CN111481193B - Fall risk assessment and early warning method and system - Google Patents
Fall risk assessment and early warning method and system Download PDFInfo
- Publication number
- CN111481193B CN111481193B CN202010303476.8A CN202010303476A CN111481193B CN 111481193 B CN111481193 B CN 111481193B CN 202010303476 A CN202010303476 A CN 202010303476A CN 111481193 B CN111481193 B CN 111481193B
- Authority
- CN
- China
- Prior art keywords
- value
- early warning
- data
- lead
- fall risk
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Active
Links
- 238000000034 method Methods 0.000 title claims abstract description 27
- 238000012502 risk assessment Methods 0.000 title claims abstract description 26
- 238000004458 analytical method Methods 0.000 claims abstract description 10
- 238000001228 spectrum Methods 0.000 claims description 35
- 238000000354 decomposition reaction Methods 0.000 claims description 16
- 238000004364 calculation method Methods 0.000 claims description 14
- 230000006870 function Effects 0.000 claims description 13
- 230000000284 resting effect Effects 0.000 claims description 8
- 238000007405 data analysis Methods 0.000 claims description 6
- 238000013210 evaluation model Methods 0.000 claims description 6
- 238000012549 training Methods 0.000 claims description 6
- 230000009466 transformation Effects 0.000 claims description 6
- 230000036772 blood pressure Effects 0.000 claims description 4
- 230000033001 locomotion Effects 0.000 claims description 4
- 238000007476 Maximum Likelihood Methods 0.000 claims description 3
- 239000011159 matrix material Substances 0.000 claims description 3
- 238000010845 search algorithm Methods 0.000 claims description 3
- 210000004556 brain Anatomy 0.000 abstract description 11
- 238000010586 diagram Methods 0.000 description 9
- 230000032683 aging Effects 0.000 description 2
- 238000006243 chemical reaction Methods 0.000 description 2
- 238000000605 extraction Methods 0.000 description 2
- 230000010354 integration Effects 0.000 description 2
- 238000005070 sampling Methods 0.000 description 2
- 206010049816 Muscle tightness Diseases 0.000 description 1
- 229940060587 alpha e Drugs 0.000 description 1
- 230000006399 behavior Effects 0.000 description 1
- 238000011161 development Methods 0.000 description 1
- 230000018109 developmental process Effects 0.000 description 1
- 239000003814 drug Substances 0.000 description 1
- 230000005611 electricity Effects 0.000 description 1
- 238000004070 electrodeposition Methods 0.000 description 1
- 238000005516 engineering process Methods 0.000 description 1
- 238000011156 evaluation Methods 0.000 description 1
- 230000005021 gait Effects 0.000 description 1
- 230000005802 health problem Effects 0.000 description 1
- 230000010365 information processing Effects 0.000 description 1
- 238000002347 injection Methods 0.000 description 1
- 239000007924 injection Substances 0.000 description 1
- 210000005036 nerve Anatomy 0.000 description 1
- 230000000474 nursing effect Effects 0.000 description 1
- 230000001575 pathological effect Effects 0.000 description 1
- 238000012545 processing Methods 0.000 description 1
- 238000004393 prognosis Methods 0.000 description 1
- 238000011160 research Methods 0.000 description 1
- 238000013058 risk prediction model Methods 0.000 description 1
- 230000001629 suppression Effects 0.000 description 1
Images
Classifications
-
- A—HUMAN NECESSITIES
- A61—MEDICAL OR VETERINARY SCIENCE; HYGIENE
- A61B—DIAGNOSIS; SURGERY; IDENTIFICATION
- A61B5/00—Measuring for diagnostic purposes; Identification of persons
- A61B5/24—Detecting, measuring or recording bioelectric or biomagnetic signals of the body or parts thereof
- A61B5/316—Modalities, i.e. specific diagnostic methods
- A61B5/369—Electroencephalography [EEG]
-
- A—HUMAN NECESSITIES
- A61—MEDICAL OR VETERINARY SCIENCE; HYGIENE
- A61B—DIAGNOSIS; SURGERY; IDENTIFICATION
- A61B5/00—Measuring for diagnostic purposes; Identification of persons
- A61B5/103—Detecting, measuring or recording devices for testing the shape, pattern, colour, size or movement of the body or parts thereof, for diagnostic purposes
- A61B5/11—Measuring movement of the entire body or parts thereof, e.g. head or hand tremor, mobility of a limb
- A61B5/1116—Determining posture transitions
- A61B5/1117—Fall detection
-
- A—HUMAN NECESSITIES
- A61—MEDICAL OR VETERINARY SCIENCE; HYGIENE
- A61B—DIAGNOSIS; SURGERY; IDENTIFICATION
- A61B5/00—Measuring for diagnostic purposes; Identification of persons
- A61B5/72—Signal processing specially adapted for physiological signals or for diagnostic purposes
- A61B5/7235—Details of waveform analysis
