Electronic Piano-Based Cognitive Ability Evaluation System

Abstract

The present invention relates to an electronic piano-based cognitive ability evaluation system, and more specifically, to an electronic piano-based cognitive ability evaluation system that collects performance information including sound start and end signals, key press intensity, and occurrence time from an electronic piano or MIDI controller in an input module, analyzes a MIDI score file to generate information on the occurrence time, duration, pitch, and velocity of a reference note, synchronizes the two pieces of information in a comparison analysis module to calculate timing error, pitch error, and duration deviation, scores the results by applying weights and deduction rules in a score calculation module, converts the user's cognitive level into step-by-step grades in a cognitive grade conversion module, and provides visual or auditory feedback through a feedback module.

Description

Technology Field

[0001] The present invention relates to an electronic piano-based cognitive ability evaluation system, and more specifically, to an electronic piano-based cognitive ability evaluation system that collects performance information including sound start and end signals, key press intensity, and occurrence time from an electronic piano or MIDI controller in an input module, analyzes a MIDI score file to generate information on the occurrence time, duration, pitch, and velocity of a reference note, synchronizes the two pieces of information in a comparison analysis module to calculate timing error, pitch error, and duration deviation, scores the results by applying weights and deduction rules in a score calculation module, converts the user's cognitive level into step-by-step grades in a cognitive grade conversion module, and provides visual or auditory feedback through a feedback module. Background Technology

[0003] Recently, as the aging of modern society has progressed rapidly, the decline in cognitive function among the elderly has emerged as a social issue; however, current cognitive assessment technologies still rely on traditional question-and-answer tests or paper-based tests, resulting in slow technological progress. In particular, while precise data-based assessments are necessary to detect early cognitive decline, such as dementia or mild cognitive impairment, conventional tests have limitations in terms of objectivity and repeatability due to large variations depending on the examiner's proficiency, the subject's language comprehension, and their education level.

[0004] Conventional computer-based cognitive training programs also largely rely on simple visual stimuli or click responses and fail to adequately reflect the complex cognitive responses that occur in real life. Since these systems do not consider the sensory and motor response characteristics of the elderly, they often underestimate actual cognitive abilities, which consequently makes them difficult to apply to personalized rehabilitation or preventive training. Furthermore, evaluation results are provided only as fragmentary scores, making it difficult to track long-term cognitive trends or provide real-time feedback on improvement effects.

[0005] Meanwhile, although some attempts were made to assess cognitive abilities using musical elements, they were limited to simply distinguishing pitches or matching rhythms, failing to comprehensively analyze multidimensional data such as timing, dynamic control, and accuracy of repetitive performance during the performance process. In particular, for the elderly, despite the fact that complex activities involving the simultaneous interaction of finger movement, auditory response, visual perception, and attention are closely linked to cognitive function, there was a lack of technical means to quantify these aspects in an integrated manner.

[0006] Furthermore, most existing research and equipment are designed for younger generations or laboratory environments, failing to adequately consider the usability of the elderly. Small text, complex operating interfaces, and unclear feedback methods actually induce stress during the testing process and hinder accurate evaluation. Moreover, despite the rapid advancement of cognitive assessment technology in the medical device sector, the field of measuring cognitive function in the elderly remains slow to adopt technology, and a standardized, data-driven evaluation system is currently lacking.

[0007] Therefore, there is a need for a new electronic piano-based cognitive ability assessment system that can efficiently diagnose and improve the cognitive functions of the elderly by providing a simple playing environment while simultaneously utilizing auditory, visual, and motor responses generated during the performance process to quantitatively evaluate complex cognitive abilities such as attention, reaction speed, memory, and consistency, and by providing real-time feedback. Prior art literature

[0009] Patent Document (001) Republic of Korea Registered Patent No. 10-2634347 (Registered on Feb. 1, 2024) The problem to be solved

[0010] The technical problem that the present invention aims to solve is to address the aforementioned problems or necessities by providing an electronic piano-based cognitive ability evaluation system capable of evaluating cognitive abilities based on objective data without being affected by the language ability or educational level of the elderly.

[0011] Another objective of the present invention is to provide an electronic piano-based cognitive ability evaluation system capable of comprehensively analyzing a user's attention, reaction speed, and memory by quantifying timing errors, pitch errors, duration deviations, etc., that occur during the music performance process.

