A method and a system for real-time movement and performance analyses for a human

Abstract

The present invention relates to a method and a system for real-time movement and performance analysis for a human using an external computer system and at least one wireless mobile monitoring device, where the monitoring device measures articular angle and / or muscular activity of an object part while the human moves the object part in accordance with pre-defined moving instructions. By processing this data, a real-time movement performance analysis indicator may be calculated indicating if the muscular activity for the muscular object is as expected or not.

Description

[0001] KIS24001PWO Rey_Baldur

[0002] Kiso ehf.

[0003] 1

[0004] A METHOD AND A SYSTEM FOR REAL-TIME MOVEMENT AND PERFORMANCE ANALYSES FOR A HUMAN

[0005] FIELD OF THE INVENTION

[0006] The present invention relates to a system and a method for real-time movement and performance analysis for a human.

[0007] BACKGROUND OF THE INVENTION

[0008] Musculoskeletal problems are among the largest health issues in the world, and the number of people requiring solutions is growing faster than healthcare providers can manage.

[0009] Musculoskeletal problems are one of the leading causes of pain and reduced mobility globally, making them a particular focus for the World Health Organization, WHO. The result is that people live with these ailments longer than necessary, leading to associated negative social impacts and costs. Up to one-third of the world's population suffers from musculoskeletal problems, with the percentage being higher in developed countries.

[0010] Physical therapists and trainers help their clients to find the best exercises. Their typical service is based on experience, knowledge from books, and / or courses rather than direct measurements. The field is characterized by subjective assessments of which exercises yield the best results. Measurements have shown that the exercises clients are given do not always produce the expected results.

[0011] Physical therapists are likely the professional group most knowledgeable about human movement. They have a solid foundational knowledge but often lack the time or resources to utilize results from scientific research and therefore don’t have good and standardized guidelines on how to use existing equipment. It is estimated that about 70% of their work involves influencing muscles. Measurements typically expected in a standard physical therapy setting are often limited to questionnaires, goniometers, or a single video recording of KIS24001PWO Rey_Baldur

[0012] Kiso ehf.

[0013] 2 movement. Thus, a typical physical therapy clinic rarely engages in objective measurements of movement ability.

[0014] Often, humans suffer from lack of muscle control and muscle imbalance which manifests in such a way that one muscle takes over the function of another or is inactive or hyperactive. This issue often goes undetected by traditional methods, leading to weeks or even months of treatment on a specific muscle without achieving the expected results.

[0015] Efforts to overcome the aforementioned challenges include, for instance, US 20200401224 discloses a rehabilitation system that uses wearable sensors, and in some embodiments video cameras, to monitor human’s movements and muscle activity. The system depends on the human’s performing a movement, such as lifting or rotating a limb, and from this action the processor estimates pose, sensor orientation, and joint angles while also capturing EMG signals. This data is then used by a therapy application to provide real-time biofeedback, visual instructions, or interactive games that guide rehabilitation. In extended versions, video data is integrated with sensor data to refine pose trajectories under biomechanical constraints, leading to more accurate joint angle calculations. At its core, the system transforms the human’s own movements into measurable data that drives personalized therapy.

[0016] The problem with this approach is that it is entirely dependent on the initial movements made by humans. If those movements are random, poorly executed, or are not relevant for the intended exercise, the system cannot distinguish them from purposeful training actions, which creates uncertainty as to how the system identifies a correct starting point for an effective therapy program. Thus, this approach relies on what may be referred to as reversed motion detection, wherein an intended motion is inferred from the user’s executed movement. In such a method, the approach attempts to recognize purpose solely based on motion data captured after the action has occurred. Consequently, if an action is poorly executed or deviates from the expected exercise pattern, the approach may fail to correctly identify the intended motion.

[0017] A further problem is that the processor does not directly measure joint pose but instead relies on estimated values for limb orientation and angle. These estimates are in turn anchored to the KIS24001PWO Rey_Baldur

[0018] Kiso ehf.

[0019] 3 quality of the human’s initial movement. As a result, two sources of inaccuracy, sensor-based pose estimation and reliance on human-provided initial movement, are compounded, leading to significant error and unreliability in the analysis. This limits the precision of feedback, the ability to track progress, and ultimately the therapeutic effectiveness of the system.

[0020] A related limitation is seen in many consumer wearables (e.g., smart watches or smart rings) that rely on a single or narrowly scoped physiological or kinematic signal (for example, a cuffless blood-pressure proxy, a photoplethysmography / heart-rate channel, or a single inertial stream) and then infer multiple secondary outcomes such as sleep quality, movement type, calorie expenditure, or stress. When downstream metrics are stacked on top of one primary estimate, any initial error propagates and may be magnified through successive inference steps. This estimation-on-estimation architecture yields outputs that can be highly sensitive to noise, motion artifacts, or human behavior, and thus materially reduces reliability for individualized guidance, mirroring the compounded-error issue described above for US 20200401224.

[0021] SUMMARY OF THE INVENTION

[0022] It is an object of the invention to address the identified challenges by providing a clinically fitting system and method capable of accurately assessing, documenting and intervening musculoskeletal issues, where objective measurements and insights into e.g. muscle control and imbalances, which are often overlooked by traditional approaches, may be solved.

[0023] In general, the invention preferably seeks to mitigate, alleviate or eliminate one or more of the above-mentioned disadvantages of prior art singly or in any combination. It may be seen as an object of embodiments of the present invention to provide precise musculoskeletal related data as input in enhancing the effectiveness of treatments and to ensure that humans receive more targeted and scientifically backed interventions, leading to better outcomes and reduced treatment times.

[0024] In a first aspect of the invention, a method is provided of real-time movement and performance analysis for a human, where the method comprises: KIS24001PWO Rey_Baldur

[0025] Kiso ehf.

[0026] 4

[0027] • providing a pre-stored moving instructions characteristic for an object part of the human,

[0028] • presenting the moving instructions to the human,

[0029] • providing at least one wireless mobile monitoring device configured to be placed on the object part of the human, comprising: o a communication module, o an orientation sensor, and / or o a muscular activity sensor,

[0030] • measuring, by the orientation and / or the muscular activity sensor, the articular angle and / or muscular activity of the object part while the human moves the object part in accordance with the moving instructions,

[0031] • communicating by the communication module the measured articular angle data and / or the muscular activity data to a computer system where the data is processed, where the processing includes,

[0032] • comparing the received articular angle position data and / or the muscular activity data with reference data and determining a real-time movement and performance analysis indicator for the object part.

[0033] The step of processing the received articular angle position data and / or the muscular activity data includes processing both the articular angle position data and the muscular activity data simultaneously. The processing may in an alternative embodiment include using the articular angle position data to register, time-align, and / or normalize the muscular activity data with corresponding reference data for each measured articular angle position. As used herein, “simultaneously” is to be understood as meaning that the processing of the articular angle position data and the muscular activity data occurs in parallel or within overlapping time intervals, such that both data types are analyzed together in real time.

[0034] Preferably, for each measured articular angle, the method further comprises forms an angle- resolved paired dataset combining the muscular activity data with the corresponding reference data and computes, in real time, a movement and performance analysis indicator for the object KIS24001PWO Rey_Baldur

[0035] Kiso ehf.

[0036] 5 part. The indicator may include one or more muscle-activation or muscle-imbalance metrics, such as an agonist-antagonist activation ratio, a side-to-side symmetry index, a co-contraction index, a deviation from an angle-specific normative envelope, or an imbalance severity score.

[0037] The method may be applied to either an injured object part, to assist in rehabilitation and restoration of balanced activation, or to a healthy object part, to enhance strength, endurance, or coordination through optimized muscle activation patterns, for example in training programs designed for professional athletes.

[0038] The method may further comprise time-aligning the muscular activity data with the reference data, where the time-alignment may be based on articular angle derived from an inertial measurement unit (IM U), such as a MARG sensor, and may further use accelerometer-based impact events (for example, heel-strike or landing) as temporal reference points.

[0039] By combining and aligning the articular angle and muscular activity data, the method produces angle-resolved datasets that provides a more accurate and physiologically meaningful assessment of muscle function. This enables early detection of muscle imbalance and maladaptive movement patterns, supporting safer and more targeted rehabilitation.

[0040] This may be performed by means of utilizing the articular angle data to align the muscular activity data with corresponding muscular activity reference data for each acquired articular angle position measured by the orientation sensor. That means as an example that for a given articular angle data a and muscular activity data MT pair (a.M^ a third dimension data point is added, namely M2resulting in (a.M-,, M2)i where M2is the muscular activity reference data for the articular angle data a., where tens, hundreds or thousands of such datapoints (a.M-,, M2)j are then be presented in real-time via any type of output means such as a display.