- A61B5/7253—Details of waveform analysis characterised by using transforms
-
- A—HUMAN NECESSITIES
- A61—MEDICAL OR VETERINARY SCIENCE; HYGIENE
- A61B—DIAGNOSIS; SURGERY; IDENTIFICATION
- A61B5/00—Measuring for diagnostic purposes; Identification of persons
- A61B5/72—Signal processing specially adapted for physiological signals or for diagnostic purposes
- A61B5/7235—Details of waveform analysis
- A61B5/7264—Classification of physiological signals or data, e.g. using neural networks, statistical classifiers, expert systems or fuzzy systems
- A61B5/7267—Classification of physiological signals or data, e.g. using neural networks, statistical classifiers, expert systems or fuzzy systems involving training the classification device
-
- A—HUMAN NECESSITIES
- A61—MEDICAL OR VETERINARY SCIENCE; HYGIENE
- A61B—DIAGNOSIS; SURGERY; IDENTIFICATION
- A61B5/00—Measuring for diagnostic purposes; Identification of persons
- A61B5/74—Details of notification to user or communication with user or patient ; user input means
- A61B5/746—Alarms related to a physiological condition, e.g. details of setting alarm thresholds or avoiding false alarms
Landscapes
- Health & Medical Sciences (AREA)
- Life Sciences & Earth Sciences (AREA)
- Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- Surgery (AREA)
- Public Health (AREA)
- Pathology (AREA)
- Veterinary Medicine (AREA)
- Biomedical Technology (AREA)
- Heart & Thoracic Surgery (AREA)
- Medical Informatics (AREA)
- Molecular Biology (AREA)
- Biophysics (AREA)
- Animal Behavior & Ethology (AREA)
- General Health & Medical Sciences (AREA)
- Artificial Intelligence (AREA)
- Physiology (AREA)
- Psychiatry (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Signal Processing (AREA)
- Psychology (AREA)
- Evolutionary Computation (AREA)
- Fuzzy Systems (AREA)
- Mathematical Physics (AREA)
- Dentistry (AREA)
- Oral & Maxillofacial Surgery (AREA)
- Measurement And Recording Of Electrical Phenomena And Electrical Characteristics Of The Living Body (AREA)
Abstract
The invention discloses a fall risk assessment and early warning method, which comprises the following steps: step 1, collecting physiological information data of a user; step 2, integrating and converting the electroencephalogram signal data acquired in the step 1 according to the prediction analysis requirement, and calculating a risk probability value; step 3, fall risk assessment; step 4, early warning; wherein: the user physiological information data includes conventional physiological information and brain electrical signal data.
Description
Technical Field
The invention relates to the technical field of information processing, in particular to a fall risk assessment and early warning method and system.
Background
Population aging is a serious civil problem faced by China, the biggest problem brought by aging is the health problem of the aged, and easy falling is a common phenomenon faced by the aged. How to predict the fall risk of the old population, so that early warning is carried out on the high-risk population in advance, and nursing intervention and behavior treatment are important challenges facing current research. The introduction of modern electronic information technology into fall risk prognosis is an urgent need for clinical medicine and technical development.
In the evaluation of the risk factors of the elderly, there are a number of risk factors that cause the elderly to fall, including lifestyle, medical history, gait, etc., and the correlation of the concentration of attention and fall is one of the widely neglected by researchers. The brain electrical signal of human body contains rich physiological, pathological and psychological information, and can directly extract the concentration degree parameter from the brain electrical signal.
Therefore, the invention provides a fall risk assessment and early warning method, which can accurately assess the fall risk and can carry out risk early warning on the basis so as to reduce the fall risk of the old.
Disclosure of Invention
The invention is realized by adopting the following technical scheme:
a fall risk assessment and early warning method comprising: step 1, collecting physiological information data of a user; step 2, integrating and converting the electroencephalogram signal data acquired in the step 1 according to the prediction analysis requirement, and calculating a risk probability value; step 3, fall risk assessment; step 4, early warning; wherein: the user physiological information data includes conventional physiological information and brain electrical signal data.