[0012] Another objective of the present invention is to provide an electronic piano-based cognitive ability evaluation system that includes a dynamic time warping (DTW)-based comparative analysis function capable of correcting non-linear tempo changes by synchronizing performance data and reference score data.

[0013] Another objective of the present invention is to provide an electronic piano-based cognitive ability evaluation system that calculates highly reliable score information by automatically applying weighting and deduction rules based on playing accuracy and timing error.

[0014] Another objective of the present invention is to provide an electronic piano-based cognitive ability evaluation system capable of clearly distinguishing the user's cognitive function level by dividing the calculated scores into predefined intervals.

[0015] Another objective of the present invention is to provide an electronic piano-based cognitive ability evaluation system that provides real-time visual or voice feedback, enabling the user to immediately recognize and correct errors during performance.

[0016] Another objective of the present invention is to provide an electronic piano-based cognitive ability evaluation system that supports personalized learning and cognitive training by storing performance data and evaluation results per session and analyzing long-term trends.

[0017] Another objective of the present invention is to provide an electronic piano-based cognitive ability evaluation system that induces an intuitive cognitive response from a user by visually distinguishing between the key currently to be played and the next reserve key by placing a light-emitting means on the keys.

[0018] Another objective of the present invention is to provide an electronic piano-based cognitive ability assessment system capable of early determination of the potential for cognitive decline and providing customized feedback by performing machine learning-based analysis in conjunction with an external server or cloud system.

[0020] Meanwhile, the technical problems to be solved by the present invention are not limited to those mentioned above, and other technical problems not mentioned will be clearly understood by those skilled in the art to which the present invention belongs from the description below. means of solving the problem

[0022] An electronic piano-based cognitive ability evaluation system according to an example of the present invention for realizing the above-described task may first include an input module (100) that collects performance information including sound start and end signals, key press intensity, and occurrence time from an electronic piano or MIDI controller.

[0023] Additionally, it may include a reference score analysis module (200) that parses a MIDI score file to generate information on the occurrence time, duration, pitch, and reference velocity of a reference note.

[0024] Additionally, it may include a comparison analysis module (300) configured to calculate a timing error by aligning performance information (110) from the input module (100) and performance reference information (210) from the reference score analysis module (200) in note units, and configured to apply dynamic time warping to correct non-linear tempo changes.

[0025] Additionally, it may include a score calculation module (400) configured to calculate performance score information (410) by applying item-specific weighting and deduction rules to the comparison output information (310) of the comparison analysis module (300).

[0026] Additionally, it may include a cognitive grade conversion module (500) configured to convert the above performance score information (410) into cognitive grade information (510) of very good, good, average, bad, or very bad according to a predefined interval.

[0027] In addition, it may include a feedback module (600) configured to provide error location, error type, and improvement decline trend in real time in visual or auditory form, together with the above performance score information (410) and the above perception grade information (510).

[0028] Additionally, it may include a storage analysis module (700) equipped to store raw performance data and output indicators, output scores and output grades per session, and to analyze long-term trends to recommend personalized tasks.

[0029] Additionally, it may include an application module (800) configured to visually display the operation results of each module on a display and to interact with the user.

[0030] Additionally, it may include a keyboard part (900) that is controlled to display a light-emitting indicator on a light-emitting means (910) placed on a keyboard based on the analysis result of the above comparison analysis module (300).

[0031] Here, the application module (800) may be characterized by being equipped with a main dashboard screen (810) that displays the user's overall cognitive state and recent evaluation results in the form of a gauge bar and a graph and instructs the start of an evaluation, an evaluation setting screen (820) that selects an evaluation MIDI score and sets the difficulty level, and a performance capture screen (830) that visualizes the keyboard input stream in real time during performance and displays the MIDI signal reception status.

[0032] Additionally, the application module (800) may be characterized by being equipped with an analysis progress screen (840) that displays timing error, pitch error, duration deviation, and dynamic time warping alignment results after the end of the performance, a result summary screen (850) that displays scores and calculated step-by-step recognition grades and displays detailed scores and graphs for each item, and a feedback screen (860) that distinguishes incorrect notes by color and provides feedback on the location of the error, including sheet music scrolling and voice guidance.

