately 0.6 mm to 1.5 mm.

[0166] Furthermore, in the static three-dimensional head and neck model created within the three-dimensional image, it is desirable that at least two grids be formed in the direction of the epiglottis thickness (for example, approximately 3.0 mm in adults and approximately 1.5 mm in infants). If the grid is too small, the computational processing burden on personal computer 2 becomes too great, which is undesirable. On the other hand, if the grid is too large, it is not possible to reproduce the fine movements of the head and neck organs, so it is desirable that the size of the grid be within the above range.

[0167] In this embodiment, we describe the case where a three-dimensional cube or rectangular prism is used as the grid for forming the head and neck organs. However, the present invention is not limited to this, and the head and neck organs may be formed using grids of various other shapes, such as triangular grids or polygonal grids.

[0168] Thus, in a swallowing simulation device that uses a dynamic three-dimensional head and neck model modeled by replacing the muscle particles of the above-described embodiment with a muscle mesh, the muscle synergy analysis processing unit 93 can analyze, similar to the above-described embodiment, the number of muscle synergies in which multiple muscle types are controlled by coordinated movement, a spatial pattern showing the degree of muscle activity of the constituent muscle types for each muscle synergy during swallowing, and a temporal pattern showing the degree of activity for each muscle synergy as it changes over time, based on the muscle activity rate of the muscle mesh for each muscle type obtained from the dynamic three-dimensional head and neck model.

[0169] Similarly, in a swallowing simulation device, a dynamic three-dimensional head and neck model is used in which muscle activity rates are set for each muscle type in the muscle mesh and these muscle activity rates can be changed. This allows for the simulation of the movement of head and neck organs during swallowing based on the grid method, and the analysis of muscle synergies regarding the movement of head and neck organs during swallowing can be performed based on the analysis results of the muscle activity rates during the simulation. Therefore, a swallowing simulation device can perform simulations that include muscle synergies regarding the movement of head and neck organs during swallowing.

[0170] In the embodiments described above, a swallowing simulation device 1 and a swallowing simulation method were described as simulation devices and simulation methods that simulate the movement of head and neck organs during swallowing of a simulated orally ingested food using three-dimensional images based on a particle method or a grid method. However, the present invention is not limited to these, and as described above, it may also be applied to a chewing simulation device and chewing simulation method that simulates the movement of head and neck organs during chewing of a simulated orally ingested food using three-dimensional images based on a particle method or a grid method. In such a chewing simulation device, as in the embodiments described above, simulations including muscle synergies can be performed regarding the movement of head and neck organs during chewing. [Explanation of Symbols]

[0171] 1. Swallowing simulation device (simulation device) 10c Dynamic three-dimensional head and neck model 50 Motion Analysis Department 83 Storage section 93 Muscle Synergy Analysis Processing Unit

Claims

1. A storage unit stores a dynamic three-dimensional head and neck model in which multiple head and neck organs are modeled by multiple particulate or mesh-like myoforms, each with identified muscle fiber direction in a three-dimensional image for each muscle type, and in which contractile stress based on the muscle fiber direction during swallowing or chewing is applied to the myoforms according to the muscle activity rate. A motion analysis unit that uses the dynamic three-dimensional head and neck model in which the muscle activity rate can be changed to simulate the movement of the head and neck organs during swallowing or chewing using the three-dimensional image based on the particle method or the grid method, A muscle synergy analysis processing unit analyzes, from the muscle activity rate of each muscle type that changes in accordance with the time course of movement of the head and neck organs, a spatial pattern showing the number of muscle synergies in which the muscle types perform coordinated movements, the degree of muscle activity of each constituent muscle type during swallowing or chewing, and a temporal pattern showing the degree of activity of each muscle synergy as it changes over time. Equipped with, The muscle synergy analysis processing unit is as follows: For each muscle synergy, the weight of each muscle type within the muscle synergy is calculated, and the muscle activity rate of each muscle type for each muscle synergy is calculated by multiplying the time-dependent activity level of each muscle synergy by the weight, and for each muscle type, an aggregated muscle activity rate is calculated by aggregating the muscle activity rates calculated for each muscle synergy. The motion analysis unit described above is A simulation device that sets the aggregated muscle activity rate calculated by the muscle synergy analysis processing unit into the dynamic three-dimensional head and neck model, and simulates the movement of the head and neck organs during swallowing or chewing using the particle method or the grid method in the three-dimensional image.

