Objective quantitative evaluation method for relieving psychological stress by piano playing

Abstract

The invention discloses an objective quantitative evaluation method for relieving psychological stress in piano playing, which comprises the following steps of: acquiring pulse wave signals of relaxation and stress states, extracting physiological characteristics, and classifying the two states based on the physiological characteristics; then, recording the playing moment, the playing key number and the playing duration during piano playing, collecting pulse wave signals after piano playing, and extracting physiological features; and finally, calculating rhythm accuracy and note accuracy in piano playing, and quantitatively evaluating the effect of piano playing on mental stress relief in combination with the synchronous change of the playing indexes and the physiological indexes. By quantitatively evaluating the psychological stress relieving effect of piano playing, a personalized music healing strategy can be formulated.

Description

Technical Field

[0001] The present invention relates to the technical field of signal processing, and in particular to an objective quantitative evaluation method for piano playing to alleviate psychological stress. Background Art

[0002] With the rapid development of modern society, more and more people are facing pressure from all aspects of society. While moderate stress can keep people motivated and help them perform better, excessive stress can threaten and damage a person's physical and mental health. Psychological stress is becoming a major health threat in modern society, with the number of people experiencing psychological and physical problems due to stress increasing year by year. Excessive stress often causes physical or psychological discomfort and may also trigger a variety of diseases, such as diabetes, cardiovascular disease, depression, and cancer. Therefore, it is essential to understand psychological stress and choose appropriate methods to reduce it.

[0003] Music therapy is a common method for alleviating psychological stress. For decades, music has been used as a stress-reducing intervention. Music therapy can be categorized into passive and active music therapy based on the participant's approach. In passive music therapy, participants simply "receive" the music, such as by listening to it. In active music therapy, participants directly participate in the process of "making music," such as playing an instrument, singing, or improvising. Piano playing is a form of active music therapy and an effective means of relieving psychological stress.

[0004] Methods for measuring psychological stress can be broadly categorized into subjective assessment and objective measurement. Subjective assessment relies on self-reports, such as questionnaires and mood ratings. However, this method is subject to potential subjective bias and individual differences, and may not fully and accurately reflect the true effects of music therapy. Objective measurement involves detecting physiological signals or indicators, such as cortisol levels, EEG signals, and pulse wave signals. Therefore, to scientifically and systematically validate the stress-reducing effects of music therapy, particularly piano playing, objective quantitative evaluation is crucial. Summary of the Invention

[0005] The purpose of the present invention is to solve the defects in the prior art and provide an objective quantitative evaluation method for piano playing to reduce psychological stress.

[0006] The purpose of the present invention can be achieved by taking the following technical solutions:

[0007] An objective quantitative evaluation method for piano playing to reduce psychological stress comprises the following steps:

[0008] S1. Collect pulse wave signals of subjects in a relaxed state and a stressed state and extract physiological features; the physiological features include heart rate variability features and pulse wave signal morphology features; calculate the importance weight of each physiological feature using random forest;

[0009] S2. The subject plays pieces of music of varying difficulty on the piano, records the playing time, key number, and playing duration, collects the subject's pulse wave signal after the performance, and extracts the physiological characteristics;

[0010] S3, calculating the subject's performance index during the piano playing process based on the playing time, the playing key number, and the playing duration, wherein the performance index includes rhythm accuracy and note accuracy;

[0011] S4, grouping the performance indicators and the physiological characteristics extracted in step S2 according to the difficulty of the music played in step S2, and calculating weighted values of the music difficulty and the performance indicators in reducing the psychological stress of the subject through regression analysis;

[0012] S5. Perform dimensionless processing on the physiological characteristics in step S1 and perform weighted value calculation according to the importance weight, and then add the weighted value calculated in step S4 to obtain a quantitative evaluation result of the effect of piano playing on reducing the psychological stress of the subject.

[0013] Furthermore, in step S1, the relaxed state refers to the baseline physiological state of the individual without task execution before the stress-induced test; and the stressed state refers to the stress response state exhibited by the individual after the stress-induced test, which is specifically manifested as significant changes in HRV characteristics and pulse wave signal morphological characteristics, wherein the HRV characteristics are the heart rate variability characteristics; wherein the stress-induced test is induced by completing the MIST mental arithmetic test;

[0014] The process of step S1 is as follows:

[0015] S101, guiding the subject to sit quietly for 5 minutes to enter a relaxed state, and collecting their pulse wave signal; then, guiding the user to complete the MIST mental arithmetic test stress induction task to induce psychological stress, and collecting the subject's pulse wave signal immediately after the task is completed;