[0041] Accordingly, the present method addresses and overcomes the limitations of prior-art rehabilitation approaches that depend on user-initiated movements and indirectly estimated joint poses. By anchoring analysis to directly measured data from orientation and muscular activity KIS24001PWO Rey_Baldur

[0042] Kiso ehf.

[0043] 6 sensors, rather than to estimated joint angles derived from initial user actions, the method eliminates compounded estimation errors caused by poorly executed or irrelevant movements.

[0044] Moreover, the method further includes guided reference postures and automatic start-point detection to ensure that data acquisition begins from a valid and biomechanically meaningful baseline, thereby reducing setup time and enabling initialization-free operation.

[0045] Direct measurement of articular angle and muscle activity allows for accurate and drift-resistant joint pose estimation, improving the fidelity of movement analysis even under motion artifacts. Through multimodal sensor fusion with quality gating and outlier rejection, the method ensures that noisy or low-confidence data are excluded from further processing, maintaining robustness under user variability. Thus, the method provides highly reliable, repeatable, and individualized feedback and objective tracking of user progress.

[0046] Further, by deriving higher-level performance metrics from validated primary measurements, rather than stacking inferences upon estimations, the method suppresses propagation of error and enhances analytical stability and therefore delivers data-driven, initialization-free calibration that significantly improves the accuracy, reliability, and efficiency of real-time movement and performance analysis, thereby providing a more effective foundation for rehabilitation and training applications.

[0047] By e.g. simultaneously processing articular-angle and muscular-activity signals and aligning EMG to angle-specific references in real time, the method yields angle-resolved, tri-variate datasets (a, Mi, M2) that expose phase-specific neuromuscular deficits invisible to angleagnostic approaches. This enables earlier and more reliable detection of poor motor control and muscle imbalance at the exact joint angles where they arise, supports lower-latency and more precise biofeedback during the instructed movement, and improves the accuracy and stability of movement / performance indicators versus conventional threshold or peak-based methods. Presenting these indicators during task execution personalizes exercise dosing (load, tempo, and range constraints) to the human’s current state, increases adherence and training efficiency, and facilitates remote supervision via the wireless mobile device. KIS24001PWO Rey_Baldur

[0048] Kiso ehf.

[0049] 7

[0050] Moreover, the use of on-body, wirelessly networked sensors reduces cabling and setup time, permits deployment outside the clinic, and maintains synchronized streams for robust analysis under motion. Overall, the method delivers objective, angle-resolved metrics that accelerate rehabilitation decision-making, reduce trial-and-error in exercise prescription, and help prevent persistent pain, restricted motion, and long-term joint wear.

[0051] The term “reference data” may according to the present invention be understood as data representing a baseline or normative condition of a physiological structure, such as a healthy muscle or tissue. Such reference data may include, for example, electromyographic signals recorded from healthy muscle under standardized contraction and relaxation, forcedisplacement curves obtained from normal biomechanical response, ultrasound or MRI imaging data characterizing muscle cross-sectional area, fiber orientation or elasticity, electrical impedance or conductivity values of tissue, or metabolic profiles such as oxygen consumption rates and lactate levels under controlled exercise loads. Reference data may further comprise kinematic information such as joint angle trajectories, velocity and acceleration patterns reflecting normal muscular activation, as well as historical datasets of clinically verified healthy individuals used for comparative analysis or for training machine-learning models. In certain embodiments, the reference data may also be generated by computational simulations, for example finite element models of muscle behavior. The availability of such reference data provides a technical advantage by allowing the system to detect deviations from normal muscular function and thereby enable robust diagnostics, personalized feedback and reliable operation even in environments where communication or sensor signals may be unstable. The reference data may in an alternative embodiment origin from a pre-stored reference database where, as addressed above, the data indicate how healthy muscular activity data should be, or it may origin from a corresponding healthy object part of the human.

[0052] The term “human” may as an example understood as a patient, i.e. someone with some kind of injuries, or someone, such as an athlete, where the aim is to improve the athlete’s condition. KIS24001PWO Rey_Baldur

[0053] Kiso ehf.

[0054] 8

[0055] For example, a human with a shoulder injury may be instructed to perform controlled arm elevations or rotations, ensuring that the captured data are directly relevant to the therapeutic objective. Because the instructions are predefined, they can be directly associated with corresponding reference data, which both simplifies the comparison process and enhances the precision of the movement analysis while reducing computational complexity.

[0056] The term object part may according to the present invention refer to a biological or mechanical segment whose state can be measured or controlled, and may be muscular or non-muscular. Muscular object parts include an individual muscle, a muscle head, a muscle group, or a functional compartment, for example biceps brachii, quadriceps femoris (including VM and VL), the hamstrings, the gastrocnemius-soleus complex, the rotator-cuff muscles, or the forearm flexors and extensors. Non-muscular object parts include body segments or structures such as joints or joint complexes (e.g., knee, shoulder, or spinal segments), bones (e.g., femur or tibia), tendons and ligaments (e.g., the patellar tendon, Achilles tendon, or ACL), fascia or defined skin regions, as well as mechanical elements such as prosthetic and orthotic components, exoskeleton links and joint modules, actuator housings, hinges, or sensorized brace elements. By way of illustration, in a rehabilitation context the system may monitor quadriceps activation (muscular) while tracking knee kinematics (non-muscular); in sports assessment it may relate hamstring activity to pelvic and lumbar motion; in wearable assistance it may control an exoskeleton knee module based on gastrocnemius EMG; and in prosthetics it may infer human intent from residual-limb muscle signals to drive a prosthetic ankle.

[0057] In the context of the present invention, the term “real-time movement” denotes that motion data from the human or human body part is captured, transmitted, processed, and analyzed substantially simultaneously with the physical movement itself.

[0058] The expression “real-time” may refers to the actual time during which an event or process takes place, and in this context may be understood that the data acquisition, computation, and feedback are performed within time intervals short enough to reflect and respond to the ongoing movement as it occurs. Real-time operation does not require absolute simultaneity but implies that latency between the physical movement and the corresponding data processing or output is KIS24001PWO Rey_Baldur

[0059] Kiso ehf.

[0060] 9 sufficiently small, typically within milliseconds, microseconds, or even nanoseconds depending on the hardware and communication constraints, so that the user perceives the feedback or analysis as instantaneous or near-instantaneous.

[0061] Moreover, in practical terms, “real-time” may encompass hardware and software systems designed to operate under a specified time constraint, ensuring that data capture, processing, and response occur deterministically within a bounded delay. This enables continuous or near- continuous movement monitoring and immediate feedback generation during exercise or performance.

[0062] In an embodiment, the method further comprises, initially, in response to selecting an object part or object parts on the human to be analyzed, providing information indicating where to place the orientation and / or the muscular activity sensor on the human. These instructions may as an example be presented in a similar way to the movement instructions for the human. For example, when the knee is selected, the present invention may instruct placement of the orientation sensor on the anterolateral tibial shaft aligned with the tibial axis and, if EMG is used, placement of differential surface electrodes over vastus medialis obliquus at a specified distance and angle from the patella and a reference channel over vastus lateralis.

[0063] When the shoulder is selected, the method may comprise step of instructing or proposing placement of the orientation sensor on the distal humerus aligned with the elbow-shoulder axis and instructing or proposing placement of EMG electrodes over the anterior or middle deltoid with a defined inter-electrode spacing and reference location to minimize cross-talk, i.e. unintended signal interference from neighboring muscles.

[0064] The guidance can further adapt to laterality (left / right), limb length, and body-mass index by scaling distances (e.g., as a percentage of segment length) and by offering alternative placements in the presence of scars, edema, or skin conditions. KIS24001PWO Rey_Baldur

[0065] Kiso ehf.

[0066] 10

[0067] To ensure correct placement before measurements begin, the method may in an embodiment include a verification step in which the computer system evaluates short calibration maneuvers (e.g., two controlled flexion-extension cycles or a brief isometric contraction) and computes signal-quality indicators such as vector magnitude stability for the orientation sensor, electrode impedance and common-mode rejection for EMG, and cross-channel correlation to detect cross-talk. If thresholds are not met, the present method may further comprise a step of issuing corrective prompts (e.g., “rotate sensor by 10°,” “increase electrode spacing to 20-25 mm,” or “re-prep skin to reduce impedance”) and repeat verification until a placement-confidence score is achieved. The method may record finalized placement metadata, including anatomical descriptions, photographs or sketches, electrode model and spacing, and calibration metrics, so that subsequent sessions can replicate the setup within predefined tolerances, supporting longitudinal comparability.

