Motor function estimating system, motor function estimating method and program
The motor function estimating system 1000 has a memory unit 1260 , a data processing unit 1220 (analyzing section), and a display unit 1400 . The memory unit 1260 memorizes waveform data about the time change of fingers tapping motion that is obtained by a motion sensor attached to a subject person who does the fingers tapping motion that is repetition of the opening and closing motion of two fingers of one hand. The data processing unit 1220 analyzes the waveform data in the memory unit 1260 . The display unit 1400 displays an analysis result. The data processing unit 1220 creates motion waveform based on waveform data, and plural characteristics based on the motion waveform. The data processing unit 1220 creates a motion disorder synthesis value that represents degree of motion disorder of the subject person by revising each of the plural characteristics based on characteristics of every age and synthesizing them.
1. A motor function estimating system comprising:
a memory unit for storing waveform data with reference to a time of a fingers tapping motion; the waveform data is obtained by a motion sensor attached to a subject person who does the fingers tapping motion that is repetition of an opening and closing motion of two fingers of one hand;
an analyzing section for analyzing the waveform data stored in the memory unit; and
a display unit for displaying an analysis result by the analyzing section;
wherein the analyzing section comprises:
a motion waveform generating section for making a motion waveform corresponding to the waveform data stored in the memory unit;
a characteristics generating section for making plural characteristics that represent features of the fingers tapping motion, based on the motion waveform;
an age-normalized characteristics generating section for creating each of plural age-normalalized characteristics by calculation a regression line or a regression curve that represents a distribution tendency depending on age of each of the plural characteristics and calculating a difference between each of the plural characteristics and a value of the age on the regression line or the regression curve;
a motion disorder synthesis value generating section for creating a motion disorder synthesis value that represents a degree of motion disorder of the subject person by comparing the plural age-normalized characteristics made by the age-normalized characteristics generating section with corresponding characteristics of healthy persons previously stored in the memory unit and synthesizing the plural characteristics.
2. The motor function estimating system according to claim 1 , wherein the characteristics generating section calculates time for which two fingers are in contact as the characteristics by using a local maximum point and local minimum value of waveform data about acceleration of the fingers tapping motion and a local maximum point and local minimum value of waveform data about jerk which is provided by differentiating the waveform data about acceleration.
3. The motor function estimating system according to claim 1 , wherein the characteristics generating section calculates a similarity degree of a Lissajous figure for the subject person and a Lissajous figure for healthy persons, the Lissajous figure being a dimensional image created from speed data and acceleration data of the fingers tapping motion as the characteristics.
4. The motor function estimating system according to claim 1 , wherein the characteristics generating section calculates, as the characteristics, a first statistical indicator value from a distance waveform of the fingers tapping motion when speed data of the fingers tapping motion become a local maximum value and a second statistical indicator value from a distance waveform of the fingers tapping motion when the speed data of the fingers tapping motion become a local minimum value.
5. The motor function estimating system according to claim 1 , wherein the analyzing section includes a motion disorder score estimating section for estimating a motion disorder score that represents a score of disease severity of a motion disorder from the motion disorder synthesis value by using a predetermined approximation function that shows relation of the motion disorder synthesis value and the motion disorder score.
6. The motor function estimating system according to claim 1 , wherein the motion disorder synthesis value generating section selects the characteristics having high value of AUC (Area Under a Receiver Operating Characteristic (ROC) Curve) if there are plural characteristics corresponding to any symptom of a motion disorder, and obtains the motion disorder synthesis value by multiplying the selected characteristics by a weight coefficient based on the value of the AUC and synthesizing the plural characteristics.
7. The motor function estimating system according to claim 6 , wherein the motion disorder synthesis value generating section normalizes each of the plural characteristics by using a distribution of healthy person group, and averages an absolute value of the normalized plural characteristics as the motion disorder synthesis value.
8. The motor function estimating system according to claim 6 , wherein the motion disorder synthesis value generating section synthesizes the plural characteristics by applying best weight coefficients to distinguish a healthy person data group from a motion disorder person data group, wherein the best weight coefficients are calculated in advance by applying statistical techniques to past information of each of the plural characteristics.
9. The motor function estimating system according to claim 1 , wherein the display unit displays each of the plural characteristics used to make the motion disorder synthesis value with a radar chart image.
10. A motor function estimating method comprising the steps of:
storing waveform data with reference to a time of a fingers tapping motion in a memory unit; wherein the waveform data is obtained by a motion sensor attached to a subject person who does the fingers tapping motion that is repetition of an opening and closing motion of two fingers of one hand;
analyzing the waveform data stored in the memory unit in an analyzing section; and
displaying an analysis result by the analyzing section on a display unit;
the analyzing step comprising:
generating, in a motion waveform generating section, a motion waveform corresponding to the waveform data stored in the memory unit;
generating, in a characteristics generating section, plural characteristics that represent features of the fingers tapping motion, based on the motion waveform;
generating, in an age-normalized characteristics generating section, each of plural age-normalized characteristics by calculating a regression line or a regression curve that represents a distribution tendency depending on age of each of the plural characteristics and calculating a difference between each of the plural characteristics and a value of the age on the regression line or the regression curve, and
generating, in a motion disorder synthesis value generating section, a motion disorder synthesis value that represents a degree of motion disorder of the subject person by comparing the plural age-normalized characteristics made by the age-normalized characteristics generating section with corresponding characteristics of healthy persons previously stored in the memory unit and synthesizing the plural characteristics.
11. The motor function estimating method according to claim 10 , further comprising the step of calculating, in the characteristics generating section, a time for which two fingers are in contact as the characteristics by using a local maximum point and a local minimum value of waveform data about acceleration of the fingers tapping motion and a local maximum point and a local minimum value of waveform data about jerk which is provided by differentiating the waveform data about acceleration.
12. The motor function estimating method according to claim 10 , further comprising the step of calculating, in the characteristics generating section, a similarity degree of a Lissajous figure for the subject person and a Lissajous figure for healthy persons, the Lissajous figure being a two dimensional image created from speed data and acceleration data of the fingers tapping motion as the characteristics.
13. The motor function estimating method according to claim 10 , further comprising the step of calculating, in the characteristics generating section, a first statistical indicator value on the value in a distance waveform of the fingers tapping motion when speed data of the fingers tapping motion become a local maximum value as the characteristics, and a second statistical indicator value from a distance waveform of the fingers tapping motion when the speed data of the fingers tapping motion become a local minimum value as the characteristics.
14. The motor function estimating method according to claim 10 , wherein the analyzing step further comprises estimating, in a motion disorder score estimating section, a motion disorder score that represents a score of disease severity of a motion disorder from the motion disorder synthesis value by using a predetermined approximation function that shows relation of the motion disorder synthesis value and the motion disorder score.
15. The motor function estimating method according to claim 10 , wherein the step of generating a motion disorder synthesis value further comprises selecting the characteristics having high value of AUC (Area Under a Receiver Operating Characteristic (ROC) Curve) if there are plural characteristics corresponding to any symptom of a motion disorder, and obtaining the motion disorder synthesis value by multiplying the selected characteristics by a weight coefficient based on the value of the AUC and synthesizing the plural characteristics.