[0016] S102, preprocessing the collected pulse wave signal, including denoising and removing baseline drift, to eliminate interference components in the pulse wave signal; then, extracting HRV features and pulse wave signal morphological features from the preprocessed pulse wave signal, where the HRV features and pulse wave signal morphological features are both the physiological features;

[0017] S103. Evaluate the effectiveness of physiological characteristics in stress detection by performing significance analysis on physiological characteristics in relaxation and stress states;

[0018] S104. Using a random forest algorithm, a binary classification is performed based on physiological characteristics in a relaxed state and a stressed state, and a psychological stress detection model is constructed;

[0019] S105. Perform feature selection, randomly construct feature subsets and score them, and finally obtain the feature subset with the best classification effect, and calculate the importance weight of each physiological feature through random forest.

[0020] Furthermore, in step S2, recording the playing time and the playing key number is completed by an electronic piano device; the electronic piano device is composed of an electronic piano and a computer system, and can collect playing data in real time; during the playing process, the electronic piano device can determine in real time whether the key is in a played state, obtain the playing key number of the played key, and calculate the playing time and playing duration of each note, and store the playing key number, the playing time and the playing duration in the computer.

[0021] Furthermore, the process of step S2 is as follows:

[0022] S201, displaying the songs played by the subject on a monitor, and classifying the songs into three difficulty levels: low, medium, and high according to the number of notes and rhythmic complexity of the songs;

[0023] S202, after the performance, obtaining the subject's playing key number, playing time, and playing duration;

[0024] S203 , collecting the pulse wave signal of the subject after playing the piano, and processing and extracting features of the pulse wave signal according to the step of S102 .

[0025] Furthermore, the performance indicators in step S3 include: rhythm accuracy and note accuracy; rhythm accuracy is defined as the root mean square error between the actual rhythm and the standard rhythm, recorded as RMSE; note accuracy is defined as the percentage of correct key presses to the total number of key presses;

[0026] The step S3 is as follows:

[0027] S301, after the subject finishes playing, the key number, playing time, and playing duration of the subject's performance are recorded to form a note sequence played by the subject, and the played note sequence is aligned with the note sequence of the piece;

[0028] S302. Calculate the number of all correct notes in the note sequence played by the subject, and calculate the percentage of the number of correct notes to the total number of notes played, as the note accuracy of the performance; calculate the deviation of the actual start time and actual duration of each note from the standard start time and standard duration, and calculate the RMSE, and standardize the RMSE to a range of 0 to 1, as the rhythm accuracy of the performance.

[0029] The present invention has the following advantages and effects compared to the prior art:

[0030] (1) The present invention verifies the changes in psychological stress levels through the classification of HRV characteristics (i.e., the heart rate variability characteristics) and pulse wave signal morphological characteristics. Combined with the pulse wave signal collection before and after piano playing, it can objectively quantify the effect of piano playing on relieving psychological stress, provide a scientific basis for psychological stress research, and solve the limitation of traditional psychological stress assessment relying on subjective questionnaires.

[0031] (2) The present invention collects pulse wave signals, extracts physiological characteristics such as HRV characteristics and pulse wave signal morphological characteristics, analyzes the changes in physiological characteristics before and after piano playing through correlation and significance tests, and objectively evaluates the effect of piano playing on alleviating individual psychological stress. It can comprehensively evaluate the effect of piano playing on relieving psychological stress, and solves the problem of lack of objective physiological indicators in traditional music stress relief research.

[0032] (3) The present invention removes the dimension of the features by dimensionless transformation and performs weighted summation based on the importance weights of each feature to obtain a quantitative scoring result of piano playing on reducing psychological stress. The quantitative scoring can help people establish an accurate and intuitive understanding of the stress-relieving effect of piano playing as a music therapy, help people make more scientific decisions, and solve the problem that traditional psychological stress assessment methods are difficult to quantify and accurately analyze. BRIEF DESCRIPTION OF THE DRAWINGS

[0033] The drawings described herein are used to provide a further understanding of the present invention and constitute a part of this application. The exemplary embodiments of the present invention and their descriptions are used to explain the present invention and do not constitute an improper limitation of the present invention. In the drawings:

[0034] Figure 1 is a flow chart of the objective quantitative evaluation method for piano playing to reduce psychological stress disclosed in an embodiment of the present invention;

[0035] Figure 2 This is an experimental flow chart in an embodiment of the present invention;

[0036] Figure 3 is a flow chart of pulse wave signal preprocessing in an embodiment of the present invention;

[0037] Figure 4 1 is a waveform diagram of the pulse wave after preprocessing in an embodiment of the present invention;

[0038] Figure 5 Schematic diagram of pulse wave signal feature point detection;

[0039] Figure 6 Schematic diagram of a music score displayed on a display according to an embodiment of the present invention; DETAILED DESCRIPTION

[0040] To make the objectives, technical solutions, and advantages of the embodiments of the present invention more clear, the technical solutions in the embodiments of the present invention will be clearly and completely described below in conjunction with the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making creative efforts shall fall within the scope of protection of the present invention.