The fall risk assessment and early warning method comprises the following steps: the step 1 comprises the following steps: (1) The user provides various conventional physiological information including age, sex, height, weight, blood pressure, quantity of motion, sleeping status and the like through the terminal equipment; (2) And acquiring electroencephalogram signal data in a resting state and an attention task state through an electroencephalogram acquisition system.
The fall risk assessment and early warning method comprises the following steps: step 2, data integration and conversion are carried out on the collected user brain electrical signal data according to the prediction analysis requirement, and the extracted user brain electrical signal characteristic value and the normal group deviation value are calculated, so that a corresponding risk probability value is obtained, and the method comprises the following steps:
2.1, respectively inputting the lead data of the electroencephalogram signals acquired in the step 1 into a power spectrum calculation module to perform classical mode decomposition and Hilbert transformation:
taking x (t) as an original electroencephalogram signal, calculating all maximum value points and minimum value points of the signal x (t), and connecting the maximum value points by an interpolation method to form an upper envelope line x 1 (t) connecting minima points to form a lower envelope x 2 (t) obtaining x 1(t) and x2 Flat of (t)The average curve is as follows:
w 1 (t)＝(x 1 (t)+x 2 (t))/2
calculating a new signal s 1 (t) the formula:
s 1 (t)＝x(t)-w 1 (t)
judging the new signal s 1 Whether all local extremum points and zero crossing points of (t) are identical or differ by at most 1, or s 1 (t) the mean value of the upper and lower envelopes is 0, if not, let x (t) =s 1 (t) repeating the above steps of obtaining the average curve and the new signal until a new signal s is obtained 1μ (t) all local extremum points and zero crossing points are identical or differ by at most 1, or s 1μ (t) the mean of the upper and lower envelopes is 0. To eliminate the riding wave, avoid the IMF component amplitude becoming constant, define the standard deviation threshold SD<0.25, the formula of which is:
wherein mu is the decomposition times when the step of calculating the average curve and the new signal is finished, and T is the signal period;
let c 1 (t)＝s 1μ (t), then c 1 (t) is the first IMF component of the original signal x (t);
solving for a first residual component r 1 (t) the formula:
r 1 (t)＝x(t)-c 1 (t)
judgment r 1 (t) whether it is a monotonic function, if so, ending the EMD decomposition, if not, letting x (t) =r 1 (t) repeating the above steps of obtaining the average curve, the new signal and the residual component to obtain n IMF components c i (t) until the residual component r obtained by the last decomposition n (t) ending the EMD decomposition when it is a monotonic function or less than 0.5;
thus, the original signal x (t) is expressed as:
hilbert transformation is carried out on each obtained IMF component, and the formula is as follows:
wherein δ is the cauchy principal component;
thus, a Hilbert spectrum H (ω, t) is obtained with amplitude over time and instantaneous frequency, the formula of which is:
wherein ,
for instantaneous amplitude +.>
Is the instantaneous frequency;
integrating H (omega, t) with time to obtain a marginal spectrum function P (omega), wherein the formula is as follows:
thereby, the power spectrum value P on the fixed frequency band of each lead under the rest state and the attention task state is obtained 1,k (f)、P 2,k (f) The formula is as follows:
wherein ,P1,k (f) Is the kth guide of the resting stateCombined power spectrum value, P 2,k (f) The power spectrum value of the kth lead for the attention task state, f represents frequency, f ∑ Represents a set of all frequency points within a fixed frequency band, ω=2pi f;
2.2 the total power spectrum value P on the fixed frequency band of each lead obtained in the step 2.1 1,k (f)、P 2,k (f) Inputting the power spectrum value to an optimal lead selection module, calculating the lead contribution degree by adopting an improved distance criterion, and defining a relative distance coefficient beta (k) by using the difference between the power spectrum value in the rest state and the task state of the same frequency band during calculation, wherein the formula is as follows:
wherein m is the number of leads of the electroencephalogram acquisition equipment, and alpha epsilon [0,1] represents the attention degree of the same lead in different states and is used for adjusting the individual specificity caused by the difference of users in the process of calculating the relative distance coefficient;
thus, the lead represented by the maximum value of beta (k) is selected as the optimal lead set k on the fixed frequency band zuiyou (j) (representing the serial number of the sampling electrode), j=1, 2,3,..;
2.3 the k obtained in step 2.2 zuiyou (j) Task state power spectrum value on optimal lead
The input deviation calculation module calculates a power spectrum mean value range matrix (I) under the same state as the normal crowd>
(m is the number of means on each lead) the deviation d, which is given by:
wherein a weight factor epsilon is introduced j ∈[0,1]Representing the relationship to each optimal leadThe degree of injection;
thus, the average distance value d is obtained, namely the deviation value d, and the absolute value |d| of different deviation values d corresponds to different risk probability values and is marked as F 1 。
