Methods for predicting the failure of bioinformatics estimation and devices for estimating bioinformatics.

Abstract

A method for predicting the failure of bioinformatics estimation and an apparatus for estimating bioinformatics are provided. The method for predicting the failure of bioinformatics estimation includes: receiving a biosignal; obtaining a failure prediction index from the biosignal received up to a current time, the failure prediction index including at least one of the following: the current number of pulses of the biosignal up to the current time, the current signal quality of the biosignal up to the current time, the predicted maximum number of pulses of the biosignal to be measured up to a time limit, and the predicted maximum signal quality of the biosignal to be measured up to the time limit; and predicting whether the bioinformatics estimation will fail based on the failure prediction index before the time limit has elapsed.

Description

[0001] This application claims priority to Korean Patent Application No. 10-2020-0107742, filed on August 26, 2020, with the Korean Intellectual Property Office, and Korean Patent Application No. 10-2021-0013962, filed on February 1, 2021, with the Korean Intellectual Property Office, the entire disclosure of which is incorporated herein by reference. Technical Field

[0002] The apparatus and methods consistent with the example embodiments relate to predicting the failure of bioinformatics estimation and estimating bioinformatics when the bioinformatics estimation is predicted to be successful. Background Technology

[0003] Recently, with an aging population, soaring medical costs, and a shortage of medical personnel for specialized medical services, research is actively underway on IT-medical convergence technologies that combine information technology (IT) with medical technology. Specifically, the monitoring of human health is no longer limited to medical institutions but is expanding into mobile healthcare, enabling users to monitor their health anytime, anywhere in daily life, whether at home or in the office. Typical examples of biosignals indicating an individual's health include electrocardiogram (ECG) signals, photoplethysmography (PPG) signals, and electromyography (EMG) signals, and various biosignal sensors have been developed to measure these signals in daily life. Specifically, PPG sensors can estimate a person's blood pressure by analyzing the shape of the pulse wave, which reflects cardiovascular status, etc.

[0004] Biosignals (such as heart rate) are typically measured over short time periods, making them susceptible to significant impact from brief movements or noise. However, measuring complex biosignals or health indicators (such as blood pressure) may require measuring biosignals over longer periods. In such cases, because the biosignals are measured over a relatively long time, brief movements or noise do not significantly affect them. However, the long measurement time also means it is difficult to determine the success or failure of the measurement. Therefore, when measuring such complex biosignals or health indicators, it is necessary to predict early on whether the estimation of the biosignals will fail. Summary of the Invention

[0005] According to one aspect of an example embodiment, a method for predicting the failure of bioinformatics estimation is provided, the method comprising: receiving a biosignal; obtaining a failure prediction index from the biosignal received up to a current time, the failure prediction index including at least one of: the current number of pulses of the biosignal up to the current time, the current signal quality of the biosignal up to the current time, the predicted maximum number of pulses of the biosignal to be measured up to a time limit, and the predicted maximum signal quality of the biosignal to be measured up to the time limit; and predicting whether the bioinformatics estimation will fail based on the failure prediction index before the time limit has elapsed.

[0006] Biological signals may include at least one of electrocardiogram (ECG), photoplethysmography (PPG), cardiac impulse gaussography (BCG), electromyography (EMG), impedance plethysmography (IPG), pressure wave gaussography, and video plethysmography (VPG).

[0007] The steps to obtain the failure prediction index may include: obtaining the current pulse count by counting the number of pulses of the biological signal input up to the current time.

[0008] The steps to obtain failure prediction metrics may include: obtaining the current signal quality by calculating the similarity between pulses of the input biological signal up to the current time.

[0009] The steps to obtain the failure prediction metric may include: predicting the number of pulses during the remaining time period from the current time until the time limit is reached, and obtaining the maximum number of pulses based on the predicted number of pulses and the current number of pulses.

[0010] The steps to obtain the failure prediction metric may include: estimating the current heart rate based on the biological signals input up to the current time, and predicting the number of pulses during the remaining time period until the time limit is reached based on the estimated current heart rate.

[0011] The steps to obtain failure prediction metrics may include: predicting the number of pulses during the remaining time period from the current time until the time limit is reached, and obtaining maximum signal quality based on a predefined similarity between the predicted number of pulses and the similarity between the pulses of the input biological signal up to the current time.

[0012] The prediction steps may include: comparing each of the obtained failure prediction metrics with a threshold to obtain a comparison result; and outputting a failure flag based on the comparison result meeting the failure prediction criteria.

[0013] The type of failure prediction indicator, the threshold, or the failure prediction criterion may be adjusted based on at least one of the type of biological signal, the type of biological information, and the length of the time period.

[0014] The prediction process may include: starting from the prediction start point or the number of prediction start pulses, comparing the failed prediction metric with a threshold.

[0015] The method may further include: outputting a failure flag in response to determining that the time limit has expired.

[0016] The method may further include: estimating biological information based on biological signals input up to the current time; and outputting a success flag in response to successful estimation of biological information.

[0017] The prediction can be performed in response to the failure to output a success flag when estimating biological information.

[0018] According to another aspect of the example embodiment, an apparatus for estimating biological information is provided, the apparatus comprising: a sensor configured to measure biological signals from an object; and a processor configured to: execute a biological information estimation algorithm and a failure prediction algorithm for predicting failure of the biological information estimation; predict failure of the biological information estimation based on biological signals continuously input from the sensor; and, in response to the biological information estimation being predicted as a failure within a first time period, terminate the biological information estimation, guide a re-measurement, or further measure the biological signals until a second time period is reached.

[0019] The failure prediction algorithm can obtain a failure prediction index based on the input biological signals up to the current time; and based on the fact that the bioinformatics estimation algorithm has not output a success flag up to the current time, the failure prediction algorithm can predict the failure of bioinformatics estimation based on the failure prediction index.