[0041] An objective quantitative evaluation method for piano playing to reduce psychological stress comprises the following steps:

[0042] S1. Collect pulse wave signals from subjects in both relaxed and stressed states and extract physiological features; the physiological features include heart rate variability (HRV) features and pulse wave signal morphology features; then classify the two states based on the physiological features, and calculate the importance weight of each feature using random forest analysis;

[0043] S2: The subjects played piano pieces of varying difficulty, and recorded the playing time, key number, and duration of their performances. Then, pulse wave signals were collected after the performances to extract physiological features.

[0044] S3, calculating two performance indices, beat stability and note accuracy, in piano performance based on playing time, playing key number, and playing duration, where the performance indices reflect the performance;

[0045] S4. Group the performance indicators and physiological characteristics extracted in S2 by repertoire difficulty, calculate the correlation between each group's performance and changes in physiological characteristics before and after performance, analyze the impact of performance and repertoire difficulty on stress relief, and calculate the weights and weighted values of the impact of repertoire difficulty and performance indicators on stress relief through regression analysis;

[0046] S5. Dimensionally transform the physiological characteristics and perform feature weighting, and add them to the weighted values of the difficulty of the repertoire and the performance index obtained in S4 to quantitatively evaluate the effect of piano playing on reducing psychological stress.

[0047] Furthermore, the process of step S1 is as follows:

[0048] S101: Instruct the subject to sit quietly for 5 minutes to relax, and collect their baseline pulse wave signal. Then, guide the subject to complete the MIST mental arithmetic test stress induction task to induce psychological stress, and collect the user's pulse wave signal immediately after the task.

[0049] S102: Preprocess the collected pulse wave signal, and then extract HRV time-frequency domain features and pulse wave morphology features from the preprocessed signal. HRV features and pulse wave morphology features are collectively referred to as physiological features. The specific steps are as follows:

[0050] S102.1. To remove high-frequency interference, such as power frequency noise and myoelectric noise, from the collected raw pulse wave signal, a Butterworth filter may be used to perform low-pass filtering on the signal.

[0051] S102.2. Remove baseline drift interference from the pulse signal by performing cubic sample interpolation on the low-pass filtered signal;

[0052] S102.3. Extracting HRV features from the processed pulse wave signal;

[0053] S102.4. Extracting pulse wave morphological features from the processed pulse wave signal;

[0054] S103. Evaluate the effectiveness of physiological characteristics in stress detection by performing significance analysis on physiological characteristics in relaxation and stress states;

[0055] S104: Use the random forest algorithm to classify relaxation and stress states based on physiological characteristics and build a psychological stress detection model. The specific steps are as follows:

[0056] S104.1. Since each individual's physiological characteristics are different, individual physiological characteristics should be standardized to ensure that all characteristics are on the same scale;

[0057] S104.2. Randomly divide the data into a training set (80%) and a test set (20%) according to the samples;

[0058] S104.3. Using the Random Forest (RF) algorithm to detect stress status;

[0059] S104.4. Optimize the hyperparameters of each classifier using a 5-fold cross-validation grid search algorithm to obtain the optimal model parameters so that the classifier can achieve relatively good classification and detection results.

[0060] S105: Perform feature selection, randomly construct feature subsets and score them, and finally obtain the feature subset with the best classification effect. Then calculate the importance weight of each feature through random forest. The specific steps are as follows:

[0061] S105.1. Perform feature selection using recursive feature elimination. Use 5-fold cross-validation to score each feature subset, using average classification accuracy as the scoring metric. Remove the features that contribute least to classification accuracy one by one until the last feature remains. Record the optimal feature subsets for each number of features retained. Obtain the relationship between the number of features in each feature subset and the estimated classification accuracy. Finally, determine the feature subset that achieves the best classification results.

[0062] S105.2. By removing different features and recalculating the scores, the importance weight of each feature in the optimal feature subset can be calculated using random forest based on the changes in the scores.