[0068] Technical advantages include improved data fidelity and repeatability through standardized, landmark-based placement with automated quality checks, which reduces inter-operator and inter-session variability and thereby increases the accuracy of downstream movement / performance indicators. By validating orientation alignment and EMG selectivity at the outset, the method lowers the incidence of drift, cross-talk, and motion artifacts, enabling more stable angle estimates and more specific muscle activation profiles. Adaptive, anatomy- scaled instructions shorten setup time and reduce human error, which improves throughput in clinical and field settings while maintaining safety (e.g., by flagging contraindicated sites). Persisting placement metadata enables reproducible, angle-resolved comparisons across days or therapists, strengthens the evidentiary value of detected improvements or regressions, and supports remote workflows where the system can confirm adequate self-placement by the human before analysis begins. Collectively, these features yield earlier, more reliable detection of muscle imbalance and impaired motor control at the exact joint angles where they arise, facilitate lower-latency and more precise biofeedback during instructed movement, and reduce trial-and-error in exercise prescription, advantages that conventional, ad-hoc placement workflows do not provide.

[0069] In an embodiment, multiple wireless mobile monitoring devices are utilized to be placed on an injured object part (first object part) and corresponding healthy object part (second object part) KIS24001PWO Rey_Baldur

[0070] Kiso ehf.

[0071] 11 on the human, which may be the same object part, where the reference data comprises the data measured on the corresponding healthy object. This may include a scenario where both object parts are moved simultaneously, or not simultaneously, e.g. where the healthy object part is first moved, and then the injured object part, or vice versa. In some scenarios the first and the second object parts are not the same object parts.

[0072] Using the healthy side as an angle-resolved reference reduces dependence on population norms and automatically normalizes for individual anatomy, limb dominance, sensor placement, and skin-electrode characteristics. This yields real-time bilateral symmetry indices, covering magnitude, timing, and co-contraction, across matched angle bins, improving sensitivity to side- to-side deficits and timing offsets that angle-agnostic methods miss. Sequential acquisition still enables template-based guidance. The method can capture kinematic / EMG envelopes from the healthy side and replay them as targets or reference for the injured side, supporting mirroring biofeedback and closed-loop assistance.

[0073] Synchronized (or time-aligned) dual-sensor acquisition also suppresses common-mode noise, drift, and environmental artifacts through differential metrics, increasing stability of joint-angle estimates and specificity of muscle-activation profiles. Setup time is reduced because a short healthy-side recording bootstraps, i.e. automatically derives individualized thresholds and progression criteria from the healthy-side data, thereby enabling safer workload titration and automatic detection of compensations. When the reference part differs (e.g., proximal-distal pairing), learned feature mappings still provide functional benchmarks to drive intent classification and disturbance rejection. Collectively, these capabilities accelerate detection of clinically meaningful asymmetries, improve low-latency feedback for corrective training, and enhance the accuracy and robustness of movement / performance indicators in real-world rehabilitation.

[0074] An example of a muscular object part is Biceps Brachii, Pectoralis Major, Rectus Abdominis, Quadriceps Femoris, Deltoid, Gastrocnemius, Gluteus Maximus, Latissimus Dorsi, just to mention few muscular object parts. KIS24001PWO Rey_Baldur

[0075] Kiso ehf.

[0076] 12

[0077] In an alternative embodiment, the orientation sensor may be placed on a non-muscular object part while measuring the articular angle data of the object part.

[0078] Mounting the orientation sensor on a non-muscular object part (e.g., a bone segment, joint region, or brace / prosthetic component) improves joint-angle accuracy by reducing soft-tissue artifact and wobble, provides a more stable and repeatable attachment aligned to anatomical axes, and keeps EMG electrode sites free for optimal placement, minimizing mechanical interference and electrical cross-talk. It is typically more comfortable and safer for the human and remains effective in the presence of edema, scars, or pain over muscle, and applies naturally to orthoses and prostheses where no muscle is available. The rigid mounting is more robust to motion and shock, lowering filtering burden and latency, and the setup is quicker and simpler (easier strapping, better hygiene and durability). Overall, this yields more accurate, stable, and repeatable joint-angle estimation, cleaner multimodal fusion with EMG, and smoother clinical and real-world workflows.

[0079] In an embodiment, the step of calculating the real-time movement and performance analysis indicator for the object part comprises calculating the difference between the muscular activity data and the corresponding muscular activity reference data for each acquired articular angle position measured by the orientation sensor, where if the difference is above or below a predefined threshold biofeedback command or information is issued. As mentioned above, the method is applicable to both injured and healthy object parts, enabling rehabilitation in the former case and improvement of strength, mobility, and coordination in the latter.

[0080] Calculating, at each joint angle, the measured muscle activity is compared to a reference value for that exact angle. If, as an example, the difference is too big (above or below a set limit), the method may be configured to immediately give feedback, like a beep, a vibration, or a visual cue, so one can adjust on the spot. Because the check is angle-by-angle, the feedback comes at the right moment in the movement, is easy to understand, and doesn’t require heavy computation. This makes exercises more targeted to the human, which reduces mistakes, and works better than methods that only look at overall peaks or ignore joint angle. KIS24001PWO Rey_Baldur

[0081] Kiso ehf.

[0082] 13

[0083] The biofeedback command may as an example include issuing moving instructions at least partly different from the pre-stored moving instructions, but this may be done visually via the display, via speech command, or simply move the object part in a different way. This may in an alternative embodiment, be continued until the difference between the first muscular activity data and the corresponding muscular activity reference data is above or below the pre-defined threshold.

[0084] In an embodiment, the step of calculating the real-time movement and performance analysis indicator for the object part comprises calculating the difference between the muscular activity data and the corresponding muscular activity reference data for each acquired articular angle position measured by the orientation sensor, where if the difference is above or below a predefined threshold biofeedback command is issued. The biofeedback command may as an example comprise issuing moving instructions at least partly different from the pre-stored moving instructions. This may in an embodiment be repeated until the difference between the first muscular activity data and the corresponding muscular activity reference data is above or below the pre-defined threshold. In that way, it is possible to adjust the movement of the object part via the biofeedback command until “sufficient” muscular activity is present and sufficient to build up the first muscular object part. In that way, it is possible to identify in real-time if the moving instructions characteristic for the muscular object are maximizing the result of the training for the human and changes in the movements may be proposed until the most optimal result is obtained.

[0085] In a second aspect of the invention, a system is provided for real-time movement and performance analysis for a human, comprising:

[0086] • at least one wireless mobile monitoring device configured to be placed on an object part of the human, comprising: o a communication module, o an orientation sensor, and / or o a muscular activity sensor, a computer system configured to present pre-stored moving instructions to the human, KIS24001PWO Rey_Baldur

[0087] Kiso ehf.

[0088] 14 where the orientation sensor and / or the muscular activity sensor is / are configured to acquire articular angle data and / or muscular activity data of the object part while the human moves the object part in accordance with the moving instructions, where the articular angle position data and / or the muscular activity data is / are communicated by the communication module to the computer system where the data is processed, where the processing includes comparing the received articular angle position data and / or the muscular activity data with reference data and determining a real-time movement and performance analysis indicator for the object part.

[0089] A technical advantage of the disclosed system is that the humans' movements are not arbitrary or random but are carried out in accordance with pre-stored moving instructions that are targeted to the human’s specific impairment, training or rehabilitation need.

[0090] Accordingly, the system mitigates compounded estimation error by eliminating dependence on human-provided “first” movements and by anchoring analysis to direct, high-fidelity measurements of joint pose and movement quality. Technical advantages include initialization- free operation through guided reference postures and automatic start-point detection that reject random or irrelevant motions; direct or better-constrained joint pose estimation (e.g., via articular angle sensing and / or kinematics constrained by anatomical priors) that reduces drift versus limb-orientation proxies; multimodal fusion (e.g., IMU / MARG, EMG, optical / encoder signals) with quality gating and outlier rejection so noisy or poorly executed motions do not seed the pipeline; confidence-weighted state estimates with uncertainty bounds that prevent low- confidence data from driving therapy decisions; closed-loop prompts on the display (or haptics / audio) that solicit a clean calibration motion when needed, shortening setup and improving repeatability; and a hierarchical inference architecture that derives secondary metrics (exercise classification, rep counting, fatigue, progress) from validated primary measurements rather than stacking inferences on inferences.

[0091] Collectively, these features suppress error propagation of the system, improve accuracy and reliability under motion artifacts and human variability, deliver more precise feedback and progress tracking, and enhance the therapeutic effectiveness of the program. KIS24001PWO Rey_Baldur

[0092] Kiso ehf.

[0093] 15

[0094] A further advantage is that the system relies on directly measured sensor data from the orientation sensor and / or muscular activity sensor, rather than on pose estimates derived from arbitrary initial movements. This avoids the compounding of inaccuracies inherent in prior-art systems, in which estimated joint angles are anchored to poorly executed initial actions. As a result, the real-time movement and performance analysis indicator for the object part becomes highly accurate and reliable, thereby delivering training programs that are both individualized and therapeutically relevant while enabling efficient real-time processing.