[0033] In addition, the application module (800) may be characterized by having a learning report screen (870) that displays the user's evaluation history, score trends, and recommended tasks, and a system settings screen (880) that sets the user profile, language, voice feedback options, and MIDI device connection status.

[0034] Meanwhile, the comparison analysis module (300) may be characterized by being configured to sort the performance standard information (210) in note units to additionally calculate pitch error, duration deviation, and whether there is omission or overperformance.

[0035] At this time, the score calculation module (400) may be characterized by being provided to include a penalty rule that applies a penalty rule according to the timing error (Δt) calculated by the comparison analysis module (300), wherein no penalty is applied when Δt is 100ms or less, 20 points or more are deducted when Δt exceeds 1000ms, and 30 points are deducted when there is no input.

[0036] Additionally, the feedback module (600) may be characterized by being configured to output a feedback message in real time, including whether there is a change and improvement compared to the user's previous performance, along with the performance score information (410) and the perception grade information (510).

[0037] At this time, the cognitive grade conversion module (500) may be characterized by classifying the user's cognitive function level into at least 5 stages based on the performance score information (410) and displaying the cognitive grades classified by stage in the form of colors, graphs, or gauge bars through a visual interface by the application module (800).

[0038] Meanwhile, the above keyboard part (900) may be characterized by being configured to control the light-emitting means (910) with the LED control module (920) based on the analysis result of the above comparison analysis module (300).

[0039] At this time, the LED control module (920) is configured to provide intuitive visual information to the user by differentiating the lighting colors of the LED of the pressing key (921) currently to be played and the LED of the reserve key (922) that is the next target of operation, and to improve the sense of playing by flashing the LED when the player misses the playing timing, and is configured to control the light emission speed and tempo of the pressing key (921) and the reserve key (922) by forming an individual practice scenario sheet that reflects the trend of change in the user's playing skill according to the calculation result of the score calculation module (400) or the cognitive grade conversion module (500).

[0040] Here, the feedback module (600) may be characterized by being configured to transmit the feedback message to an external system (610) via a network, and to store user performance information in conjunction with the external system (610), analyze the user's performance information, and provide feedback.

[0041] At this time, the external system (610) may be characterized by analyzing the user's performance information collected using a machine learning-based algorithm, determining the possibility of cognitive function decline if the user's performance is maintained below a specific standard or decreases rapidly, and displaying this on the learning report screen (870) through the application module (800). Effects of the invention

[0043] If using an electronic piano-based cognitive ability evaluation system according to one embodiment of the present invention,

[0044] First, it has the effect of enabling objective and quantitative evaluation of cognitive ability based on the user's performance data, without being influenced by language proficiency or educational level.

[0045] Second, by analyzing timing errors, pitch errors, and duration deviations that occur during the musical performance process, it is effective in precisely measuring complex cognitive elements such as attention, reaction speed, and memory.

[0046] Third, by applying a dynamic time warping algorithm, it is possible to accurately compare and analyze performances even when non-linear tempo changes occur.

[0047] Fourth, it has the effect of automatically applying weighting and deduction rules to produce a highly reliable score based on the user's playing accuracy.

[0048] Fifth, converting the score results into step-by-step cognitive grades has the effect of enabling users to clearly recognize their cognitive level and intuitively understand the direction for improvement.

[0049] Sixth, by providing real-time visual and auditory feedback, users can immediately recognize errors and simultaneously improve playing accuracy and cognitive response abilities through repetitive learning.

[0050] Seventh, by storing performance data and evaluation results for each session, it is possible to track long-term changes in cognitive function and provide personalized training programs.

[0051] Eighth, by visually distinguishing the current and next keys to be played through a light source mounted on the keyboard, it has the effect of improving the user's concentration and reaction speed.

[0052] Ninth, through machine learning analysis linked with an external server, it automatically analyzes the user's playing patterns and changes in cognitive function, effectively enabling early detection and prevention of potential cognitive decline.