[0063] Furthermore, in step S2, the recording of the playing time and the playing key number is completed by an electronic piano device in the laboratory. The device consists of an electronic piano and a supporting computer system, which can collect playing data in real time. During the playing process, the system can determine in real time whether the key is being played, obtain the key number of the played key (i.e., the playing key number), accurately calculate the start time and playing duration of each note, and store various playing information in the computer.

[0064] Furthermore, the process of step S2 is as follows:

[0065] S201, classifying the music to be played into three difficulty levels: low, medium, and high, based on the number of notes and rhythmic complexity of the music, displaying the music to be played by the user on a display, and guiding the user to complete the performance according to the music score;

[0066] The difficulty of a piece of music is measured by a difficulty score, which is calculated from the two dimensions of notes and rhythm. The definition and calculation method of each dimension are as follows:

[0067] 1) Note score: It is composed of note diversity and note span. The note diversity and note span are added together to get the note score.

[0068] Note diversity is used to measure the richness of different note types in a track. Continuous repetition of notes will reduce diversity. The calculation method is as follows:

[0069]

[0070] Among them, unique note types refers to the number of different pitches (e.g., C4, D#5) in the piece, and maximum consecutive repetitions refers to the maximum number of times the same note appears consecutively;

[0071] Note span is used to measure the interval span between adjacent notes. The larger the span, the higher the difficulty. The specific calculation method is as follows:

[0072]

[0073] Among them, the average interval is the average of the interval spans between adjacent notes, the maximum interval is the total span of the repertoire, and the maximum span is the number of note intervals exceeding one octave;

[0074] 2) Rhythm score: used to measure the complexity of rhythm changes in a piece of music. The rhythm of a piano piece is expressed by note value, beat and tempo. Note value defines the duration of each note. Common note types include whole notes (with a duration of 4 beats), quarter notes (with a duration of 1 beat), eighth notes (with a duration of 0.5 beats), etc.; while beat specifies the number of beats per measure, usually written at the beginning of the score in the form of a fraction (such as 4 / 4, 3 / 4); tempo defines the speed of the piece, usually expressed in beats per minute (BPM), which determines the actual duration of each note. The specific calculation method of the rhythm score is as follows:

[0075]

[0076] The number of rhythm changes is the number of times different note types (such as quarter notes, eighth notes, etc.) in the track are switched, and the benchmark BPM is the average BPM of all tracks in the track library.

[0077] The difficulty score of the piece is calculated by adding the note score and rhythm score and normalizing them, which ranges from [0,1].

[0078] According to the difficulty score, the songs are divided into three levels: low, medium and high: songs with a difficulty score range of 0 to 0.4 are low difficulty; songs with a score range of 0.4 to 0.7 are medium difficulty; and songs with a score range of 0.7 to 1 are high difficulty.

[0079] S202, obtaining the key number of the piano played by the subject (i.e., the playing key number), the playing time, and the playing duration;

[0080] S203 , collecting the pulse wave signal of the subject after playing the piano, and processing and extracting features of the pulse wave signal according to the step of S102 .

[0081] Furthermore, the performance metrics in step S3 include rhythm accuracy and note accuracy. Rhythm accuracy is defined as the root mean square error (RMSE) between the actual rhythm and the standard rhythm, while note accuracy is defined as the percentage of correct key presses to the total number of key presses. Together, these two metrics reflect the performance of a performance.

[0082] Furthermore, the step S3 is as follows:

[0083] S301. After the subject finishes playing, the key numbers and playing times of the piano keys played by the subject are recorded to form a note sequence played by the user, and the played note sequence is aligned with the note sequence of the music piece.

[0084] S302. Count all correct notes in the note sequence played by the subject and calculate the percentage of correct notes to the total number of notes played as the note accuracy for that performance; calculate the deviation of the actual start time and duration of each note from the standard start time and duration, and calculate the RMSE. The RMSE is normalized to a range of 0 to 1 and used as the rhythm accuracy for that performance. Values closer to 1 indicate more accurate rhythms, and values closer to 0 indicate greater deviations. The performance accuracy is calculated by combining the note accuracy and rhythm accuracy.