[0020] Failure prediction metrics may include at least one of the following: the current number of pulses of the biological signal up to the current time, the current signal quality of the biological signal up to the current time, the maximum number of pulses of the biological signal to be measured up to the first time period, and the maximum signal quality of the biological signal to be measured up to the first time period.

[0021] The failure prediction algorithm can obtain the current pulse count by counting the number of pulses of the input biological signal up to the current time.

[0022] The failure prediction algorithm can determine the current signal quality by calculating the similarity between the pulses of the input biological signal up to the current time.

[0023] The failure prediction algorithm can predict the number of pulses during the remaining time period from the current time until the first time limit is reached, and can obtain the maximum number of pulses based on the predicted number of pulses and the current number of pulses.

[0024] The failure prediction algorithm can estimate the current heart rate based on the biological signals input up to the current time, and can predict the number of pulses during the remaining time period from the current time until the first time limit is reached based on the estimated current heart rate.

[0025] The failure prediction algorithm can predict the number of pulses during the remaining time period from the current time until the first time limit is reached, and can obtain the maximum signal quality based on the predefined similarity between the predicted number of pulses and the similarity between the pulses of the input biological signal up to the current time.

[0026] The failure prediction algorithm can compare each of the obtained failure prediction indicators with a threshold to obtain a comparison result, and can output a failure flag based on the comparison result meeting the failure prediction criteria.

[0027] The type of failure prediction indicator to be obtained, the threshold or failure prediction criterion may be adjusted based on at least one of the type of biological signal, the type of biological information and the length of the first time period.

[0028] The failure prediction algorithm can start from the prediction start point or the prediction start pulse number and compare at least some of the failure prediction metrics with a threshold.

[0029] In response to the output of a failure flag, the processor can also be configured to output information about the failure of the bioinformatics estimation, guidance information about remeasurement, and information about additional measurements up to the second time limit.

[0030] In response to the bioinformatics estimation not being predicted as a failure until the first time limit is reached, the failure prediction algorithm can output a failure flag.

[0031] The bioinformatics estimation algorithm can estimate bioinformatics based on the input biosignals up to the current time, and can output a success flag in response to successful bioinformatics estimation.

[0032] Bioinformation may include at least one of the following: blood pressure, vascular age, arterial stiffness, aortic pressure waveform, vascular compliance, stress index, and fatigue level.

[0033] According to another aspect of an example embodiment, an apparatus for estimating biological information is provided, the apparatus comprising: a sensor configured to measure biological signals from an object; and a processor configured to: count the current number of pulses of the biological signal from a preset start time up to a current time; while the sensor continuously measures the biological signal, predict, based on the current number of pulses and a remaining time period from the current time to a preset end time, a maximum number of pulses of the biological signal to be received from the preset start time up to the preset end time; predict whether the estimation of the biological information will fail based on the predicted maximum number of pulses; and output a failure signal based on the prediction that the estimation of the biological information will fail, so as to cause the sensor to stop measuring the biological signal or to start re-measuring the biological signal. Attached Figure Description

[0034] Figure 1 This is a block diagram illustrating a device for estimating biological information according to an example embodiment;

[0035] Figure 2 This is a flowchart illustrating a method for predicting the failure of bioinformatics estimation according to an example embodiment;

[0036] Figure 3 This is a flowchart illustrating a method for predicting the failure of bioinformatics estimation according to another example embodiment;

[0037] Figure 4 This is a flowchart illustrating a method for predicting the failure of bioinformatics estimation according to yet another example embodiment;

[0038] Figure 5 This is a diagram illustrating the prediction of failure over time.

[0039] Figure 6 This is a block diagram illustrating an apparatus for estimating biological information according to another example embodiment;

[0040] Figure 7 This is a flowchart illustrating a method for estimating biological information according to an example embodiment; and

[0041] Figures 8 to 10 This is a diagram illustrating various examples of electronic devices, including those used for estimating biological information. Detailed Implementation

[0042] Details of other embodiments are included in the following detailed description and accompanying drawings. The advantages and features of the invention, as well as methods of implementing the invention, will become clearer from the embodiments described in detail below with reference to the accompanying drawings. Throughout the drawings and detailed description, the same reference numerals will be understood to denote the same elements, features, and structures, unless otherwise described.

[0043] It will be understood that although the terms first, second, etc., may be used herein to describe various elements, these elements should not be limited by these terms. These terms are used only to distinguish one element from another. Furthermore, unless the context clearly indicates otherwise, the singular form is intended to include the plural form as well. It will also be understood that, unless explicitly stated to the contrary, when an element is referred to as “including” another element, that element is not intended to exclude one or more other elements, but rather to include one or more other elements. In the following description, terms such as “unit” and “module” indicate units for performing at least one function or operation, and they may be implemented using hardware, software, or a combination thereof.

[0044] Figure 1 This is a block diagram illustrating a device for estimating biological information according to an example embodiment.

[0045] Reference Figure 1 The device 100 for estimating biological information includes a sensor 110 and a processor 120.

[0046] Sensor 110 can measure biological signals from a subject. Specifically, the biological signals can be measured continuously over a predetermined time period and can have repetitive pulse waveforms. Examples of biological signals may include at least one of the following: electrocardiogram (ECG) signals, photoplethysmography (PPG) signals, cardiac impulse plethysmography (BCG) signals, electromyography (EMG) signals, impedance plethysmography (IPG) signals, pressure wave signals, video volume plethysmography (VPG) signals, etc.

[0047] Sensor 110 may include, for example, a PPG sensor for measuring PPG signals. The PPG sensor may include one or more light sources and a detector. The one or more light sources emit light onto a user's object, and the detector detects the light emitted onto the object and subsequently reflected or scattered from it. The light source may include a light-emitting diode (LED), a laser diode (LD), a phosphor, etc. The light source may be configured as a single light source or an array of two or more light sources. Each light source may emit light of a different wavelength. Furthermore, the detector may include a photodiode, a phototransistor (PTr), a complementary metal-oxide-semiconductor (CMOS) image sensor, a charge-coupled device (CCD) image sensor, etc., and may be configured as a single detector or an array of two or more detectors.