[0054] Meanwhile, the effects obtainable from the present invention are not limited to those mentioned above, and other unmentioned effects will be clearly understood by those skilled in the art to which the present invention belongs from the description below. Brief explanation of the drawing

[0056] The following drawings attached to this specification illustrate a preferred embodiment of the present invention and serve to further enhance understanding of the technical concept of the present invention together with the detailed description of the invention; therefore, the present invention should not be interpreted as being limited only to the matters described in such drawings. FIGS. 1 to 6 are conceptual diagrams illustrating the process flow of input and output of an electronic piano-based cognitive ability evaluation system according to an embodiment of the present invention, and each configuration for determining the decline or improvement of cognitive ability by analyzing collected performance information. FIGS. 7 to 12 are web screen dumps to explain the application screen configuration of an electronic piano-based cognitive ability evaluation system according to one embodiment of the present invention. FIGS. 13 to 15 are conceptual diagrams for explaining the keyboard contents of an electronic piano-based cognitive ability evaluation system according to one embodiment of the present invention. Specific details for implementing the invention

[0057] Specific structural or functional descriptions of embodiments according to the concept of the present invention disclosed herein are provided merely for the purpose of explaining embodiments according to the concept of the present invention, and embodiments according to the concept of the present invention may be implemented in various forms and are not limited to the embodiments described herein.

[0058] Embodiments according to the concept of the present invention may be subject to various modifications and may take various forms; therefore, embodiments are illustrated in the drawings and described in detail in this specification. However, this is not intended to limit the embodiments according to the concept of the present invention to specific disclosed forms, and includes all modifications, equivalents, or substitutions that fall within the spirit and scope of the present invention.

[0059] FIGS. 1 to 6 are conceptual diagrams illustrating the process flow of input and output of an electronic piano-based cognitive ability evaluation system according to an embodiment of the present invention and each configuration for determining the decline or improvement of cognitive ability by analyzing collected performance information; FIGS. 7 to 12 are web screen dumps for explaining the application screen configuration of an electronic piano-based cognitive ability evaluation system according to an embodiment of the present invention; and FIGS. 13 to 15 are conceptual diagrams for explaining the keyboard content of an electronic piano-based cognitive ability evaluation system according to an embodiment of the present invention.

[0061] An electronic piano-based cognitive ability evaluation system according to one embodiment of the present invention includes an input module (100), a standard score analysis module (200), a comparison analysis module (300), a score calculation module (400), a cognitive grade conversion module (500), a feedback module (600), a storage analysis module (700), an application module (800), and a keyboard unit (900).

[0062] First, the input module (100) is configured such that the hardware interface can use MIDI 1.0 serial, USB MIDI, or BLE MIDI to collect performance information (110) including sound start and end signals, key press intensity, and occurrence time from an electronic piano or MIDI controller, and a photocoupler is employed in the receiver for electrical isolation.

[0063] Here, the standard score analysis module (200) parses a MIDI score file and generates the timing of occurrence, duration, pitch, and standard key press intensity information of the standard note as performance standard information (210).

[0064] Additionally, the comparison analysis module (300) is configured to align performance information (110) and performance standard information (210) in note units, apply dynamic time warping to correct non-linear tempo changes, and calculate timing errors. The score calculation module (400) calculates performance score information (410) by applying item-specific weights and deduction rules to the comparison calculation information (310), and the perception grade conversion module (500) is configured to convert the performance score information (410) into stage perception grade information (510) ranging from very good to very bad according to a prior section. The feedback module (600) provides the error location, error type, and improvement / decrease trend in real-time via visual or auditory means along with the score and grade, and the storage analysis module (700) is configured to store raw data, indicators, scores, and grades per session to analyze long-term trends and recommend personalized tasks. Herein, the application module (800) includes a main dashboard screen (810), an evaluation setting screen (820), a performance capture screen (830), an analysis progress screen (840), a result summary screen (850), a feedback screen (860), a learning report screen (870), and a system setting screen (880). The keyboard section (900) includes a light-emitting means (910), an LED control module (920), a push key (921), and a spare key (922). The light-emitting means uses a surface-mount LED, and a diffuser made of PMMA or PC material is applied to improve diffusion. The circuit board may use an FR-4 epoxy glass substrate, the keyboard housing may use ABS, PC, or PBT, and the key tops may use a combination of POM, ABS, and an elastomer coating to improve tactile sensation. It is configured to convert actual performance actions into data, convert them into cognitive indicators, provide immediate feedback, and connect to customized training through long-term accumulation.