[0085] Furthermore, the step S4 is as follows:

[0086] S401: Group the performance indicators and the physiological characteristics extracted in S203 by the difficulty of the repertoire. Analyze the impact of performance at different difficulty levels on stress relief by calculating the correlation between performance at different difficulty levels and changes in physiological characteristics before and after the performance. Then, analyze the impact of performance and repertoire difficulty on stress relief. Calculate the influence weights of performance and repertoire difficulty through regression analysis to quantify the impact of performance and repertoire difficulty on stress relief. The specific steps are as follows:

[0087] S401.1. Group the performance indicators and the physiological characteristics extracted in S203 according to the difficulty of the music;

[0088] S401.2. For each group, calculate the Spearman rank correlation coefficient between the performance index and the changes in physiological characteristics before and after performance. The calculated correlation coefficient shows that the greater the difficulty, the smaller the correlation between the performance index and the changes in physiological characteristics before and after performance, indicating that the greater the difficulty, the worse the stress relief effect.

[0089] S401.3. To ensure the reliability of the correlation analysis results, the Kruskal-Wallis H test was performed to analyze whether there were significant differences in the changes in physiological characteristics after stress relief in each group, further verifying the conclusions of S401.2.

[0090] S401.4. Calculate the influence weights of performance indicators and piece difficulty scores on stress reduction through Bayesian regression fitting;

[0091] S402: Multiply the performance index and the piece difficulty score by their corresponding influence weights to obtain corresponding weighted values.

[0092] Furthermore, the step S5 is as follows:

[0093] S501, calculating the importance weight of each feature of the optimal feature subset obtained in S1, and removing the dimension of each feature of the feature subset by dimensionless conversion;

[0094] S502: Multiply each physiological feature by its importance weight to obtain a weighted value of each human body feature;

[0095] S503: Add the weighted values of the physiological characteristics and the weighted values of the performance indicators obtained in step S402 to obtain a quantitative evaluation score of the effect of piano performance on alleviating psychological stress.

[0096] Example 1

[0097] This embodiment discloses an objective quantitative evaluation method for piano playing to reduce psychological stress, such as Figure 1 The specific steps are as follows:

[0098] S1. Collect pulse wave signals from subjects in both relaxed and stressed states and extract physiological features. Physiological features include heart rate variability (HRV) features and pulse wave signal morphology features. Then, classify the two states based on the physiological features and calculate the importance weight of each feature using a random forest. The specific steps are as follows:

[0099] S101: Guide the subject to sit quietly for 5 minutes to relax, and collect their pulse wave signals. Then, guide the user to complete the MIST mental arithmetic test stress induction task to induce psychological stress, and collect the subject's pulse wave signals immediately after the task is completed. Figure 2 The overall experimental flow chart is presented;

[0100] S102, pre-processing the collected pulse wave signal, and then extracting HRV time-frequency domain features and pulse wave morphology features from the pre-processed signal. HRV features and pulse wave morphology features are collectively referred to as physiological features. Figure 3 The flowchart of pulse wave signal processing and feature extraction is shown. The specific steps are as follows:

[0101] S102.1. To remove high-frequency interference, such as power frequency noise and myoelectric noise, from the collected raw pulse wave signal, a Butterworth filter may be used to perform low-pass filtering on the signal.

[0102] S102.2. Find the starting point of the pulse wave cycle by the point of maximum slope of the pulse wave signal. Determine the starting point of each pulse wave cycle by searching for the zero point of the first-order difference based on the maximum slope point of each cycle.

[0103] S102.3. After determining the cycle start point, remove baseline drift interference from the pulse signal by performing cubic sample interpolation.

[0104] S102.4. Calculate the Pearson correlation coefficient between adjacent cycles, filter the pulse wave signals based on the correlation coefficient, and eliminate unqualified pulse wave cycles. Figure 4 The waveform of the pulse wave signal after preprocessing is shown;

[0105] S102.5. Extracting HRV-related time domain and frequency domain features from the preprocessed pulse wave signal;

[0106] S102.6. Identify the pulse wave signal feature points on a cycle-by-cycle basis for the preprocessed pulse wave signal, and extract the morphological features of each cycle of the pulse wave signal based on this. Figure 5 A schematic diagram showing the results of pulse wave signal feature point detection is shown; the pulse wave signal feature points include the pulse starting point, main wave peak point, and dicrotic wave peak point of the pulse wave signal; the morphological features include main wave amplitude, dicrotic wave amplitude, descending branch time, descending branch area, ascending branch time, ascending branch area, reflection index, reflection time, and pulse width;

[0107] S102.7. Extract three time domain features, namely, mean AVE, variance VAR, and root mean square difference RMSSD, from the extracted pulse wave morphology feature sequence;

[0108] S103. Evaluate the effectiveness of HRV characteristics in stress detection by calculating the correlation between changes in HRV characteristics and pulse wave morphology characteristics and changes in psychological stress levels.