[0048] The processor 120 can be electrically connected to the sensor 110 directly or via wireless communication. The processor 120 can control the sensor 110 to measure biological signals used for estimating biological information within a time period (e.g., 60 seconds). The time period can be set to various values, taking into account the type of biological information to be estimated, the computing power of the device 100 used for estimating the biological information, the measurement location of the biological signal, etc. Furthermore, the biological signals measured by the sensor 110 can be continuously input to the processor 120 in real time during the measurement.

[0049] Upon receiving a biosignal from sensor 110, processor 120 can remove noise (such as motion noise) from the biosignal by using various noise removal methods (such as filtering, smoothing, etc.). For example, if the biosignal is a PPG signal, processor 120 can perform bandpass filtering at a cutoff frequency of 1 Hz to 10 Hz.

[0050] When a biosignal is input from sensor 110, processor 120 can execute a failure prediction algorithm to continuously predict the probability of bioinformation estimation failure, such that even if bioinformation is continuously estimated before the time limit is reached, processor 120 can predict early on whether there is a probability that bioinformation estimation will fail. Bioinformation may include at least one of, for example, blood pressure, vascular age, arterial stiffness, aortic pressure waveform, vascular compliance, pressure index, fatigue level, etc., but examples of bioinformation are not limited to these.

[0051] Specifically, the failure prediction algorithm can be pre-generated by the processor 120 or an external device. When generating the failure prediction algorithm, valid and invalid biological signals from samples are collected, wherein valid biological signals enable successful bioinformation estimation, and invalid biological signals cause bioinformation estimation failure. The performance of the failure prediction algorithm is evaluated based on any one or any combination of the following: (1) sensitivity determined by comparing cases where bioinformation estimation is predicted to fail with cases where bioinformation estimation actually fails; (2) singularity determined by comparing cases where bioinformation estimation is predicted to fail with cases where bioinformation estimation actually succeeds; and (3) failure prediction time.

[0052] In the following text, reference will be made to Figures 2 to 5 Various embodiments are described to predict the probability of failure of a predictive bioinformatics estimate performed by processor 120.

[0053] Figure 2 This is a flowchart illustrating a method for predicting the failure of bioinformatics estimation according to an example embodiment.

[0054] First, in operation 211, the processor 120 can receive biological signals from the sensor 110. The sensor 110 can continuously measure the biological signals over a time period, and after the measurement of the biological signals is completed, the measured biological signals can be input to the processor 120 in real time.

[0055] Then, by using a failure prediction algorithm, processor 120 can perform a real-time prediction operation 200, estimating whether the biological information will lead to failure, while continuously receiving biological signals. The prediction can be performed without waiting for the time limit for measuring the biological signal to be reached.

[0056] In operation 212, by using a failure prediction algorithm, processor 120 can obtain a failure prediction index based on the biological signals received up to the current time. The failure prediction index may include the current number of pulses, the current signal quality, the maximum number of pulses, or the maximum signal quality, and the type of failure prediction index to be obtained can be predetermined according to various conditions (such as the type of biological signal, the type of biological information, the length of the set time period, etc.).

[0057] Figure 5 This is a graph illustrating the failure prediction metrics obtained over time. (See reference.) Figure 5 By using the measurement start time 0 and the current time T current The time period between them, the preset prediction start time T start With current time T current The time period between, or corresponding to the predicted start pulse number (also known as the predicted start pulse) B start The time point and the current time T current During the time interval between input biosignals, the number of pulses is counted over time. The failure prediction algorithm obtains the cumulative number of pulses counted so far as the current pulse number. A preset prediction start time T is used. start Can be compared with the predicted start pulse number B start Match or no match. For example, the preset prediction start time T. start It can be set to a preset time to start measurement from measurement start time 0, and predicts the start pulse B. start It can be set to be the nth pulse counted from the measurement start time 0 (e.g., as...). Figure 5 The second pulse shown in the figure), where n represents a positive integer.

[0058] Furthermore, the failure prediction algorithm can calculate up to the current time T. currentThe similarity between two input pulses can be calculated, and the statistical values ​​of the calculated similarity (e.g., average, median, maximum, etc.) can be used to determine the current signal quality. For example, the failure prediction algorithm can calculate the similarity between temporally consecutive and adjacent pulses (e.g., the first pulse and the second pulse, the second pulse and the third pulse, etc.). Optionally, the failure prediction algorithm can also calculate the similarity between any two pulses among all detected pulses, or between two pulses separated by a predetermined interval (e.g., the first pulse and the third pulse, the second pulse and the fourth pulse, etc.). However, the similarity is not limited to these.

[0059] The similarity between pulses can be represented by Euclidean distance, Pearson correlation coefficient, Spearman correlation coefficient, cosine similarity, etc., but the disclosed embodiments are not limited to these.

[0060] Furthermore, the failure prediction algorithm can predict the failure at the current time T. current With time limit T Timeout The remaining time period (T) between Timeout -T current The number of pulses available during the period can be determined, and the maximum number of pulses can be obtained based on the predicted number of pulses and the current number of pulses. For example, the failure prediction algorithm can obtain the current heart rate based on the input biosignals up to the current time; the predicted number of pulses can be obtained by multiplying the obtained current heart rate by the remaining time until the time limit; and the value obtained by adding the obtained predicted number of pulses and the current number of pulses can be used as the value between the start time 0 and the time limit T. Timeout The maximum number of pulses during the time period between the two points.

[0061] Furthermore, the failure prediction algorithm can predict the failure at the current time T. current With time limit T Timeout The remaining time period (T) between Timeout -T current The number of pulses to be obtained during the prediction period can be determined, and the maximum signal quality can be obtained by using the similarity between the predicted number of pulses and the similarity between the pulses obtained up to the current time. For example, by setting the similarity between the predicted number of pulses to a predetermined value (e.g., "1") and by setting the similarity between the pulses obtained up to the current time to a statistical value (e.g., maximum, average, median, etc.), the failure prediction algorithm can obtain the statistical value (e.g., average, median, maximum, etc.) of the total value obtained by adding all the set values ​​and the similarity between the pulses obtained up to the current time as the maximum signal quality.