[0095] In an embodiment, the orientation sensor and / or the muscular activity sensor comprises an amplifier configured to process low-level analog signals in close proximity to a microprocessor and / or a Bluetooth transmitter. The amplifier placement is arranged to minimize signal path length and associated noise pickup while maintaining electromagnetic compatibility with the nearby digital and RF components. This configuration is achieved through specific printed circuit board (PCB) design considerations, including grounding strategies, component shielding, and trace routing optimized to reduce crosstalk and electromagnetic interference. The integration of the amplifier in this configuration enables compact packaging and improved signal fidelity without requiring physical separation between analog and digital domains.

[0096] In a further embodiment, the orientation sensor and / or the muscular activity sensor includes onboard memory configured to store data that are not processed in real time. Such data storage allows deferred signal analysis, local buffering, and event logging even when wireless transmission or real-time computation is unavailable. This architecture supports intermittent connectivity and ensures data integrity during periods of transmission loss or processor sleep modes.

[0097] In another embodiment, the sensor device incorporates passive contact-detection circuitry configured to assess the coupling between the sensor and the user’s skin. The circuitry operates without active current injection and relies on variations in impedance, capacitance, or potential at the skin interface to confirm stable contact. This feature provides reliable operation KIS24001PWO Rey_Baldur

[0098] Kiso ehf.

[0099] 16 in physiological monitoring applications and reduces power consumption compared to active contact-sensing techniques.

[0100] In yet another embodiment, the orientation sensor and / or the muscular activity sensor comprises a dual inertial measurement unit (IMU) configuration including two sets of accelerometer, gyroscope, and optionally magnetometer sensors. The dual-IMU arrangement allows cross-validation of motion data, correction of sensor drift, improved noise performance, and redundancy in dynamic conditions. The data from both IMlls may be fused through a common filtering algorithm to enhance kinematic accuracy.

[0101] In a related embodiment, a mobile or desktop application is associated with the system comprising a local database configured to enable data acquisition and temporary storage in offline mode, thereby allowing continuous recording in the absence of an internet connection. When connectivity is restored, the recorded data are synchronized with a remote database. This feature ensures uninterrupted data capture and reliable synchronization across multiple user sessions.

[0102] Each of the above features may be implemented independently or in combination within the same sensor system, depending on the desired configuration and scope of protection.

[0103] In an embodiment, the system is configured to present the performance analysis indicator as a game-based user interface in which the indicator may reflect a deviation of the first muscular activity data from the muscular activity reference data. In this embodiment, the deviation is mapped to one or more game variables, such as a score, progress meter, avatar position, or “target zone” occupancy. For example, when the deviation remains within a specified tolerance band around the reference data, the system awards points, advances levels, or maintains an on-screen avatar within a target corridor. When the deviation exceeds the tolerance, the system may deduct points, applies penalties, or displaces the avatar outside the corridor. KIS24001PWO Rey_Baldur

[0104] Kiso ehf.

[0105] 17

[0106] The system may further quantize deviation magnitudes into tiers (e.g., green / yellow / red) with corresponding rewards or penalties, adjust difficulty adaptively by narrowing or widening the tolerance band based on user performance over time; deliver multimodal feedback (visual, audio, and / or haptic) synchronized to the deviation to reinforce corrective actions; and implement timed challenges in which maintaining the deviation below a threshold for a predetermined duration yields bonus points or unlocks levels. In certain implementations, the score may be normalized to session length to enable cross-session comparisons, and anonymized results may be summarized for coach / clinician dashboards. This gamified mode increases engagement and adherence while preserving the underlying function of the performance analysis indicator as a real-time representation of deviation from the muscular activity reference data.

[0107] The pre-stored moving instructions presented by the computer system may be task-specific prompts that guide the human through defined motions rather than random movements. The instructions may call for basic joint excursions such as bending and straightening the elbow, rotating the wrist clockwise and counter-clockwise, lifting the arm to shoulder height and lowering it, performing internal and external shoulder rotation, flexing and extending the knee while seated, dorsiflexing and plantarflexing the ankle by lifting the heel or the toes, and slowly rotating the head left and right.

[0108] The pre-stored moving instructions may further specify resistance or strength tasks, for example holding a small dumbbell and performing repeated biceps curls, pulling a resistance band apart at chest level, executing a front squat with a light bar, stepping up to and down from a low platform, performing bilateral or single-leg calf raises, or pressing against a wall to simulate a chest press. Balance and stability prompts can include standing on one leg for a timed interval, shifting body weight rhythmically from left to right, walking heel-to-toe in a straight line, or performing repeated sit-to-stand transitions from a chair.

[0109] Where appropriate, the instructions may reflect occupational demands; for instance, a construction worker may be instructed to lift a 5-kg object from floor to waist height using a hiphinge pattern, an office worker may simulate prolonged typing posture followed by overhead KIS24001PWO Rey_Baldur

[0110] Kiso ehf.

[0111] 18 arm elevation, a nurse may practice controlled bending and straightening as if assisting a human, and a delivery worker may simulate carrying a package up a single step. Similarly, sport-specific prompts can be presented, such as a runner performing alternating high-knee drive to emulate stride mechanics, a swimmer executing controlled shoulder circles mimicking a freestyle stroke, a cyclist simulating pedal rotation while seated, or a tennis player rehearsing a forehand swing with or without a racket. Clinical or rehabilitation instructions can include, by way of example, post-operative knee extension to terminal range with a brief hold, controlled shoulder flexion to 90 degrees followed by return, repetitive grasp-and-release of a soft ball for hand function retraining, or a supine pelvic tilt for lumbar control. Resistance, tempo, range of motion, and repetition counts may be embedded in the instruction text to standardize execution.

[0112] The system may tailor these pre-stored instructions to the individual based on human-provided inputs such as occupation, age, gender, training status, and activity level. Thus, elderly, low- activity human may receive seated knee extensions, chair rises, and gentle arm lifts with slow tempo and short bouts, whereas a young, highly active human may be presented with weighted squats, resisted curls, or light plyometric hops with specified rest intervals. An office worker with reported postural complaints may be guided through scapular retraction drills, chest-opening stretches, and thoracic rotations, while a manual laborer may receive simulated lifting, carrying, and controlled twisting tasks using a prescribed load. In all cases, the instructions can be expressed in precise, stepwise prompts, for example, “hold a 2-kg weight in the right hand; perform ten controlled curls in five seconds each; maintain elbow close to torso; rest fifteen seconds; repeat three sets”, so that the resulting articular-angle and muscular-activity data are directly traceable to a defined movement template, enabling time alignment with instruction timelines and comparison to instruction-specific reference profiles.

[0113] In an embodiment, the computer system is configured to process both the articular angle position data and the muscular activity data, where the processing includes using the articular angle position data to register, time-align and / or normalize the muscular activity data with corresponding angle-specific muscular activity reference data for each articular angle position measured by the orientation sensor. KIS24001PWO Rey_Baldur

[0114] Kiso ehf.

[0115] 19

[0116] The computer system is further configured to, for each measured articular angle position, forming an angle-resolved paired dataset comprising the muscular activity data and the corresponding reference data and computing, in real time, a movement and performance analysis indicator for the object part, the indicator including at least one muscle activation or muscle-imbalance metric such as agonist-antagonist activation ratio, a side-to-side symmetry index, a co-contraction index, a deviation from an angle-specific normative envelope, and an imbalance severity score. As used herein, ‘time-align’ may include alignment based on articular angle derived from an IMU (e.g., a MARG sensor) and may additionally use other IMU-derived signals, such as accelerometer-based impact estimation; such impact events (e.g., heel-strike during gait or landing in jumping) can serve as fiducial markers for temporal alignment.

[0117] Accordingly, by processing both the articular angle position data and the muscular activity data in combination, and by aligning the muscular activity data with angle-specific reference data for each measured articular angle position, the system provides angle-resolved datasets that enable a more accurate and physiologically meaningful assessment of muscle function. This structured processing allows the computer system to, e.g. compute, in real time, movement and performance analysis indicators that explicitly incorporate muscle-imbalance metrics such as agonist-antagonist activation ratios, side-to-side symmetry indices, co-contraction indices, deviations from angle-specific normative envelopes, and imbalance severity scores. As a result, the system may be utilized to improve the ability to detect lack of muscle control and muscle imbalance that may persist despite exercise, thereby providing earlier identification of maladaptive movement patterns, reducing the risk of continued pain and restricted joint motion, and enabling targeted interventions that help prevent long-term joint wear and deterioration.

[0118] Example: This has been applied to a professional handball player after patellar-tendon repair, these indicators guided VM-focused interventions, especially in terminal knee extension, with progressive remote biofeedback. Outcomes included left knee-extension strength improving from 40 Nm (Oct 2023) to 125 Nm (Apr 2024) and thigh-girth deficits shrinking from 3 cm / 5 cm at +5 / +15 cm above the patella to 0 cm / 2 cm; angle-resolved metrics showed VM / VL imbalance normalizing toward the reference envelope alongside functional gains (light jogging, direction changes, heavier lifts). A remaining issue, lateral peripatellar pain during heavy terminal extension, is now targeted using VM-biased technique constraints informed by the indicators. KIS24001PWO Rey_Baldur

[0119] Kiso ehf.