[0109] S104. Use the random forest algorithm to perform binary classification based on the physiological characteristics of the relaxed state and the stressed state, and build a psychological stress detection model. The specific steps are as follows:

[0110] S104.1. Since each individual's physiological characteristics are different, individual physiological characteristics should be standardized to ensure that all characteristics are on the same scale;

[0111] S104.2. Randomly divide the data into a training set (80%) and a test set (20%) according to the samples;

[0112] S104.3. Use the Random Forest (RF) algorithm to detect stress status;

[0113] S104.4. Use a 5-fold cross-validation grid search algorithm to optimize the classifier's hyperparameters to obtain the optimal model parameters so that the classifier can achieve relatively good classification and detection results.

[0114] S105: Perform feature selection, randomly construct feature subsets and score them, and finally obtain the feature subset with the best classification effect. Then calculate the importance weight of each feature through random forest. The specific steps are as follows:

[0115] S105.1. Perform feature selection using recursive feature elimination. Use 5-fold cross-validation to score each feature subset, using average classification accuracy as the scoring metric. Remove the features that contribute least to classification accuracy one by one until the last feature remains. Record the optimal feature subsets for different numbers of features retained. Obtain the relationship between the number of features in each feature subset and the evaluated classification accuracy. Finally, determine the feature subset that achieves the best classification results.

[0116] S105.2. By removing different features and recalculating the scores, the importance weights of each feature in the optimal feature subset that can be calculated using random forest are as follows based on the changes in the scores:

[0117] Table 1. Feature weights

[0118]

[0119] S2. The subject plays piano pieces of varying difficulty and records the playing time, key number, and duration. Then, pulse wave signals are collected after the performance to extract physiological features. The specific steps are as follows:

[0120] S201, classifying the repertoire into three levels of difficulty: low, medium, and high, based on the complexity of the notes and rhythm of the repertoire, displaying the repertoire to be performed by the subject on a monitor, and guiding the subject to complete the performance according to the score;

[0121] like Figure 6 The figure shows a schematic diagram of a performance piece on a display. The horizontal axis is time. The leftmost vertical line in the figure is the starting point, and the time is recorded as 0. Each note is displayed from left to right according to the order in which it is played. The leftmost vertical line in the figure slides from left to right to indicate the starting time and playing duration of each note, guiding the user to complete the performance according to the music score.

[0122] After each stress induction, a piece of music with a certain difficulty level (low, medium, or high) was randomly selected and the subjects were guided to play it.

[0123] The difficulty of a piece of music is measured by a difficulty score, which is calculated from the two dimensions of notes and rhythm. The definition and calculation method of each dimension are as follows:

[0124] 1) Note score: It consists of note diversity and note span.

[0125] Note diversity is used to measure the richness of different note types in a track. Continuous repetition of notes will reduce diversity. The calculation method is as follows:

[0126]

[0127] Among them, unique note types refers to the number of different pitches (e.g., C4, D#5) in the piece, and maximum consecutive repetitions refers to the maximum number of times the same note appears consecutively;

[0128] Note span is used to measure the interval span between adjacent notes. The larger the span, the higher the difficulty. The specific calculation method is as follows:

[0129]

[0130] Among them, the average interval is the average of the interval spans between adjacent notes, the maximum interval is the total span of the repertoire, and the maximum span is the number of note intervals exceeding one octave;

[0131] Note diversity and note span are added together to get the note score.

[0132] 2) Rhythm score: used to measure the complexity of rhythm changes in a piece of music. The rhythm of a piano piece is expressed by note value, beat and tempo. Note value defines the duration of each note. Common note types include whole notes (with a duration of 4 beats), quarter notes (with a duration of 1 beat), eighth notes (with a duration of 0.5 beats), etc.; while beat specifies the number of beats per measure, usually written at the beginning of the score in the form of a fraction (such as 4 / 4, 3 / 4); tempo defines the speed of the piece, usually expressed in beats per minute (BPM), which determines the actual duration of each note. The specific calculation method of the rhythm score is as follows:

[0133]

[0134] The number of rhythm changes is the number of times different note types (such as quarter notes, eighth notes, etc.) in the track are switched, and the benchmark BPM is the average BPM of all tracks in the track library.

[0135] The difficulty score of the piece is calculated by adding the note score and rhythm score and normalizing them, which ranges from [0,1].

[0136] According to the difficulty score, the songs are divided into three levels: low, medium and high: songs with a difficulty score range of 0 to 0.4 are low difficulty; songs with a score range of 0.4 to 0.7 are medium difficulty; and songs with a score range of 0.7 to 1 are high difficulty.