[0062] Then, in operation 213, the failure prediction algorithm can predict the failure of bioinformatics estimation by using the obtained failure prediction indices. For example, by comparing the obtained failure prediction indices with thresholds set for each failure prediction index, and if the comparison results meet predefined failure prediction criteria, the failure prediction algorithm can predict that the bioinformatics estimation will fail. The prediction can be made within a time period T. Timeout Previously completed. Thresholds or failure prediction criteria for each failure prediction indicator can be set and adjusted based on at least one of the following: type of biological signal, type of biological information, length of time period, etc.

[0063] In addition, from the preset prediction start time T start And / or predict the start pulse number B start Initially, the failure prediction algorithm may perform processing to obtain failure prediction metrics and / or compare thresholds with individual failure prediction metrics. The prediction start time T can be set differently for each failure prediction metric. start Or predict the number of starting pulses B start In addition, one or both of the prediction start time or prediction start pulse number can be set for each failure prediction metric.

[0064] For example, refer to Figure 5 The failure prediction algorithm can start from a preset prediction start time T. start The process begins by comparing the current pulse count, current signal quality, maximum pulse count, and maximum signal quality—all indicators of the failure prediction metrics—with thresholds. In another example, the failure prediction algorithm may start when the cumulative pulse count is greater than or equal to a preset prediction start pulse count B. start At the start of the time, the process will compare the current signal quality, maximum pulse count, and maximum signal quality with the threshold in the failure prediction metrics.

[0065] Subsequently, when the estimation of biological information is predicted to fail in operation 213, the failure prediction algorithm can output a failure flag in operation 214.

[0066] If, based on the biological signals obtained up to the current time, the estimation of biological information at the current time in operation 213 is not predicted as a failure, the failure prediction algorithm may determine in operation 215 whether the current time has exceeded the time limit. Upon determination, if the estimation up to the current time has not been predicted as a failure and the time limit has not been exceeded, the failure prediction algorithm may proceed to operation 211 to perform the failure prediction process in operation 200 using biological signals at a subsequent time. Upon determination, if the estimation up to the current time has not been predicted as a failure and the time limit has been exceeded, the failure prediction algorithm may output a failure flag in operation 214 to indicate prediction failure and / or indicate that the estimation of biological information has failed.

[0067] The time limit can be equal to the time interval during which sensor 110 measures the biosignal, or it can be set to be shorter than the time interval, and can be adjusted if necessary. By setting the time limit as described above, the failure prediction algorithm can predict the probability of failure of biosignals at an early stage.

[0068] Figure 3 This is a flowchart illustrating a method for predicting bioinformatics estimation failure according to another example embodiment.

[0069] First, in operation 211, processor 120 receives biological signals from sensor 110.

[0070] Then, by executing a failure prediction algorithm while biosignals are continuously input, processor 120 can perform the processing of predicting failures in bioinformation estimation in operation 200. (Refer to the above...) Figure 2 As described, in operation 212, the failure prediction algorithm can obtain a failure prediction index using biological signals obtained up to the current time, and in operation 213, it can predict the failure of the bioinformatics estimation by using the obtained failure prediction index. Once the bioinformatics estimation is predicted to fail based on the prediction in operation 213, the failure prediction algorithm can output a failure flag in operation 214, and once the bioinformatics estimation is not predicted to fail, the failure prediction algorithm can determine in operation 215 whether the current time has exceeded the time limit. If the current time has exceeded the time limit, the failure prediction algorithm can output a failure flag in operation 214 to indicate prediction failure and / or failure of the bioinformatics estimation.

[0071] Furthermore, according to the example embodiment, while the biosignal is input in operation 211, the processor 120 may execute a bioinformation estimation algorithm for estimating bioinformation based on the biosignals input up to the current time. The bioinformation estimation algorithm may be executed concurrently with or before the processing in operation 200 that uses a failure prediction algorithm to predict failure.

[0072] In operation 311, the bioinformatics estimation algorithm can estimate bioinformatics using the biosignals input up to the current time. Specifically, the bioinformatics estimation algorithm can receive some necessary data (e.g., acquired pulses, etc.) from the failure prediction algorithm and can share the data. Optionally, if the bioinformatics estimation algorithm performs processing to generate some data required by the failure prediction algorithm, the bioinformatics estimation algorithm can share the generated data with the failure prediction algorithm.

[0073] Bioinformatics estimation algorithms can extract features needed to estimate bioinformatics by using input biosignals up to the current time, differential signals of biosignals (e.g., first-order differential signals, second-order differential signals, etc.), and can estimate bioinformatics by using the extracted features and a predefined bioinformatics estimation model. Specifically, features may include the maximum amplitude value of the biosignal, the area of ​​the waveform, the time value and / or amplitude value associated with the propagating and reflected waves, or combinations thereof.

[0074] In operation 312, if bioinformation estimation is successfully performed using biosignals input up to the current time, the bioinformation estimation algorithm may output a success flag in operation 313. Once the success flag is output, the processor 120 may terminate the failure prediction algorithm and control the sensor 110 to stop measuring biosignals. In another example, when the success flag is output, the processor 120 may stop executing the failure prediction algorithm, but may control the sensor 110 to continue measuring biosignals until the time limit is reached, in order to obtain a final estimation result based on the biosignals received up to the time limit.

[0075] When determined in operation 312, if the estimation of biological information has not been successful up to the current time, the failure prediction algorithm can determine whether the failure prediction criteria are met in operation 213 by using the comparison results of the failure prediction indicators.