[0120] 20

[0121] Technically, the angle-resolved EMG-kinematics approach exposes phase-specific neuromuscular deficits that angle-agnostic methods miss, enables precise angle-windowed prescriptions (load, tempo, ROM) to recruit VM and limit VL dominance exactly where needed, and delivers real-time, actionable guidance that accelerates detection of maladaptive control, reduces pain and restrictions, and helps prevent long-term joint wear. The rehabilitation became significantly more efficient, and the training program was built around exercises that would likely never would have been thought to use otherwise. The handball player's progress was evident, both in strength and capacity, as well as in the quality of his movements.

[0122] In an embodiment, the system is further configured to operate in a “relaxation mode” in which it processes both articular angle position data and muscular activity data. The processing includes using the articular angle position data to register, time-align, and / or normalize the muscular activity data with corresponding angle-specific muscular activity reference data for each articular angle position measured by the orientation sensor. In this mode, the system is configured to monitor and maintain inactivity of a specified inflamed or overactive muscle during functional movement or exercise. The system detects the muscle’s activity level and generates a performance analysis indicator only when the measured activity exceeds a defined relaxation threshold, for example when the activity rises above a fixed percentage of the muscle’s maximum voluntary contraction (MVC) or above a resting baseline for a predetermined duration. Further, this mode enables verification that the targeted muscle remains relaxed during tasks in which it should not engage, thereby assisting in rehabilitation of e.g. inflamed muscles, such as those affected by myositis.

[0123] Another scenario where the system is configured to operate in a “relaxation mode” is to ensure a specified inflamed muscle (e.g., a shoulder muscle with myositis) stays quiet during a task where it should not engage. The performance analysis indicator triggers then only when that muscle’s activity rises above a relaxation threshold (for example, above a fixed percentage of MVC or above resting baseline for a short, defined duration). The thresholds may be fixed or adaptive and may also be expressed as a ratio to a target muscle to detect excessive coactivation. As an example, during knee-extension rehab, an athlete trains the quadriceps while the hamstrings should remain relaxed. If hamstring activity exceeds 10% MVC for longer than KIS24001PWO Rey_Baldur

[0124] Kiso ehf.

[0125] 21

[0126] 200 ms, or if the hamstring / quadriceps ratio crosses a preset limit, the system alerts the user or clinician.

[0127] The computer system may in an embodiment comprise a computer device comprising a device receiver, a display and a device processor for controlling the device receiver and the display. The computer device may as an example comprise a stationary computer, or any type of a portable device such as a tablet device, mobile phone and the like. The device receiver is configured to receive the data communicated by the communication module and the display is in an embodiment configured to display the moving instructions via the display. The communication of the data via the communication module to the device receiver may be done via communication protocol such as, but not limited to, WIFI or Bluetooth. The computer device may also comprise a memory for storing the received data, which may be beneficial in case the communication is not present, or the signal is bad.

[0128] In an embodiment, the computer system further comprises a computer platform comprising a platform memory, a platform transceiver and a platform processor for controlling the platform memory and the platform transceiver. The moving instructions may be stored by the platform memory, and the platform processor may be configured to instruct the platform transceiver to transmit the stored moving instructions to the computer device where the moving instructions are stored in a device memory present in the computer device. This may as an example be done via a Transmission Control Protocol / lnternet Protocol (TCP / IP) protocol. The transmitted stored moving instructions may in another alternative embodiment be displayed in real-time at the computer device via the display.

[0129] Technical advantages include decoupling content generation from device execution to reduce device-side compute and power consumption; reliable, in-order, congestion-controlled delivery and broad network interoperability via TCP / IP; local persistence in the device memory enabling offline execution, fast resume after link interruptions, and versioned rollbacks; low-latency visualization when instructions are streamed for real-time display; centralized scheduling and throttling at the platform for bandwidth management and prioritization; one-to-many distribution (e.g., multicast or sequential unicast) for synchronized updates across multiple devices; and KIS24001PWO Rey_Baldur

[0130] Kiso ehf.

[0131] 22 improved security and auditability when the platform applies transport-layer protections (e.g., TLS), message authentication, and logging. Collectively, these features improve robustness, scalability, and determinism of instruction delivery and presentation while reducing latency and maintenance burden on the endpoint devices.

[0132] In an embodiment, the orientation sensor comprises an Inertial Measurement Unit (IMU) sensor or Magnetic Angular Rate and Gravity (MARG) sensor.

[0133] Using IMU / MARG is suitable when the orientation is of relevance, but these sensors can also measure acceleration and the magnetic field, which might be of interest, for example, for automatically detecting steps. Therefore, it might be worth mentioning that the motion sensors may also be used to record acceleration, for example for impact detection, and to make angle measurement more accurate. Thus, the acceleration may also be measured for e.g. impact detection, to ensure more accurate articular angle detection.

[0134] Moreover, this yields technical advantages including high-rate inertial and magnetic measurements that can be fused (e.g., complementary / Kalman filtering) to suppress gyro drift and accelerometer noise, thereby providing more accurate and stable orientation and joint angle estimates over time; gravity-vector and heading references that improve long-term stability and dynamic performance during fast motion; robust operation under shock and vibration with outlier rejection and adaptive weighting to handle magnetic disturbances; reduced latency and power due to on-package integration of gyroscope, accelerometer, and (optionally) magnetometer, thereby reducing on-device compute requirements and enabling real-time control loops; secondary functionality from the same sensors, such as step / impact detection and motion classification, without additional hardware; simplified calibration and continuous bias compensation that maintain accuracy across temperature and aging; synchronized rate / acceleration streams that enhance kinematic modeling; and lower BOM cost and size from consolidating sensing functions into a single IMU / MARG module. KIS24001PWO Rey_Baldur

[0135] Kiso ehf.

[0136] 23

[0137] In an embodiment, the muscular activity sensor comprises an Electromyography (EMG) sensor, such as surface Electromyography (EMG) sensor and where the muscular activity data is a surface or non-surface muscular activity data, e.g. needle or fine wire emg.

[0138] Technical advantages include early detection of muscle activation and human intent prior to observable motion, enabling lower-latency control and improved closed-loop responsiveness; rich amplitude-frequency features (e.g., RMS, median frequency) that support fatigue estimation, force / torque inference, and adaptive assistance; high signal quality via differential electrode configurations with common-mode rejection to suppress power-line and motion artifacts; flexible deployment from non-invasive sEMG for ease of use and repeatability to intramuscular EMG for deeper muscle selectivity; reduced sensor count and BOM by deriving intent and state directly from bioelectrical activity rather than indirect kinematic proxies; improved robustness through on-sensor preprocessing (filtering, rectification, normalization) and automatic gain / impedance adaptation across skin types and temperatures; and seamless fusion with inertial signals (when present) to enhance joint angle estimation, intent classification, and disturbance rejection.

[0139] In an embodiment, and similarly as addressed above, the step of processing the articular angle position data and / or the muscular activity data includes processing both the articular angle position data and the muscular activity data, preferably simultaneously. The processing may be performed by a device processor comprised in the wireless mobile monitoring device(s) or by the platform processor. The processing comprises:

[0140] • utilizing the articular angle data to align the muscular activity data with corresponding muscular activity reference data for each acquired articular angle position measured by the orientation sensor, and

[0141] • utilizing the resulting muscular activity data pair for each acquired articular angle position measured by the orientation sensor in calculating the real-time movement and performance analysis indicator for the muscular object part.

[0142] The system enables, while performing the above-mentioned movements and calculations in real-time while the human is performing the moving instructions / exercises, acquiring the KIS24001PWO Rey_Baldur

[0143] Kiso ehf.

[0144] 24 necessary data needed as input data to make an individualized training program and determine the exercises that maximizes the result of the training for the human when handling musculoskeletal problems and shortens the treatment time.

[0145] The platform processor may in an embodiment further be configured to present in real-time the measured muscular activity data and / or the muscular activity reference data. This may as an example be done via the display where the human and / or any physical therapists or medical expert may visually see in real-time the muscular activity, e.g. compared to target activity, where the human and / or the medical expert may act on it accordingly, e.g. by changing the movements if there is too much deviation between the muscular activity compared to target activity indicated by the performance analysis indicator.

[0146] As will be discussed in more details later, multiple of the wireless mobile monitoring device may be placed on the object part during the above-mentioned processing, such as placing a wireless mobile monitoring device on a healthy object part and corresponding injured object part, e.g. right and left shoulder, right and left arm, right and left knee etc., where the object part “pairs” may be moved simultaneously, or individually.