2. The muscle synergy analysis processing unit is as follows: From the simulation results of the movement of the head and neck organs during swallowing or chewing performed by the aforementioned motion analysis unit, a muscle activity matrix X is generated that represents the change over time in the muscle activity rate for each muscle type during swallowing or chewing. The simulation apparatus according to claim 1, wherein the number of muscle synergies set is used to convert the muscle activity matrix X into a spatial pattern matrix W representing the degree of muscle activity for each of the constituent muscle types for each muscle synergy, and a temporal pattern matrix H representing the change in the degree of activity over time for each muscle synergy, and the spatial pattern matrix W and the temporal pattern matrix H are calculated as the results of the analysis of the spatial pattern and the temporal pattern.

3. The muscle synergy analysis processing unit is as follows: Based on the characteristic movements of the head and neck organs during swallowing or chewing, the muscle activity matrix X is divided into multiple time-compartmental muscle activity matrices X in a time-series manner. ｎ Generates the multiple time-divided muscle activity matrices X ｎ The simulation apparatus according to claim 2, which analyzes the spatial pattern matrix W and the time pattern matrix H for each.

4. The muscle synergy analysis processing unit is as follows: The aforementioned time pattern matrix H is divided into multiple time domains along the time series to form multiple time-division time pattern matrices H ｎ The spatial pattern matrix W is generated, and the plurality of time-division time pattern matrices H ｎ The simulation apparatus according to claim 2, having a configuration corresponding to the above.

5. The muscle synergy analysis processing unit is as follows: The aforementioned time-division time pattern matrix H ｎ The simulation apparatus according to claim 4, which interpolates the muscle activity rate of the muscle type in the time domain.

6. The simulation device according to claim 1, wherein the number of muscle synergies is analyzed based on the results of simulating the movement of the head and neck organs during swallowing or mastication using the dynamic three-dimensional head and neck model in which the aggregated muscle activity rate is set, based on the particle method or the grid method in the three-dimensional image.

7. The simulation device according to claim 1, further comprising a simulation determination unit that determines whether or not analysis by the muscle synergy analysis processing unit is necessary based on the simulation results of the movement of the head and neck organs during swallowing or chewing, performed by the motion analysis unit using the dynamic three-dimensional head and neck model in which the aggregated muscle activity rate has been set.

8. A storage step of storing in a memory unit a dynamic three-dimensional head and neck model in which multiple head and neck organs are modeled by multiple particulate or mesh-like myoforms, each with identified muscle fiber direction in a three-dimensional image for each muscle type, and contractile stress based on the muscle fiber direction during swallowing or chewing is applied to the myoforms according to the muscle activity rate, A motion analysis step in which the motion analysis unit uses the dynamic three-dimensional head and neck model, in which the muscle activity rate can be changed, to simulate the movement of the head and neck organs during swallowing or chewing using the three-dimensional image based on the particle method or the grid method, A muscle synergy analysis processing step involves analyzing, using a muscle synergy analysis processing unit, the number of muscle synergies in which the muscle types perform coordinated movements, a spatial pattern showing the degree of muscle activity of each constituent muscle type during swallowing or chewing, and a temporal pattern showing the degree of activity of each muscle synergy as it changes over time, based on the muscle activity rate of each muscle type that changes in accordance with the changes in movement of the head and neck organs over time. The muscle synergy analysis processing unit calculates the weight of each muscle type within each muscle synergy, multiplies the time-dependent activity level of each muscle synergy by the weight to calculate the muscle activity rate of each muscle type for each muscle synergy, and calculates an aggregated muscle activity rate for each muscle type by aggregating the muscle activity rates calculated for each muscle synergy. The motion analysis unit sets the aggregated muscle activity rate calculated by the muscle synergy analysis processing unit into the dynamic three-dimensional head and neck model, and simulates the movement of the head and neck organs during swallowing or chewing using the three-dimensional image based on the particle method or the grid method. A simulation method that includes this.

Images