[0076] Furthermore, when executing both the bioinformatics estimation algorithm and the failure prediction algorithm simultaneously, even if the bioinformatics estimation algorithm outputs a success flag first, the failure prediction algorithm may ultimately output a failure flag if it predicts that the bioinformatics estimation will fail based on the biosignal at the current time. Even if the bioinformatics estimation algorithm successfully estimates bioinformatics using the input biosignal up to the current time, noise (such as motion noise) may be generated in the biosignal, increasing the probability that the bioinformatics estimation result may be relatively inaccurate. Therefore, the failure prediction algorithm may ultimately predict that the bioinformatics estimation will fail, allowing the user to remeasure the biosignal or perform further continuous measurements.

[0077] Figure 4 This is a flowchart illustrating a method for predicting the failure of bioinformatics estimation according to yet another example embodiment.

[0078] Figure 4 The above references are shown. Figure 2 and Figure 3 The described failure prediction algorithm describes the overall process of predicting failures in operation 200, and shows an example of obtaining the current pulse count, maximum pulse count, current signal quality, and maximum signal quality as failure prediction metrics and using all obtained failure prediction metrics in failure prediction.

[0079] First, when receiving a biosignal from sensor 110 at the current time, processor 120 may perform detection of the received biosignal (e.g., bit detection, but not limited thereto) in operation 411 to obtain the current pulse count in operation 412 and the current heart rate in operation 413.

[0080] Then, processor 120 can accumulate pulses by detection in operation 414, and can obtain the current signal quality in operation 415 by using the accumulated pulses. As described above, processor 120 can calculate the similarity between the pulses accumulated so far, and can obtain a statistical value (e.g., average value) of the calculated similarity as the current signal quality. Specifically, processor 120 can calculate the similarity between temporally consecutive and adjacent pulses (e.g., first pulse and second pulse, second pulse and third pulse, etc.), or calculate the similarity between any two pulses among all detected pulses, or the similarity between two pulses separated by a predetermined interval.

[0081] Subsequently, in operation 416, processor 120 can obtain the maximum number of pulses by using the current number of pulses obtained in operation 412 and the current heart rate obtained in operation 413. For example, by using the remaining time period until the time limit is reached and the current heart rate, processor 120 can predict the number of pulses that can be detected during the remaining time period, and by adding the predicted number of pulses to the current number of pulses, processor 120 can obtain the maximum number of pulses to be detected throughout the entire time period.

[0082] Then, in operation 417, processor 120 may obtain the maximum signal quality based on the number of pulses predicted in operation 416 during the remaining time period up to the time limit, and based on the current signal quality obtained in operation 415. For example, processor 120 may obtain the maximum signal quality by setting the similarity between pulses predicted during the remaining time period before the time limit to any value (e.g., a maximum value "1"), setting the similarity between pulses obtained up to the current time to a statistical value (e.g., maximum, average, median, etc.), and by obtaining the average of the similarities between all pulses.

[0083] Subsequently, in operations 418, 419, 420, and 421, the processor 120 compares the obtained failure prediction metrics with thresholds set for each failure prediction metric. As shown in Table 1 below, the comparisons of the current pulse count, maximum pulse count, maximum signal quality, and current signal quality with the corresponding thresholds in operations 418, 419, 420, and 421 can be performed starting from a predetermined prediction start time. Furthermore, as shown in Table 1 below, the comparisons of the maximum pulse count, maximum signal quality, and current signal quality with the corresponding thresholds in operations 419, 420, and 421 can be performed starting from the predicted pulse count. Table 1 below only shows examples of thresholds and comparison times for each failure prediction metric, and conditions including thresholds and comparison times can be adjusted.

[0084] [Table 1]

[0085] Current pulse count B1＜5 30 seconds - Current signal quality Q1＜0.52 0 seconds 7 people Maximum number of pulses B2＜20 30 seconds 7 people Maximum signal quality Q2＜0.8 30 seconds 7 people

[0086] In Table 1, B1 and B2 represent the current number of pulses and the predicted maximum number of pulses to be measured up to the end of the time limit, respectively. Q1 and Q2 represent the current signal quality and the predicted maximum signal quality to be measured up to the end of the time limit, respectively.

[0087] In operation 422, processor 120 may determine whether the failure prediction criteria are met based on the comparison results of various failure prediction metrics. If the bioinformatics estimation algorithm is executed together with the failure prediction algorithm, then the determination of whether the failure prediction criteria are met can be performed in operation 422 when the bioinformatics estimation algorithm does not output a success flag. However, the failure prediction algorithm is not limited to this, and operation 422 can be executed regardless of whether the bioinformatics estimation algorithm outputs a success flag.

[0088] Failure prediction criteria can be pre-set and adjusted based on at least one of various conditions, such as the biological signal, biological information, or time frame to be estimated. For example, an estimate can be predicted as a failure if all four failure prediction metrics are below a threshold, if two or more failure prediction metrics are below a threshold, or if a specific failure prediction metric specified for each particular case is below a threshold.

[0089] Then, if the failure prediction criteria are met in operation 422, the processor 120 may output a failure flag in operation 423; and if the failure prediction criteria are not met in operation 422, then as described above... Figure 2 and Figure 3 As described, processor 120 can perform operation 215 to determine whether a time limit has expired.

[0090] Return to reference Figure 1 The processor 120 can estimate biological information (e.g., blood pressure) based on biological signals.

[0091] For example, as mentioned above Figure 3 As described, processor 120 can execute bioinformatics estimation algorithms and failure prediction algorithms, and can perform bioinformatics estimation and failure prediction simultaneously using the biosignals received so far.

[0092] Once the bioinformatics estimation algorithm outputs a bioinformatics estimation success flag, the processor 120 can terminate the failure prediction algorithm, control the sensor 110 to stop measuring biosignals, and provide the bioinformatics estimation results to the user through an output device included in the device 100 for estimating bioinformatics or a connected external output device.

[0093] If the failure prediction algorithm outputs a bioinformatics estimation failure flag within the time limit, the processor 120 may terminate the execution of the bioinformatics estimation algorithm and / or control the sensor 110 to stop measuring the biosignal, and may provide guidance information on bioinformatics estimation failure and / or remeasurement through a connected external output device.