[0147] As mentioned above, in some instances, humans suffer from lack of muscle control and muscle imbalance which manifests in such a way that one muscle takes over the function of another. Causes of muscle imbalance can be, for example, after an accident or strain. As a result, the current exercises the human is doing may not improve this condition at all, resulting in that pain persists, as well as restricted joint motion. Prolonged imbalance can lead to joint wear and tear, among other issues, which eventually may require surgical intervention.

[0148] Thus, by measuring the muscular activity of more than one muscular object part, it is possible to identify in real-time via the performance analysis indicator if one muscle has “taken over” another muscle for which the muscular training program is designed for to build up the muscular activity. This may as an example be based on if the deviation is above or below a pre-defined threshold, the performance analysis indicator may indicate this via any type of a command, sound or visually to the human or physiotherapist. KIS24001PWO Rey_Baldur

[0149] Kiso ehf.

[0150] 25

[0151] Based on the above, a compact and human-friendly system is provided for such a movement and performance analysis that does not e.g. require spacious area.

[0152] In general, the various aspects of the invention may be combined and coupled in any way possible within the scope of the invention. These and other aspects, features and / or advantages of the invention will be apparent from and elucidated with reference to the embodiments described hereinafter.

[0153] BRIEF DESCRIPTION OF THE DRAWINGS

[0154] Embodiments of the invention will be described, by way of example only, with reference to the drawings, in which

[0155] Figure 1 shows an embodiment of a system according to the present invention for real-time movement and performance analysis for a human,

[0156] Figure 2 shows a flowchart of a method according to the present invention method of real-time movement and performance analysis for a human,

[0157] Figure 3 depicts graphically where a human follows moving instructions shown by a silhouette on a laptop, tablet, mobile phone,

[0158] Figure 4 shows the embodiment in figure 3 but where the moving instructions are given orally by a physical therapist or a medical expert, KIS24001PWO Rey_Baldur

[0159] Kiso ehf.

[0160] 26

[0161] Figure 5 shows an embodiment where a second wireless mobile monitoring device, comprising a second transmitter, a second orientation sensor and a second muscular activity sensor is placed on the human,

[0162] Figure 6 shows another scenario where the three wireless monitoring devices being utilized on each arm, and

[0163] Figure 7 shows an exemplary embodiment where the human starts with moving the right healthy arm with the second wireless mobile monitoring device in accordance to the moving instructions as discussed in relation to figures 3 and 4.

[0164] DESCRIPTION OF EMBODIMENTS

[0165] Figure 1 shows an embodiment of a system 100 according to the present invention for real-time movement and performance analysis for a human 102, where the system comprises a wireless mobile monitoring device 101 , a computer device 112 and an external computer platform 107.

[0166] The first wireless mobile monitoring device 101 is attached, e.g. via strip or adhesive electrodes, to an object part of the human 102 which may be a muscular object part or non-muscular object part. The monitoring device comprises a communication module (C_M) 104, an orientation sensor (O_S) 106, a memory 103 and a muscular activity sensor (M_A) 105.

[0167] The orientation sensor may comprise an Inertial Measurement Unit (IMU) sensor or Magnetic Angular Rate and Gravity (MARG) sensor and the muscular activity sensor may in an embodiment comprise a surface ElectroMyoGraphy (EMG) sensor and where the muscular activity data is a surface muscular activity data where the EMG sensor preferably comprises at least two electrodes.

[0168] In the embodiments discussed in relation to the figures, both the articular angle position data anand the muscular activity data mnis measured simultaneously and where the data pair (an, mn) KIS24001PWO Rey_Baldur

[0169] Kiso ehf.

[0170] 27 113 is communicated in real time by the communication module 104 via communication protocol such as WIFI or Bluetooth 140 to the computer device 112. In the embodiment illustrated here, the data is forwarded to the external platform computer 107 where the data is processed. These data may also be stored in the memory 103, which may be beneficial in case the communication is not present or the signal is bad.

[0171] It should be noted that in another alternative embodiment the data may be processed and displayed at the computer device 112, i.e. where the processing steps illustrated here are performed by the computer device 112.

[0172] The processing includes comparing the received articular angle position data and the muscular activity data with reference data and determining a real-time movement and performance analysis indicator for the object part.

[0173] The computer device 112 may be any type of portable computer such as a tablet computer, mobile phone and the like comprising a device transceiver (D_T) 135, a display (D) 132, a device memory (D_M) 131 and a device processor (D_P) 134 for controlling the device receiver, the display and the device memory. The device transceiver 135 is configured for receiving the data 113 communicated by the communication module 104 and the display 132 is configured to display the above-mentioned moving instructions.

[0174] The external platform computer 107 comprises a platform memory (P_M) 111 , a platform transceiver (P_T) 109 and a platform processor (P_P) 108 for controlling the platform memory and the platform transceiver. The moving instructions may be stored in the platform memory, and the platform processor may be configured to instruct the platform transceiver to transmit the stored moving instructions to the computer device 112 where the moving instructions may subsequently be stored in the device memory 131. The step of transmitting may as an example be done via a Transmission Control Protocol / lnternet Protocol (TCP / IP) protocol. The transmitted stored moving instructions may in another alternative embodiment be displayed in real-time at the computer device 112 via the display 132. KIS24001PWO Rey_Baldur

[0175] Kiso ehf.

[0176] 28

[0177] Processing the data pair (an, mn) 113 includes utilizing the first articular angle data to align the first muscular activity data with corresponding muscular activity reference data r_mnfor each acquired articular angle position measured by the first orientation sensor, and utilizing the resulting data point (an, mn, r_mn) for each acquired articular angle position measured by the first orientation sensor in calculating a real-time movement and performance analysis indicator for the first muscular object part.

[0178] The external computer system 107 is further configured to transmit the processed data (an, mn, r_mn) 114 to the computer device 112 where the data points and the performance analysis indicator may be presented in real-time for the human. The performance analysis indicator may in the embodiment illustrated here reflect a deviation of the first muscular activity data from the muscular activity reference-data as indicated by the arrow 120.

[0179] In an alternative embodiment, the performance analysis indicator may be presented as a gamebased III in which the indicator reflects a deviation of the first muscular activity data from the muscular activity reference data. The deviation may be mapped to game variables, e.g., score, progress meter, avatar position, or target-zone occupancy. When the deviation stays within a tolerance band, the system awards points or advances levels; when it exceeds the band, it deducts points or displaces the avatar. Also, the system may adapt difficulty by adjusting the tolerance band, provide visual / audio / haptic feedback synchronized to deviation, and run timed challenges that reward maintaining deviation below threshold for a set duration.

[0180] As mentioned before, the processing of the data may be done at the computer device 112 side and presented in real time. This data may be sent to the external platform computer 107 where the data is amongst others stored.

[0181] Figure 2 shows a flowchart of a method according to the present invention method of real-time movement and performance analysis for a human. KIS24001PWO Rey_Baldur

[0182] Kiso ehf.

[0183] 29

[0184] In step (SO) 200, before it comes to the movement, a medical or physical therapist needs to know how to set up the equipment, i.e. which equipment goes where. These instructions may as an example be presented in a similar way to the movement instructions for the human discussed here below.

[0185] In step (S1) 201, pre-stored moving instructions are provided characteristic for an object part of the human and presented to the human and / or medical or physical therapists.

[0186] In step (S2) 202, a first wireless mobile monitoring device is provided configured to be placed on the object part of the human, which may be a muscular or non-muscular object part.

[0187] In step (S3) 203, the articular angle position of the object part is measured by the orientation sensor while the human moves the first object part in accordance with the moving instructions.

[0188] In step (S4) 204, the muscular activity of the object part is measured by the muscular activity sensor, simultaneous to the measured articular angle position of the object, while the human moves the object part in accordance with the moving instructions.

[0189] In step (S5) 205, the articular angle position data and the muscular activity data is communicated in real-time to a computer device such as laptop, mobile phone, and the like via Bluetooth or WiFi, where the data is forwarded to a computer platform where the data is processed. The processing many include utilizing the articular angle data to align the muscular activity data with corresponding muscular activity reference data for each acquired articular angle position measured by the orientation sensor and utilizing the resulting muscular activity data pair for each acquired articular angle position measured by the orientation sensor in calculating a real-time movement and performance analysis indicator for the object part. The performance analysis indicator may include calculating the difference between the first muscular activity data and the corresponding muscular activity reference data for each acquired articular angle position measured by the orientation sensor, where if the difference is above or below a pre-defined threshold biofeedback command is issued. This may be any type of visual feedback KIS24001PWO Rey_Baldur

[0190] Kiso ehf.

[0191] 30 presented as an example on a display device where the human and medical expert are able to monitor the muscular activity in real-time. If the muscular activity is above or below a predefined threshold, i.e. not as it should be, the biofeedback command may include issuing moving instructions at least partly different from the pre-stored moving instructions. This may be repeated until the difference between the muscular activity data and the corresponding muscular activity reference data is within a threshold window.