[0065] Meanwhile, the comparison analysis module (300) calculates additional pitch error, duration deviation, and whether there is omission or over-performance through alignment with the performance standard information (210). The score calculation module (400) includes a penalty rule based on the timing error Δt, and if Δt is 100ms or less, no penalty is applied; if Δt exceeds 1000ms, 20 points or more are deducted; and if there is no input, 30 points are deducted. 100ms or less is a practical allowable range considering beat units and human perceptual tolerance, and is significant in that it prevents excessive over-judgment. Exceeding 1000ms implies a delay of one or several beats completely out of step, which corresponds to a failure of cognitive response, so it is reasonable to apply a weighted penalty of 20 points or more. The 30-point deduction for no input is a value designed to make the model sensitive to omission detection by reflecting the severity of omission events, which are distinguished from mis-tab or delay. The above figures are valid in general learning scenarios where the tempo range of the song is 60 BPM to 180 BPM, and for special songs with a tempo less than 60 BPM or more than 180 BPM, a correction factor can be applied to maintain the same relative judgment. This configuration is effective in producing highly reliable scores by ensuring consistency and fairness in judgment in various performance environments.

[0066] Here, the feedback module (600) outputs a real-time feedback message by combining performance score information (410) and cognitive grade information (510) with information on changes compared to past sessions. The change information may include a decrease or increase in the average Δt, an improvement or deterioration in the pitch error ratio, and whether the duration deviation has stabilized. The cognitive grade conversion module (500) classifies the user's cognitive function level into at least 5 stages and displays it in the form of colors, graphs, or gauge bars in the application module (800). The minimum of 5 stages is a number of intervals that is neither too dense nor too coarse, satisfying both the user's ease of understanding and discriminability, and is suitable for goal setting and tracking management in clinical or educational settings. This configuration strengthens user motivation and self-directed learning and increases the sensitivity to detect trend changes within the same user.

[0067] Meanwhile, the keyboard section (900) receives the analysis results of the comparative analysis module (300) as input, and the LED control module (920) controls the light-emitting means (910). The LED control module displays the currently playing target key (921) and the next reserve key (922) with different colors or different flashing patterns. For example, the key can be continuously lit, and the reserve key can be flashed at low brightness to enhance intuitiveness. The driving method finely controls the brightness using PWM control, and it is desirable to include overcurrent protection and temperature protection functions in the driver IC. The light-emitting means may use white or multi-color LEDs in 2835 to 3528 packages, and to improve color reproduction and reduce fatigue, a color rendering index of 80 or higher and a flicker index reduction circuit are adopted. The light-diffusing diffuser uses PMMA or PC and considers heat resistance and impact resistance. The rubber dome of the key mechanism uses silicone rubber to reduce noise and standardize tactile sensation. This configuration enhances visual guidance accuracy and intuitively supports reaction speed training, thereby simultaneously improving the reliability of cognitive assessment and training efficiency.

[0068] Additionally, the feedback module (600) transmits feedback messages to the external system (610) via a network, and the external system collects, stores, or analyzes the user's performance information. The external system tracks features between sessions using a machine learning-based algorithm to determine the possibility of cognitive function deterioration when performance skills remain below a specific standard or decline rapidly, and displays warnings and suggested training tasks on the learning report screen (870) through the application module (800). It is desirable for the external system to include encryption during the transmission phase and de-identification processing during the storage phase to protect privacy. The cloud infrastructure may use container orchestration for availability and scalability, and may apply edge caching to minimize latency. This configuration has the effect of utilizing collective intelligence beyond individual device resources to detect early signs of anomalies and automate customized interventions.

[0069] Here, the main board is manufactured using an FR-4-based 4 to 6-layer PCB, and grounding and shielding treatments are applied to reduce EMI. The keybed frame may utilize an aluminum alloy 6xxx series or a high-rigidity ABS housing. Considering the characteristics of long-term continuous use, it is desirable to use a copper core substrate or thermal pads for the light-emitting means to dissipate heat. For the exterior housing, UL94 V-0 grade PC or PC-ABS alloy resin may be selected for environments requiring a flame-retardant rating. The front panel display is protected by tempered glass or a polycarbonate window, and an anti-reflective coating is applied to reduce glare. The speaker for voice feedback uses a neodymium magnet full-range unit and is configured to have an internal EVA or silicone damper attached to suppress vibration.