[0094] Optionally, if the bioinformatics estimation is not predicted to be a failure based on the failure prediction index within the first time period, and a failure flag is output after the time period has elapsed, the processor 120 may provide the user with guidance information regarding the failure and / or remeasurement of the bioinformatics estimation via an external device included in or connected to the device 100 for estimating bioinformatics, or may further measure the biosignal during a predetermined second time period, or may further perform bioinformatics estimation and / or failure estimation.

[0095] In another example, as referenced above. Figure 2 As described, processor 120 may first perform failure prediction processing by executing a failure prediction algorithm based on the biological signals received so far, and if a failure flag is output after a time limit has elapsed, processor 120 may also perform a bioinformation estimation algorithm based on the biological signals input up to the time limit. If the bioinformation estimation fails, processor 120 may provide guidance information to the user regarding the failure of bioinformation estimation and / or remeasurement, or may further measure the biological signals during a predetermined second time limit, or may further perform bioinformation estimation and / or failure estimation.

[0096] Figure 6 This is a block diagram illustrating a device for estimating biological information according to another example embodiment.

[0097] Reference Figure 6The device 600 for estimating biological information includes a sensor 110, a processor 120, an output interface 610, a storage device 620, and a communication interface 630. The sensor 110 and processor 120 have been described in detail above, therefore their descriptions will be omitted.

[0098] The output interface 610 can provide the user with the results processed by the processor 120 through various visual / non-visual methods, such as a display, sound output device, or tactile device. For example, the output interface 610 can provide the user with information about failed bioinformatics estimations, guidance on remeasurement, and / or progress on further measurements.

[0099] Storage device 620 may store various information required for estimating biometrics and / or predicting failure. For example, storage device 620 may store failure prediction metrics to be obtained, thresholds for each failure prediction metric, failure prediction criteria, prediction start time, prediction start pulse number, time limit, biometric estimation model, and user characteristic information (such as the user's age, gender, health status, etc.).

[0100] Storage device 620 may include, but is not limited to, at least one of the following storage media: flash memory, hard disk memory, multimedia card micro memory, card memory (e.g., SD memory, XD memory, etc.), random access memory (RAM), static random access memory (SRAM), read-only memory (ROM), electrically erasable programmable read-only memory (EEPROM), programmable read-only memory (PROM), magnetic memory, magnetic disk and optical disk.

[0101] The communication interface 630 can communicate with external devices using various communication technologies to send data required for estimating biological information and / or predicting failure, as well as data generated and processed by sensor 110 or processor 120, to the external devices, and to receive data required for estimating biological information and / or predicting failure, as well as data generated and processed by sensor 110 or processor 120, from the external devices. External devices may include various information processing devices (such as smartphones, tablet PCs, desktop computers, laptop computers, etc.).

[0102] Examples of communication technologies may include, but are not limited to, Bluetooth communication, Bluetooth Low Energy (BLE) communication, Near Field Communication (NFC), WLAN communication, Zigbee communication, Infrared Data Association (IrDA) communication, Wi-Fi Direct (WFD) communication, Ultra Wideband (UWB) communication, Ant+ communication, Wi-Fi communication, and 3G, 4G, and 5G communication.

[0103] Figure 7 This is a flowchart illustrating a method for estimating biological information according to an example embodiment.

[0104] Figure 7 This is an example of a method for estimating biological information performed by the aforementioned devices 100 and 600 for estimating biological information.

[0105] First, in operation 711, sensor 110 continuously measures biosignals from the subject over a predetermined time period.

[0106] Then, in operation 712, processor 120 may receive biosignals measured by sensor 110 up to the current time.

[0107] Subsequently, when a biosignal measured up to the current time is received from sensor 110 in operation 712, processor 120 executes a bioinformation estimation algorithm and a failure prediction algorithm in operations 713 and 714 to simultaneously perform bioinformation estimation and failure prediction based on the biosignal received up to the current time.

[0108] Next, if the bioinformation estimation is successful in operation 713 and a success flag is output, the processor 120 can output the bioinformation estimation result in operation 716 and end the measurement of biosignals, the execution of the failure prediction algorithm, etc.

[0109] If a bioinformation estimation success flag is not output in operation 715, and in operation 717 the estimation failure is predicted by a failure prediction algorithm before the time limit expires or a failure flag is output after the time limit expires, the processor 120 can determine whether to perform additional measurements. In this case, whether to perform additional measurements is predetermined. For example, if the estimation is predicted to fail before the time limit expires, it can be set to end the bioinformation estimation without additional measurements; or if the estimation is not predicted to fail within the time limit but a failure flag is output after the time limit expires, it can be set to further perform biosignal measurements, bioinformation estimation, and failure prediction during the additional time period.

[0110] Then, when it is determined in operation 718 that an additional measurement should be performed, processor 120 may increase the measurement time in operation 720 and control sensor 110 to continue measuring biosignals in operation 711; if "No", processor 120 may terminate the biosignal measurement and bioinformation estimation algorithm and the failure prediction algorithm, and may output guidance information indicating the failure of bioinformation measurement or guidance information about remeasurement in operation 719.

[0111] When determining in operation 717, if the estimate is not predicted as a failure at the current time and the time limit has not expired so that no failure flag is output, the processor 120 may proceed to operation 712 of receiving biosignals to estimate bioinformation in operation 713 and predict failure in operation 714 by using biosignals from subsequent times.

[0112] Figures 8 to 10 This is a diagram illustrating various examples of electronic devices, including those used for estimating biological information.

[0113] The electronic devices including the aforementioned devices 100 and 600 for estimating biological information may include, for example: Figures 8 to 10 The smartwatch-type wearable device 800, mobile device 900 such as a smartphone, and ear-worn device 1000 are shown. However, the electronic devices are not limited to these and may include smart wristbands, smart glasses, smart rings, smart patches, smart necklaces, tablet PCs, etc. The electronic devices include devices 100 and 600 for estimating biometric information, and components of devices 100 and 600 for estimating biometric information may be installed in a single electronic device or in two or more electronic devices.