[0192] In step (S6) 206, another wireless mobile monitoring device (second) is provided configured to be placed on another object part of the human different from the previous object part. More than two such monitoring devices may obviously be used.

[0193] In step (S7) 207, the articular angle position of the other object part is measured by the second orientation sensor while the human moves the object part in accordance with the moving instructions.

[0194] In step (S8) 208, the articular angle position data and the muscular activity data measured by the second monitoring device is transmitted in real-time to the computer device that forwards the data to the platform computer where the data is stored and processed as discussed above.

[0195] The other object part may be the same muscle as the previous object part but on the other side of the human, e.g. biceps on right arm and left arm, or left and right shoulder parts.

[0196] In an embodiment, the object part may be the injured object part and the previous object part may be the healthy part, where the muscular activity reference data is the measured muscular activity data from the healthy muscular object part.

[0197] In another embodiment, the muscular activity reference data comprises pre-stored data from plurality of healthy muscular objects from other humans. KIS24001PWO Rey_Baldur

[0198] Kiso ehf.

[0199] 31

[0200] At the end of the recording, a summary of the results may be presented, e.g. displayed on the display, in a simple format so that the medical or physical therapists can make a decision about the next steps. This could as an example be where the muscle is: Left / right deltoid, left / right upper-trapezius, left / right lower-trapezius, left / right serratus, where the data presented in the overview might be max voltages together with the time.

[0201] Figure 3 depicts graphically where a human 302 follows moving instructions shown by a silhouette 302 on a display of a computer device, which may be a laptop, tablet, mobile phone 112, PC computer, where the accumulated muscular activity data and the articular angle data 113a,b,c as discussed in relation to figures 1 to 2 for each angular position aO-an is transmitted in real-time to the computer device 112, that forwards the data to the external computer platform 107, where the processing steps discussed in relation to figure 2 are performed. The processed data 114 is presented on the display of the computer device 112. As discussed previously, the performance analysis indicator 120 may reflect a deviation of the first muscular activity data from the muscular activity reference-data as indicated by the arrow. In that way, the human and / or medical expert may monitor the muscular activity in real-time and adjust the exercises if needed until the deviation is below a pre-defined threshold value.

[0202] Figure 4 shows the embodiment in figure 3 but where the moving instructions are given orally by a physical therapist or a medical expert 403.

[0203] Figure 5 shows an embodiment where a second wireless mobile monitoring device 501 , comprising a second transmitter, a second orientation sensor and a second muscular activity sensor. The second wireless mobile monitoring device is placed on the other upper arm on the human 302 on a second object part of the human different from the first muscular object part, which as shown here is the same but on the other arm.

[0204] Similar measurements are done where the articular angle position of the second object part is measured together with the muscular activity data while the human moves the second object part in accordance with the moving instructions, where the data is transmitted in real-time to the computer device as discussed previous figures. KIS24001PWO Rey_Baldur

[0205] Kiso ehf.

[0206] 32

[0207] The first object part shown in the figures may be seen as injured muscle part, whereas the second object part may be seen as a healthy object part. In that way, the measurements for both the muscular object parts may be compared in real-time.

[0208] The movement of the arms does not necessarily need to be done simultaneously as shown here.

[0209] It should be noted that the object part may also include muscle or non-muscle object parts on the legs on the human and / or the back and / or the stomach, just to mention few.

[0210] Figure 6 shows another scenario where the three wireless monitoring devices 101a-c and 501a- c are being utilized.

[0211] The number of devices may vary depending on the injury, in some instances, e.g.16 devices and even over 30 devices may be used.

[0212] Figure 7(a) shows an exemplary embodiment where the human starts with moving the right healthy arm with the second wireless mobile monitoring device 501 in accordance to the moving instructions as discussed in relation to figures 3 and 4, where for each articular angle position of the second muscular object a anthe muscular activity is measured r_m-! - r_mnand transmitted to the external computer system where the measured muscular activity is used as said reference data.

[0213] Figure 7(b) shows corresponding data a’ra’n and r’_m! - r’_mnwhere the movement for the other injured arm is moved according to the moving instruction and transmitted (not shown here) as shown in figure 3 and transmitted to the computer device 112, where the above mentioned processing steps are performed, using the measured data of the healthy arm as reference data. KIS24001PWO Rey_Baldur

[0214] Kiso ehf.

[0215] 33

[0216] This issue could also be solved with the same measuring unit by moving it between (then the inactive hand in the image would be without the measuring device). This way, for example, one could compare 4 muscles on the right and 4 muscles on the left side even though using four wireless mobile monitoring devices.

[0217] Begin example.

[0218] Angle-Resolved EMG-Kinematics in Post-Surgical Knee Rehabilitation

[0219] Overview

[0220] In one implementation, the system integrates articular (joint) angle position data with muscular activity (EMG) data and uses the angle data to register, time-align, and normalize the EMG signals against angle-specific reference profiles. For each measured articular angle, the system forms an angle-resolved paired dataset (measured EMG and the corresponding angle-specific reference) and computes, in real time, one or more performance indicators, including at least one muscle-imbalance metric selected from: an agonist-antagonist activation ratio, a side-to- side symmetry index, a co-contraction index, a deviation from an angle-specific normative envelope, and an imbalance severity score. This example illustrates the operation and clinical outcome in a professional athlete following patellar tendon repair.

[0221] 1. Human and Clinical Context

[0222] Subject: Professional handball player (male), Club US Ivry (Paris), age 24 at time of injury; multiple youth national caps and two senior caps.

[0223] Injury event (24 Feb 2023): Complete rupture of the left infrapatellar tendon during final practice before debut match.

[0224] Surgery: Early March 2023 in Paris; knee braced for 8 weeks; light mobilization initiated at week 4. KIS24001PWO Rey_Baldur

[0225] Kiso ehf.

[0226] 34

[0227] 2. Conventional Rehabilitation (Pre-lmplementation Snapshot)

[0228] May-Aug 2023: In-clinic rehabilitation in France; manipulation under anesthesia in early July to improve range of motion (ROM). Subsequent rehabilitation block in Iceland (10 Jul-15 Aug) focused on ROM maintenance, light strengthening, and proprioception, with limited progress.

[0229] Status on 15 Oct 2023 (before system introduction): Antalgic gait; unable to ascend stairs; marked quadriceps atrophy / shortening; isokinetic strength left 40 Nm vs right 159 Nm (-25%); thigh-girth deficit 3 cm at +5 cm and 5 cm at +15 cm above patella.

[0230] 3. System Configuration and Data Acquisition

[0231] Sensors:

[0232] Orientation sensor affixed to the shank to measure knee angle throughout flexion-extension.

[0233] EMG electrodes placed over quadriceps heads, including vastus medialis (VM) and vastus lateralis (VL).

[0234] Processing (computer system):

[0235] Acquire continuous knee-angle and EMG signals during therapeutic exercises (open-chain knee extensions and closed-chain drills).

[0236] Register and time-align EMG to the instantaneous articular angle.

[0237] Normalize EMG against angle-specific reference datasets for each monitored muscle.

[0238] For each discrete angle bin, form an angle-resolved paired dataset (measured EMG and reference EMG).

[0239] Compute real-time indicators: agonist-antagonist activation ratio (VM / VL), side-to-side symmetry index, co-contraction index, deviation from a normative envelope, and an imbalance severity score.

[0240] Display indicators as real-time biofeedback to guide exercise execution and loading.

[0241] 4. EMG-Guided Intervention

[0242] Initiation: Early December 2023 in collaboration with the rehabilitation center. KIS24001PWO Rey_Baldur

[0243] Kiso ehf.

[0244] 35

[0245] Baseline finding: Minimal VM activation under standard protocols; pronounced VL dominance, especially near terminal knee extension (TKE).

[0246] Adaptation: Based on the computed indicators, exercises were re-parameterized (joint-angle windows, tempo, range constraints, foot position, and load) to enhance VM recruitment in the terminal arc while constraining VL over-activation.

[0247] Adjuncts: Neuromuscular electrical stimulation (as tolerated), soft-tissue therapy, progressive strength training (machines, free weights, bodyweight), stability / mobility drills, cycling, and ROM work. 5. Quantitative Outcomes

[0248] (a) Strength (isokinetic peak torque)

[0249] (b) Morphology (thigh girth) KIS24001PWO Rey_Baldur

[0250] Kiso ehf.

[0251] 36

[0252] (c) Functional milestones

[0253] Mid-Feb 2024: Initiated light jogging on turf, direction-change ladder drills, jump rope, and heavier lifts. Persistent lateral peripatellar pain during TKE was detected by the system as coinciding with VM under-recruitment and VL dominance in the final 10-15° of extension. 3 May 2024: Continued training remotely in Paris using an app-based biofeedback client; performed higher-load knee extensions with VM-biased control in the terminal arc.