[0137] S202, after the performance, obtaining the key number, playing time, and playing duration of the piano played by the subject;

[0138] S203 , collecting the pulse wave signal of the subject after playing the piano, and processing and extracting features of the pulse wave signal according to the step of S102 to extract physiological features.

[0139] S3. Calculate two performance indicators, rhythm accuracy and note accuracy, in piano performance based on the playing time, the playing key number, and the playing duration. The performance indicators reflect the performance. The specific steps are as follows:

[0140] S301. After the subject finishes playing, the key numbers and playing times of the piano keys played by the subject are recorded to form a note sequence played by the subject, and the played note sequence is aligned with the note sequence of the piece.

[0141] S302. Count the number of all correct notes in the note sequence played by the subject, and calculate the percentage of the number of correct notes to the total number of notes played as the note accuracy of the performance; calculate the deviation between the actual start time and duration of each note and the standard start time and duration, and calculate RMSE. Normalize RMSE to a range of 0 to 1 and use it as the rhythm accuracy of the performance. The closer the value is to 1, the more accurate the rhythm; the closer the value is to 0, the greater the deviation. The two indicators together reflect the performance of a performance;

[0142] S4. Group the performance indicators and physiological characteristics extracted in S2 by repertoire difficulty, calculate the correlation between each group's performance and the changes in physiological characteristics before and after performance, analyze the impact of performance and repertoire difficulty on stress relief, and calculate the weights and weighted values of the impact of repertoire difficulty and performance indicators on stress relief through regression analysis. The specific steps are as follows:

[0143] S401: Group the performance indicators and the physiological characteristics extracted in S203 by the difficulty of the repertoire. Analyze the impact of performance at different difficulty levels on stress relief by calculating the correlation between performance at different difficulty levels and changes in physiological characteristics before and after the performance. Then, analyze the impact of performance and repertoire difficulty on stress relief. Calculate the influence weights of performance and repertoire difficulty through regression analysis to quantify the impact of performance and repertoire difficulty on stress relief. The specific steps are as follows:

[0144] S401.1. Group the performance indicators and the physiological characteristics extracted in S203 according to the difficulty of the music;

[0145] S401.2. For each group, calculate the Spearman rank correlation coefficient between the performance index and the changes in physiological characteristics before and after performance. The calculated correlation coefficient shows that the greater the difficulty, the smaller the correlation between the performance index and the changes in physiological characteristics before and after performance, indicating that the greater the difficulty, the worse the stress relief effect.

[0146] S401.3. To ensure the reliability of the correlation analysis results, the Kruskal-Wallis H test was performed to analyze whether there were significant differences in the changes in physiological characteristics after stress relief in each group, further verifying the conclusions of S401.2.

[0147] S401.4. Calculate the influence of performance index on stress relief using Bayesian regression fitting, with a weight of 0.2568 and a weight of -0.2937 for the influence of repertoire difficulty on stress relief, to quantify the effects of performance and repertoire difficulty on stress relief.

[0148] S402: Multiply the performance index and the piece difficulty score by their corresponding influence weights to obtain corresponding weighted values.

[0149] S5. Dimensionally transform the physiological characteristics and perform feature weighting. The weighted values of the difficulty of the repertoire and the performance index obtained in S4 are added together to quantitatively evaluate the effect of piano playing on reducing psychological stress. The specific steps are as follows:

[0150] S501. Remove the dimension of each physiological feature of the optimal feature subset obtained in S105 by using the dimensionless method corresponding to the table below, so that the value range of all features is between [0, 1].

[0151] Table 2. Dimensionless table

[0152] Target range Dimensionless method LF / HF [0,1] naturalization Mean RR interval [0,1] Anti-naturalization Rising branch time AVE [0,1] Anti-naturalization Reflection time AVE [0,1] Anti-naturalization Pulse width AVE [0,1] naturalization RMSSD of ascending limb area [0,1] Anti-naturalization

[0153] S502: multiply each physiological feature by its importance weight to obtain a weighted value of each physiological feature;

[0154] S503: Add the weighted value of each physiological characteristic and the weighted value obtained in step S402 to obtain a quantitative evaluation score of the effect of piano playing on alleviating psychological stress.

[0155] The above embodiments are preferred implementation modes of the present invention, but the implementation modes of the present invention are not limited to the above embodiments. Any other changes, modifications, substitutions, combinations, and simplifications that do not deviate from the spirit and principles of the present invention should be considered as equivalent replacement methods and are included in the scope of protection of the present invention.