[0114] Reference Figure 8 The smartwatch-type wearable device 800 includes a main body 810 and a strap 830.

[0115] The straps 830 connected to both ends of the body 810 can be flexible to wrap around the user's wrist. The straps 830 may include a first strap and a second strap, separate from each other. One end of the first and second straps is connected to the body 810, and the other ends of the first and second straps can be connected to each other via a connector. The connector may be implemented to provide a magnetic connection, a Velcro connection, a pin connection, etc., but the connection type is not limited to these. Air may be injected into the straps 830, or the straps 830 may be provided with an air bladder to be elastic according to changes in pressure applied to the wrist and to transmit changes in wrist pressure to the body 810. A battery may also be embedded in the body 810 or the straps 830 to power the wearable device 800.

[0116] Sensor 820 of devices 100 and 600 for estimating biological information may be mounted on the rear surface of body 810 and may include a light source and a detector. Processor of devices 100 and 600 for estimating biological information may be mounted in body 810, may be connected to various components (such as sensor 820, etc.) to control components, and may perform the aforementioned functions of estimating biological information, predicting estimation failure, and / or various functions performed by other electronic devices.

[0117] Furthermore, the body 810 may include storage devices for storing various information required for estimating biological information and / or predicting failure, as well as other information processed by the components. Additionally, the body 810 may include a manipulator 840 disposed on a side surface of the body 810, which receives control commands from the user and transmits the received control commands to a processor. The manipulator 840 may have a power button for inputting commands to turn the wearable device 800 on / off.

[0118] Furthermore, a display for outputting information to the user can be mounted on the front surface of the main body 810. The display may have a touchscreen for receiving touch input. The display can receive touch input from the user and send the touch input to the processor, and can display the processing results of the processor.

[0119] In addition, the main body 810 may include a communication interface for communicating with external devices. The communication interface can send and receive data with external devices.

[0120] Reference Figure 9 The smartphone-type mobile device 900 may include a main body and a display panel. The main body forms the appearance of the mobile device 900. The main body has a first surface on which the display panel and a cover glass are sequentially disposed, and the display panel is exposed to the outside through the cover glass. The main body has a second surface 910 on which a sensor 930 is disposed. The sensor 930 may include one or more light sources 931 and detectors 932. In addition, a camera module 920 and / or an infrared sensor, etc., may be disposed on the second surface 910 of the main body. Processors and other components (such as communication interfaces, storage devices, etc.) for estimating biological information of devices 100 and 600 may be disposed in the main body, and may estimate biological information and / or predict failures, as well as store processing results and communicate with external devices.

[0121] Reference Figure 10 The ear-worn device 1000 may include a main body and an ear strap.

[0122] Users can wear the wearable device by attaching an ear strap to their ear. Depending on the shape of the ear-worn device 1000, the ear strap may be omitted. The main body can be inserted into the external auditory canal. The sensor 1020 can be installed in the main body. The ear-worn device 1000 can provide the user with bio-information estimation results and / or failed prediction results as sound, or can transmit the results to an external device (e.g., a mobile device, tablet PC, personal computer, etc.) via a communication module installed in the main body.

[0123] While not limited thereto, the exemplary embodiments can be implemented as computer-readable code on a computer-readable recording medium. A computer-readable recording medium is any data storage device capable of storing data that can subsequently be read by a computer system. Examples of computer-readable recording media include read-only memory (ROM), random access memory (RAM), CD-ROM, magnetic tape, floppy disk, and optical data storage devices. The computer-readable recording medium can also be distributed across a networked computer system, such that the computer-readable code is stored and executed in a distributed manner. Furthermore, the exemplary embodiments can be written as computer programs transmitted via a computer-readable transmission medium (such as a carrier wave) and received and implemented in a general-purpose or special-purpose digital computer executing the program. Moreover, it is understood that in the exemplary embodiments, one or more units of the above-described devices and apparatus may include circuits, processors, microprocessors, etc., and are capable of executing computer programs stored on computer-readable media.

[0124] The foregoing exemplary embodiments and advantages are merely illustrative and should not be construed as limiting. This teaching can be readily applied to other types of devices. Furthermore, the description of the exemplary embodiments is intended to be illustrative and should not limit the scope of the claims, and many alternatives, modifications, and variations will be apparent to those skilled in the art.

Claims

1. A method for predicting the failure of bioinformatics estimation, the method comprising: Receive biological signals, wherein the biological signals include at least one of electrocardiogram signals, photoplethysmography signals, cardiac impact recording signals, electromyography signals, impedance volume recording signals, pressure wave signals, and video volume recording signals; The failure prediction index is obtained from the biological signals received up to the current time. The failure prediction index includes the predicted maximum number of pulses of the biological signal to be measured up to the time limit. Predicting whether bioinformatics estimation will fail based on failure prediction metrics; and In response to the bioinformatics estimation not being predicted as a failure, determine whether the timeframe has expired. The steps for obtaining the failure prediction index include: predicting the number of pulses during the remaining time period from the current time until the time limit is reached, and obtaining the maximum number of pulses based on the predicted number of pulses and the current number of pulses of the biosignal up to the current time.

2. The method of claim 1, wherein, Failure prediction metrics also include: the current number of biological signal pulses up to the current time. The step of obtaining the failure prediction index also includes obtaining the current pulse count by counting the number of pulses of the biological signal input up to the current time.

3. The method according to claim 1, wherein, Failure prediction metrics also include: the current signal quality of the biological signal up to the present time. The step of obtaining the failure prediction index also includes: obtaining the current signal quality by calculating the similarity between the pulses of the input biological signals up to the current time.

4. The method according to claim 1, wherein, The steps to obtain the failure prediction metric include: estimating the current heart rate based on the biological signals input up to the current time, and predicting the number of pulses during the remaining time period until the time limit is reached based on the estimated current heart rate.