[0070] Meanwhile, the no-deduction for Δt ≤ 100ms is an allowable range that aligns with the division of beats and human perception thresholds within the tempo range of 60 BPM to 180 BPM, thereby reducing unnecessary judgments, while the deduction of 20 points or more for Δt > 1000ms and the deduction of 30 points for no input ensure the discriminability of judgments through weighted penalties proportional to the severity of the event. A minimum of 5 steps simultaneously satisfies the intuitiveness of grade interpretation and ease of management, making it suitable for setting goals in clinical or educational settings.

[0071] As described above, the present invention converts time-series data acquired from actual instrument performance into multidimensional indicators and corrects non-linear elements through dynamic time warping to consistently provide scores, grades, and feedback, and realizes a virtuous cycle of training and evaluation through visual-guided keyboard display and integration with external systems.

[0072] Here, delay correction calibration is configured to provide a calibration routine that measures device or network delay and is configured to remove the offset before Δt calculation.

[0073] In addition, a reinforcement learning-based model can be configured to learn a difficulty function from a performance history to generate tempo, note density, or rhythmic complexity, recognize the actual keybed with a tablet camera, and display the pressed keys and spare keys superimposed on the screen as AR.

[0074] Meanwhile, for users with color vision deficiencies, the configuration may be modified to provide equivalent guidance quality by enhancing color contrast and adding patterns, vibrations, or voice cues.

[0075] In addition, it is configured to learn the user's unique performance signature using a session-to-session hidden Markov model or a variational autoencoder to detect anomalies based on performance patterns, and to detect abrupt deviations as anomalies.

[0077] The present invention has been described above through a preferred embodiment. However, the above-described embodiment is merely an illustrative explanation of the technical concept of the present invention, and those skilled in the art will understand that various changes are possible within the scope of the technical concept of the present invention. Therefore, the scope of protection of the present invention should be interpreted by the matters described in the patent claims rather than by specific embodiments, and all technical concepts within the corresponding scope should also be interpreted as being included within the scope of rights of the present invention. Explanation of the symbols

[0079] 100 ... input module 110 ... Performance Information 200 ... Standard Score Analysis Module 210 ... Performance Standard Information 300 ... Comparative Analysis Module 310 ... Comparative Output Information 400 ... Score Calculation Module 410 ... Performance score information 500 ... Cognitive Grade Conversion Module 510 ... Cognitive rating information 600 ... feedback module 610 ... external system 700 ... Storage Analysis Module 800 ... application module 810 ... Main Dashboard Screen 820 ... Evaluation Settings Screen 830 ... Performance capture screen 840 ... Analysis progress screen 850 ... Result Summary Screen 860 ... Feedback screen 870 ... Learning Report Screen 880 ... System Settings Screen 900 ... keyboard section 910 ... means of light 920 ... LED control module 921 ... push keys 922 ... spare keyboard