Claims

1. An objective quantitative evaluation method for piano playing to reduce psychological stress, characterized by: The following steps are involved: S1. Collect pulse wave signals of subjects in a relaxed state and a stressed state and extract physiological features; the physiological features include heart rate variability features and pulse wave signal morphology features; calculate the importance weight of each physiological feature using random forest; S2. The subject plays pieces of music of varying difficulty on the piano, records the playing time, key number, and playing duration, collects the subject's pulse wave signal after the performance, and extracts the physiological characteristics; S3, calculating the subject's performance index during the piano playing process based on the playing time, the playing key number, and the playing duration, wherein the performance index includes rhythm accuracy and note accuracy; S4, grouping the performance indicators and the physiological characteristics extracted in step S2 according to the difficulty of the music played in step S2, and calculating weighted values of the music difficulty and the performance indicators in reducing the psychological stress of the subject through regression analysis; S5. Perform dimensionless processing on the physiological characteristics in step S1 and perform weighted value calculation according to the importance weight, and then add the weighted value calculated in step S4 to obtain a quantitative evaluation result of the effect of piano playing on reducing the psychological stress of the subject.

2. The objective quantitative evaluation method for piano playing to reduce psychological stress according to claim 1, characterized in that: In step S1, the relaxed state refers to the baseline physiological state of the individual without task execution before the stress-induced test; and the stressed state refers to the stress response state exhibited by the individual after the stress-induced test, which is specifically manifested by significant changes in HRV characteristics and pulse wave signal morphological characteristics. The HRV characteristics are the heart rate variability characteristics; wherein the stress-induced test is induced by completing the MIST mental arithmetic test; The process of step S1 is as follows: S101, guiding the subject to sit quietly for 5 minutes to enter a relaxed state, and collecting their pulse wave signal; then, guiding the user to complete the MIST mental arithmetic test stress induction task to induce psychological stress, and collecting the subject's pulse wave signal immediately after the task is completed; S102, preprocessing the collected pulse wave signal, including denoising and removing baseline drift, to eliminate interference components in the pulse wave signal; then, extracting HRV features and pulse wave signal morphological features from the preprocessed pulse wave signal, where the HRV features and pulse wave signal morphological features are both the physiological features; S103. Evaluate the effectiveness of physiological characteristics in stress detection by performing significance analysis on physiological characteristics in relaxation and stress states; S104. Using a random forest algorithm, a binary classification is performed based on physiological characteristics in a relaxed state and a stressed state, and a psychological stress detection model is constructed; S105. Perform feature selection, randomly construct feature subsets and score them, and finally obtain the feature subset with the best classification effect, and calculate the importance weight of each physiological feature through random forest.

3. The objective quantitative evaluation method for piano playing to reduce psychological stress according to claim 2, characterized in that: In step S2, recording the playing time and the playing key number is completed by an electronic keyboard device; the electronic keyboard device is composed of an electronic keyboard and a computer system, and can collect playing data in real time; During the playing process, the electronic piano device can determine in real time whether the keys are in a played state, obtain the playing key number of the played keys, calculate the playing time and playing duration of each note, and store the playing key number, the playing time and the playing duration in a computer.

4. The objective quantitative evaluation method for piano playing to reduce psychological stress according to claim 3, characterized in that: The process of step S2 is as follows: S201, displaying the songs played by the subject on a monitor, and classifying the songs into three difficulty levels: low, medium, and high according to the number of notes and rhythmic complexity of the songs; S202, after the performance, obtaining the subject's playing key number, playing time, and playing duration; S203 , collecting the pulse wave signal of the subject after playing the piano, and processing and extracting features of the pulse wave signal according to the step of S102 .

5. The objective quantitative evaluation method for piano playing to reduce psychological stress according to claim 1, characterized in that: The performance indicators in step S3 include: rhythm accuracy and note accuracy; rhythm accuracy is defined as the root mean square error between the actual rhythm and the standard rhythm, recorded as RMSE; note accuracy is defined as the percentage of correct key presses to the total number of key presses; The step S3 is as follows: S301, after the subject finishes playing, the key number, playing time, and playing duration of the subject's performance are recorded to form a note sequence played by the subject, and the played note sequence is aligned with the note sequence of the piece; S302. Calculate the number of all correct notes in the note sequence played by the subject, and calculate the percentage of the number of correct notes to the total number of notes played, as the note accuracy of the performance; calculate the deviation of the actual start time and actual duration of each note from the standard start time and standard duration, and calculate the RMSE, and standardize the RMSE to a range of 0 to 1, as the rhythm accuracy of the performance.

Images