5. The method according to claim 1, wherein, Failure prediction metrics also include: the predicted maximum signal quality of the biological signal to be measured up to the time limit; The steps for obtaining failure prediction metrics also include: obtaining maximum signal quality based on a predefined similarity between the predicted number of pulses and the similarity between the pulses of the input biological signal up to the current time.

6. The method according to claim 1, wherein, The prediction steps include: Each of the obtained failure prediction metrics is compared with a threshold to obtain the comparison results; and Based on the comparison results meeting the failure prediction criteria, a failure flag is output.

7. The method according to claim 6, wherein, The type of failure prediction indicator, the threshold, and the failure prediction criteria are adjusted based on at least one of the type of biological signal, the type of biological information, and the length of the time period.

8. The method according to claim 6, wherein, The prediction process includes comparing the failure prediction metric with a threshold, starting from the prediction start time or prediction start pulse.

9. The method according to claim 1, further comprising: In response to the determination that the time limit has expired, a failure flag is output.

10. The method according to any one of claims 1 to 9, further comprising: Estimating biological information based on biological signals input up to the present time; as well as In response to successful bioinformatics estimation, a success flag is output.

11. The method according to claim 10, wherein, In response to the failure to output a success flag when estimating bioinformation, a step is performed to predict whether the bioinformation estimation will fail.

12. An apparatus for estimating biological information, the apparatus comprising: A sensor is configured to measure biosignals from a subject, wherein the biosignals include at least one of electrocardiogram (ECG) signals, photoplethysmography (PPG) signals, cardiac impaction (CA) signals, electromyography (EMG) signals, impedance plethysmography (IPL) signals, pressure wave signals, and video plethysmography (VLP) signals; and The processor is configured as follows: A failure prediction algorithm for performing bioinformatics estimation and predicting the failure of bioinformatics estimation. Failure to predict bioinformation estimation based on continuous biosignals input from sensors; and If the bioinformatics estimation is predicted to fail within the first time period, the bioinformatics estimation is terminated, and a remeasurement is initiated, or the bioinformatics signal is measured again until the second time period is reached. The failure prediction metrics include: the predicted maximum number of biosignal pulses to be measured up to the first time limit. The failure prediction algorithm predicts the number of pulses during the remaining time period from the current time until the first time limit is reached, and obtains the maximum number of pulses based on the predicted number of pulses and the current number of biological signals up to the current time.

13. The device according to claim 12, wherein, The failure prediction algorithm obtains failure prediction indices based on the input biosignals up to the current time; and Since the bioinformatics estimation algorithm has not output a success flag up to the current time, the failure prediction algorithm predicts the failure of bioinformatics estimation based on failure prediction indicators.

14. The device according to claim 13, wherein, The failure prediction metrics also include at least one of the following: the current number of pulses of the biological signal up to the current time, the current signal quality of the biological signal up to the current time, and the predicted maximum signal quality of the biological signal to be measured up to the first time period.

15. The device according to claim 14, wherein, The failure prediction algorithm also obtains the current pulse count by counting the number of pulses of the input biological signal up to the current time.

16. The device according to claim 14, wherein, The failure prediction algorithm also obtains the current signal quality by calculating the similarity between the pulses of the input biological signal up to the current time.

17. The device according to claim 14, wherein, The failure prediction algorithm estimates the current heart rate based on the input biosignals up to the current time, and predicts the number of pulses during the remaining time period from the current time until the first time limit is reached based on the estimated current heart rate.

18. The device according to claim 14, wherein, The failure prediction algorithm also obtains maximum signal quality based on a predefined similarity between the predicted number of pulses and the similarity between the pulses of the input biological signal up to the current time.

19. The device according to claim 13, wherein, The failure prediction algorithm compares each of the obtained failure prediction metrics with a threshold to obtain a comparison result, and outputs a failure flag based on whether the comparison result meets the failure prediction criteria.

20. The device according to claim 19, wherein, The type of failure prediction indicator to be obtained, the threshold, and the failure prediction criteria are adjusted based on at least one of the type of biological signal, the type of biological information, and the length of the first time period.

21. The device according to claim 19, wherein, Starting from the prediction start time or prediction start pulse, the failure prediction algorithm compares at least some of the failure prediction metrics with a threshold.

22. The device according to claim 19, wherein, In response to the output of a failure flag, the processor is also configured to output information about the failure of the bioinformatics estimation, guidance information about remeasurement, and information about additional measurements up to the second time limit.

23. The device according to claim 13, wherein, In response to the bioinformatics estimation not being predicted as a failure and the first time limit having passed, the failure prediction algorithm outputs a failure flag.

24. The device according to any one of claims 12 to 23, wherein, The bioinformatics estimation algorithm estimates bioinformatics based on the input biosignals up to the current time, and outputs a success flag in response to successful bioinformatics estimation.

25. The device according to claim 12, wherein, Bioinformation includes at least one of the following: blood pressure, vascular age, arterial stiffness, aortic pressure waveform, vascular compliance, stress index, and fatigue level.

26. An apparatus for estimating biological information, the apparatus comprising: A sensor is configured to measure biosignals from a subject, wherein the biosignals include at least one of electrocardiogram (ECG) signals, photoplethysmography (PPG) signals, cardiac impaction (CA) signals, electromyography (EMG) signals, impedance plethysmography (IPL) signals, pressure wave signals, and video plethysmography (VLP) signals; and The processor is configured as follows: Count the current number of biological signals from the preset start time up to the current time; While the sensor continuously measures biological signals, based on the current number of pulses and the remaining time from the current time to the preset end time, the maximum number of biological signal pulses to be received from the preset start time until the preset end time is reached is predicted. Will the estimation of biological information based on the predicted maximum pulse number fail? The estimation based on biological information will fail to predict and output a failure signal, so that the sensor stops measuring the biological signal or starts to remeasure the biological signal.

Images