Disease risk assessment assistance system

Abstract

This invention relates to a disease risk assessment assistance system, comprising: a non-invasive sensing device for sensing a photoplethysmography (PPG) signal of a user; a signal processing unit for capturing and processing the PPG signal to generate pulse information; a feature extraction module for obtaining time-domain feature information, frequency-domain feature information, and a waveform sharpness index from the pulse information; and a disease risk assessment module for assessing disease risk information of the user based on the time-domain feature information, the frequency-domain feature information, and the waveform sharpness index.

Claims

1. A disease risk assessment assistance system, comprising: A non-invasive sensing device senses a photoplethysmography (PPG) signal of a user; a signal processing unit coupled to the non-invasive sensing device captures and processes the PPG signal to generate pulse information; a feature extraction module coupled to the signal processing unit obtains a time-domain feature information, a frequency-domain feature information, and a waveform sharpness index from the pulse information; and a disease risk assessment module coupled to the feature extraction module assesses the disease risk information of the user based on the time-domain feature information, the frequency-domain feature information, and the waveform sharpness index; wherein the pulse information includes multiple main peaks, each main peak defining multiple corresponding pulse widths at multiple different heights, at least one main peak defining a maximum bottom width, and the ratio of the multiple pulse widths of the at least one main peak to the at least one maximum bottom width forms the waveform sharpness index; The frequency domain feature information includes an autonomic nerve parameter and a harmonic energy ratio.

2. The disease risk assessment auxiliary system as described in claim 1, wherein the time-domain feature information includes pulse amplitude (AMP), systolic time (Ts), diastolic time (Td), systolic slope (Ss), diastolic slope (Ds), area, systolic-to-diastolic ratio (Ts / Td), standard deviation of pulse area (Std_AREA), variability of pulse area (RMSSD_AREA), position of primary peak (P1), position of secondary peak (P2), time difference between primary and secondary peaks (ΔT, delta T), proportion of multiple peak positions (loc), variability of primary wave first-order slope (1D_RMSSD), standard deviation of NN intervals (SDNN), and root mean square of successive rhythm variation. Differences (RMSD), a ratio of adjacent cycle differences > 50 milliseconds (PNN50), a pulse rate, or any combination thereof.

3. The disease risk assessment assistance system as described in claim 1 further includes a visualization module coupled to the feature extraction module. The visualization module generates a physiological data integration chart based on the time-domain feature information, the frequency-domain feature information, the waveform sharpness index, or any combination thereof. The physiological data integration chart includes at least one physiological data trend chart, which is generated based on multiple measurements of the time-domain feature information, the frequency-domain feature information, the waveform sharpness index, or any combination thereof.

4. The disease risk assessment assistance system as described in claim 3, wherein the physiological data integration chart further includes a pulse waveform comparison chart, the pulse waveform comparison chart including multiple overlapping pulse waveforms, the pulse waveforms being generated based on the pulse information measured by the user at different time points.

5. A disease risk assessment assistance system as described in any one of claims 1 to 4, wherein the disease risk assessment module compares the time-domain feature information, the frequency-domain feature information, and the waveform sharpness index with validation data to perform logistic regression and machine learning to assess the user's disease risk information; wherein, The verification data includes diagnostic data for multiple users, each of which includes time-domain feature information, frequency-domain feature information, waveform sharpness index, and a disease label, which corresponds to the time-domain feature information, frequency-domain feature information, and waveform sharpness index.

6. The disease risk assessment assistance system as claimed in claim 1, wherein the autonomic nervous system parameters include low-frequency energy (LF), high-frequency energy (HF), and a low-frequency energy (LF) / high-frequency energy (HF) ratio; the harmonic energy ratios include a second harmonic ratio, a third harmonic ratio, and a fourth harmonic ratio, wherein the second harmonic ratio is the second peak amplitude / main peak amplitude, the third harmonic ratio is the third peak amplitude / main peak amplitude, and the fourth harmonic ratio is the fourth peak amplitude / main peak amplitude; the frequency domain feature information further includes a spectral centroid information, a frequency slope, and a frequency distribution width; the time domain feature information includes a pulse amplitude (AMP), a systolic time (Ts), a diastolic time (Td), a systolic slope (Ss), and a diastolic slope (Td). Ds), Area of ​​a single wave (Area), Contraction-to-diastolic ratio (Ts / Td), Standard deviation of pulse area (Std_AREA), Variation of pulse area (RMSSD_AREA), Position of primary peak (P1), Position of secondary peak (P2), Time difference between primary and secondary peaks (ΔT, delta T), Proportion of multiple peak positions (loc), Variation of first-order slope of primary wave (1D_RMSSD), Standard Deviation of NN Intervals (SDNN), Root Mean Square of Successive Differences (RMSSD), Proportion of adjacent cycle differences > 50 milliseconds (PNN50), Pulse rate (Pulse The disease risk assessment assistance system further includes a visualization module coupled to the feature extraction module. The visualization module generates a physiological data integration graph based on the time-domain feature information, the frequency-domain feature information, the waveform sharpness index, or any combination thereof. The physiological data integration graph includes at least one physiological data trend graph, which is generated based on multiple measurements of the time-domain feature information, the frequency-domain feature information, the waveform sharpness index, or any combination thereof. The physiological data integration graph further includes at least one physiological data trend graph. The system includes at least one physiological data variability rate, which is a comparison result of physiological data generated by the previous measurement of the time-domain feature information, the frequency-domain feature information, or the waveform sharpness index; the physiological data trend graph and the physiological data variability rate are displayed on a display unit; the physiological data summary graph further includes a pulse waveform comparison graph, which includes multiple overlapping pulse waveforms, which are generated based on the pulse information measured by the user at different time points; the pulse waveform comparison graph can be displayed on the display unit;The disease risk assessment module compares the time-domain feature information, the frequency-domain feature information, and the waveform sharpness index with validation data to perform logistic regression and machine learning to assess the user's disease risk information; where, ; The verification data includes diagnostic data for multiple users, each of which includes time-domain feature information, frequency-domain feature information, waveform sharpness index, and a disease label, which corresponds to the time-domain feature information, frequency-domain feature information, and waveform sharpness index.

7. The disease risk assessment assistance system as claimed in any one of claims 1 to 4, wherein the non-invasive sensing device includes a first component and a second component, the first component and the second component being pivotally connected and forming a finger placement space therebetween, the second component including a base, a movable member and a sensing unit, the movable member being movably disposed on the base, the sensing unit being disposed on the base, the movable member being able to be supported by one of the user's fingertips, and the sensing unit being able to project a light through one of the user's fingers and receive the light reflected by the finger.

8. The disease risk assessment assistance system as described in claim 7, wherein the sensing unit includes a light-emitting area and a light-receiving area disposed adjacent to each other, the light-emitting area is provided with a first polarizer, the light-receiving area is provided with a second polarizer, the light-emitting area projects light through the first polarizer to the finger, and the light reflected from the finger passes through the second polarizer to the light-receiving area; the first polarizer has a first polarization direction, the second polarizer has a second polarization direction, and the first polarization direction is orthogonal to the second polarization direction.

Images