[0254] 6. Angle-Resolved Indicators and Findings

[0255] Agonist-antagonist ratio (VM / VL): Initially below the reference envelope in TKE; improved to within or near the normative envelope after VM-focused training.

[0256] Side-to-side symmetry index: Gradual improvement toward symmetry across angles, with minor residual asymmetry near TKE.

[0257] Deviation from normative envelope: Decreased across mid-range, with residual deviation at angles corresponding to reported pain. Imbalance severity score: Declined progressively during EMG-guided training, corresponding with clinical improvements.

[0258] Table I — Summary of Quantitative Outcomes KIS24001PWO Rey_Baldur

[0259] Kiso ehf.

[0260] 37

[0261] 7. Technical Advantages Demonstrated

[0262] Revealed phase-specific (angle-dependent) neuromuscular deficits not detectable using angleagnostic methods.

[0263] Enabled targeted exercise prescription (load, angle windows, tempo) to promote VM activation and reduce VL dominance at specific joint angles.

[0264] Allowed early identification of maladaptive activation patterns linked to pain and restricted motion, guiding corrective interventions that reduced symptoms and improved recovery trajectory.

[0265] 8. Current Status and Outlook

[0266] Between 2023 and 2024, the athlete progressed from marked weakness and atrophy to clinically meaningful strength and function. Ongoing objectives include: (i) elimination of lateral peripatellar pain during heavy terminal knee extension (TKE); (ii) further VM hypertrophy and improved VM / VL neuromuscular balance in the terminal range; and (iii) resolution of the residual +15 cm girth / strength asymmetry. The angle-resolved indicators continue to guide loading and technique during remote rehabilitation.

[0267] End example.

[0268] While the invention has been illustrated and described in detail in the drawings and foregoing description, such illustration and description are to be considered illustrative or exemplary and not restrictive; the invention is not limited to the disclosed embodiments. Other variations to the KIS24001PWO Rey_Baldur

[0269] Kiso ehf. 38 disclosed embodiments can be understood and effected by those skilled in the art in practicing the claimed invention, from a study of the drawings, the disclosure, and the appended claims. In the claims, the word "comprising" does not exclude other elements or steps, and the indefinite article "a" or "an" does not exclude a plurality. The mere fact that certain measures are recited in mutually different dependent claims does not indicate that a combination of these measures cannot be used to advantage.

Claims

KIS24001PWO Rey_BaldurKiso ehf.39CLAIMS1. A method of real-time movement and performance analysis for a human, where the method comprises:• providing a pre-stored moving instructions characteristic for an object part, such as muscular or non-muscular object part, of the human,• presenting the moving instructions to the human,• providing at least one wireless mobile monitoring device configured to be placed on the object part of the human, comprising: o a communication module, o an orientation sensor, and / or o a muscular activity sensor,• measuring, by the orientation and / or the muscular activity sensor, the articular angle and / or muscular activity of the object part, while the human moves the object part in accordance with the moving instructions,• communicating, by the communication module, the measured articular angle data and / or the muscular activity data to a computer system where the data is processed, where the processing includes,• comparing the received articular angle position data and / or the muscular activity data with reference data and determining a real-time movement and performance analysis indicator for the object part.

2. The method according to claim 1 , further comprising initially, in response to selecting an object part or object parts on the human to be analyzed, providing information indicating where to place the orientation and / or the muscular activity sensor on the human.KIS24001PWO Rey_BaldurKiso ehf.

403. The method according to claim 1 or 2, wherein the step of processing the received articular angle position data and / or the muscular activity data includes processing both the articular angle position data and the muscular activity data, where the step of comparing comprises:• utilizing the articular angle data to align the muscular activity data with corresponding muscular activity reference data (114) for each acquired articular angle position measured by the orientation sensor, and• utilizing the resulting muscular activity data pair for each acquired articular angle position measured by the orientation sensor in calculating the real-time movement and performance analysis indicator (120) for the muscular object part.

4. The method according to any of the preceding claims, wherein multiple of wireless mobile monitoring devices are utilized to be placed on an injured object part and corresponding healthy object part on the human, where the reference data comprises the data measured on the corresponding healthy object.

5. The method according to any of the preceding claims, wherein muscular activity reference data comprises pre-stored data from plurality of healthy muscular objects from other humans.

6. The method according to any of the preceding claims, wherein step of calculating the realtime movement and performance analysis indicator for the object part comprises calculating the difference between the muscular activity data and the corresponding muscular activity reference data for each acquired articular angle position measured by the orientation sensor, where if the difference is above a pre-defined threshold biofeedback command is issued.

7. The method according to claim 6, wherein the biofeedback command comprises issuing moving instructions at least partly different from the pre-stored moving instructions.KIS24001PWO Rey_BaldurKiso ehf.

418. The method according to claim 7, wherein the biofeedback command is issued until the difference between the first muscular activity data and the corresponding muscular activity reference data is above or below the pre-defined threshold.

9. The method according to any of the preceding claims, wherein the step of processing the received articular angle position data and / or the muscular activity data, and comparing the received articular angle position data and / or the muscular activity data with reference includes:• processing the received articular angle position data and compare it with articular angle position reference data, or• processing the received muscular activity data and compare it with muscular activity reference data.

10. The method according to any of the preceding claim, wherein processing the articular angle position data and / or the muscular activity data (113) includes processing both the articular angle position data and the muscular activity data by the computer system, where the processing includes using the articular angle position data to register, time-align and / or normalize the muscular activity data with corresponding angle-specific muscular activity reference data (114) for each articular angle position measured by the orientation sensor, wherein the computer system is used to, for each measured articular angle position, forming an angle-resolved paired dataset comprising the muscular activity data and the corresponding reference data and computing, in real time, a movement and performance analysis indicator (120) for the object part, the indicator including at least one muscle-imbalance metric such as agonist-antagonist activation ratio, a side-to-side symmetry index, a co-contraction index, a deviation from an angle-specific normative envelope, and an imbalance severity score.

11. The method according to any of the preceding claim, wherein orientation sensor comprises an Inertial Measurement Unit (IMU) sensor or Magnetic Angular Rate and Gravity (MARG) sensor, and / or wherein muscular activity sensor comprises an Electromyography (EMG) sensor.KIS24001PWO Rey_BaldurKiso ehf.4212. A system (100) for real-time movement and performance analysis for a human (102, 302), comprising:• at least one wireless mobile monitoring device (101) configured to be placed on an object part of the human, comprising: o a communication module (104), o an orientation sensor (106), and / or o a muscular activity sensor (105),• a computer system configured to present pre-stored moving instructions to the human, where the orientation sensor and / or the muscular activity sensor is / are configured to acquire articular angle position data and / or muscular activity data of the object part while the human moves the object part in accordance with the moving instructions, where the articular angle position data and / or the muscular activity data (113) is / are communicated by the communication module to the computer system where the data is processed, where the processing includes comparing the received articular angle position data and / or the muscular activity data with reference data and determining a real-time movement and performance analysis indicator for the object part.

13. The system according to claim 12, wherein processing, by the computer system, the articular angle position data and / or the muscular activity data (113) includes processing both the articular angle position data and the muscular activity data by the computer system, where the processing includes using the articular angle position data to register, time-align and / or normalize the muscular activity data with corresponding angle-specific muscular activity reference data (114) for each articular angle position measured by the orientation sensor, wherein the computer system is further configured to, for each measured articular angle position, forming an angle-resolved paired dataset comprising the muscular activity data and the corresponding reference data and computing, in real time, a movement and performance analysis indicator (120) for the object part, the indicator including at least one muscle-imbalance metric such as agonist-antagonist activation ratio, a side-to-side symmetry index, a cocontraction index, a deviation from an angle-specific normative envelope, and an imbalance severity score.KIS24001PWO Rey_BaldurKiso ehf.4314. The system according to claim 12 or 13, wherein the computer system comprises a computer device comprising a device receiver, a display and a device processor for controlling the device receiver and the display, where the device receiver is configured for receiving the data communicated by the communication module and where the display is configured to display the moving instructions via the display.

15. The system according to any of the claims 12 to 14, wherein the computer system further comprises a computer platform comprising a platform memory, a platform transceiver and a platform processor for controlling the platform memory and the platform transmitter, where the moving instructions are stored by the platform memory, and where the platform processor is configured to instruct the platform transceiver to transmit the stored moving instructions to the computer device where the moving instructions are stored in a device memory comprised in the computer device or displayed in real-time at the computer device via the display.

16. The system according to any of the claims 12 to 15, wherein orientation sensor comprises an Inertial Measurement Unit (IMU) sensor or Magnetic Angular Rate and Gravity (MARG) sensor.

17. The system according to any of the claims 12 to 16, wherein muscular activity sensor comprises an Electromyography (EMG) sensor.

Images