Claims

Claim 1 An input module (100) that collects performance information including start and end signals of a sound, key press intensity, and occurrence time from an electronic piano or MIDI controller; a reference score analysis module (200) that parses a MIDI score file to generate information on the occurrence time, duration, pitch, and reference velocity of a reference note; a comparison analysis module (300) that is configured to calculate a timing error by aligning performance information (110) from the input module (100) and performance reference information (210) from the reference score analysis module (200) in note units, and is configured to apply dynamic time warping to correct non-linear tempo changes; a score calculation module (400) that is configured to calculate performance score information (410) by applying item-specific weighting and deduction rules to the comparison calculation information (310) of the comparison analysis module (300); and a performance score information (410) that is classified into stages of very good, good, average, bad, or very bad according to a predefined interval. A cognitive grade conversion module (500) configured to convert into cognitive grade information (510); a feedback module (600) configured to provide error location, error type, and improvement / decrease trend in real time via visual or auditory means together with the performance score information (410) and the cognitive grade information (510); a storage analysis module (700) configured to store raw performance data, output indicators, output scores, and output grades per session, analyze long-term trends, and recommend personalized tasks; an application module (800) configured to visually display the operation results of each module on a display and to interact with the user; and a keyboard unit (900) controlled to display light on a light-emitting means (910) placed on the keyboard based on the analysis results of the comparison analysis module (300).The comparison analysis module (300) is configured to sort the performance standard information (210) in note units to additionally calculate pitch error, duration deviation, and whether there is omission or over-performance, and the score calculation module (400) is configured to apply a deduction rule according to the timing error (Δt) calculated by the comparison analysis module (300), wherein no deduction is made if Δt is 100ms or less, 20 points or more are deducted if Δt exceeds 1000ms, and 30 points are deducted if there is no input, and the feedback module (600) is configured to output a feedback message in real time including whether there is a change and improvement compared to the user's previous performance, together with the performance score information (410) and the cognitive grade information (510), and the cognitive grade conversion module (500) classifies the user's cognitive function level into at least 5 stages based on the performance score information (410). The application module (800) is configured to display the cognitive grade, which is divided into stages, in the form of a color, graph, or gauge bar through a visual interface. The application module (800) includes: a main dashboard screen (810) that displays the user's overall cognitive state and recent evaluation results in the form of a gauge bar or graph and instructs the start of the evaluation; an evaluation setting screen (820) for selecting a MIDI score for evaluation and setting the difficulty level; a performance capture screen (830) that visualizes the keyboard input stream in real-time during performance and displays the MIDI signal reception status; an analysis progress screen (840) that displays timing error, pitch error, duration deviation, and dynamic time warping alignment results after the performance ends; a result summary screen (850) that displays the score and the calculated cognitive grade for each stage, and displays detailed scores and graphs for each item; a feedback screen (860) that distinguishes incorrect notes by color and provides feedback on the location of the error, including score scrolling and voice guidance; and a learning report screen (870) that displays the user's evaluation history, score trends, and recommended tasks. and system settings screen (880) for setting user profile, language, voice feedback options and MIDI device connection status;The comparison analysis module (300) is configured to perform delay correction calibration to measure device or network delay occurring in the electronic piano, MIDI controller, application module (800), or network environment before calculating the timing error (Δt), and to remove the delay offset calculated by the delay correction calibration before calculating the timing error (Δt); the storage analysis module (700) is configured to store the user's performance history including raw performance data per session, calculation indicators, calculation scores, and calculation grades, and to enable a reinforcement learning-based model to learn an individual difficulty function based on the performance history; the application module (800) or the keyboard unit (900) is configured to adjust or generate the tempo, note density, or rhythm complexity of the sheet music used for evaluation or training according to the individual difficulty function learned by the reinforcement learning-based model; and the application module (800) recognizes the keybed of the actual electronic piano or MIDI controller with a tablet camera and the currently to be played An electronic piano-based cognitive ability evaluation system characterized by being configured to superimpose the positions of a pressed key (921) and a reserve key (922), which is the next action target, on a screen in an augmented reality manner, wherein the storage analysis module (700) or the feedback module (600) learns a user's unique performance signature using a hidden Markov model or a variational autoencoder based on the raw performance data and output indicators for each session, and detects an abnormal state when the performance pattern of a subsequent session deviates abruptly from the user's unique performance signature, and provides a state of potential cognitive function deterioration or the need for additional verification through the feedback module (600). Claim 2 delete Claim 3 delete Claim 4 In claim 1, the keyboard section (900) is configured to control the light-emitting means (910) to the LED control module (920) based on the analysis result of the comparison analysis module (300), wherein the LED control module (920) is configured to provide intuitive visual information to the user by differentiating the lighting color of the LED of the press key (921) currently to be played and the LED of the reserve key (922) that is the next operation target, and to improve the sense of playing by flashing the LED when the player misses the playing timing, and is configured to form an individual practice scenario sheet reflecting the user's playing skill change trend according to the calculation result of the score calculation module (400) or the cognitive grade conversion module (500), and to adjust the light-emitting speed and tempo of the press key (921) and the reserve key (922). Claim 5 In claim 4, the feedback module (600) is configured to transmit the feedback message to an external system (610) via a network, and is configured to store user performance information in conjunction with the external system (610), analyze the user's performance information, and provide feedback. The external system (610) analyzes the collected user's performance information using a machine learning-based algorithm, and if the user's performance is maintained below a specific standard or decreases rapidly, it determines the possibility of cognitive function decline and displays it on the learning report screen (870) through the application module (800). This characterizes an electronic piano-based cognitive ability evaluation system.

Images