The fall risk assessment and early warning method comprises the following steps:
3.1, performing Bayesian network learning by utilizing a tabu search algorithm, determining the influence weight of each node of the Bayesian network, and establishing a Bayesian network model, namely searching a network structure which is best matched with a training sample set on the premise of giving one data sample set;
3.2, carrying out parameter learning on the probability of each node by utilizing a maximum likelihood estimation algorithm, namely searching the probability distribution of each node of the network on the basis of sample data, and determining the conditional probability density at each node of the Bayesian network model by utilizing a network topological structure, a training sample set and priori knowledge: each item of conventional physiological information of the user is respectively marked as x 1 ，x 2 ，……x i ,F(x i ) To generate x i Probability of F (x) j |x i ) Represents x i Under the condition x occurs j Is a fall risk probability value F (x|x) 1 ,x 2 ,......x i ) The calculation is as follows:
F(x|x 1 ,x 2 ,......x i )＝ΠF(x i |x 1 ,......x i-1 )......F(x 2 |x 1 )F(x 1 )
wherein i, j is the number of normal physiological information, i+.j
3.3, calculating a user falling risk probability value F (x), wherein the formula is as follows:
F(x)＝F 1 ×F(x|x 1 ,x 2 ,...... xi )。
the fall risk assessment and early warning method comprises the following steps: and (3) performing risk early warning by using the fall risk probability value F (x) calculated in the step (3).
The fall risk assessment and early warning system comprises a data acquisition module, a data analysis module, an assessment model module and an early warning module, wherein: the data acquisition module is used for collecting physiological information data of the user; the data analysis module is used for integrating and converting data according to the electroencephalogram signal data acquired by the data acquisition module and the prediction analysis requirement, and calculating a risk probability value; the evaluation model module is used for establishing a fall risk evaluation model and calculating a fall risk value; the early warning module is used for early warning.
The fall risk assessment and early warning system, wherein: the system is configured to perform the fall risk assessment and early warning method as described in one of the above.
Drawings
FIG. 1 is a block diagram of a fall risk assessment and early warning system according to the present invention;
FIG. 2 is a block diagram of a conventional physiological information acquisition system of the present invention;
FIG. 3 is a block diagram of an electroencephalogram signal acquisition system of the present invention;
FIG. 4 is a block diagram of an electroencephalogram signal-based attention feature extraction and probability calculation system according to the present invention;
fig. 5 is a block diagram showing the correspondence between absolute values of deviation values and fall risk probability values according to the present invention;
FIG. 6 is a flowchart of the establishment of a Bayesian network-based fall risk assessment model in accordance with the present invention;
FIG. 7 is a Bayesian network-based fall risk prediction model in accordance with the present invention;
fig. 8 is a block diagram showing a correspondence between a fall risk probability value and a fall risk degree according to the present invention.
Detailed Description
Specific embodiments of the present invention will be described in detail below with reference to fig. 1-8.
As shown in fig. 1, the fall risk assessment and early warning system of the present invention includes a data acquisition module, a data analysis module, an assessment model module and an early warning module. The data acquisition module is used for collecting physiological information data of the user; the data analysis module is used for integrating and converting data according to the electroencephalogram signal data acquired by the data acquisition module and the prediction analysis requirement, and calculating a risk probability value; the evaluation model module is used for establishing a fall risk evaluation model and calculating a fall risk value; the early warning module is used for early warning.
The fall risk assessment and early warning method comprises the following steps:
step 2, integrating and converting the electroencephalogram signal data acquired in the step 1 according to the prediction analysis requirement, and calculating a risk probability value;
step 3, establishing a calculation model and calculating a falling risk value;
and step 4, early warning.
Specifically:
the step 1 comprises the following steps: (1) The user provides various conventional physiological information including age, sex, height, weight, blood pressure, quantity of motion, sleeping status and the like through terminal equipment such as a computer and the like; (2) The electroencephalogram data in a resting state and an attention task state are collected through an electroencephalogram collection system, and the collection duration is consistent with the task duration;
as shown in FIG. 2, the system is a block diagram of a conventional physiological information acquisition system of the invention, and uses a terminal device such as a computer to input various conventional physiological information of a user, including age, gender, height, weight, blood pressure, exercise amount, sleeping condition and the like.
As shown in fig. 3, which is a block diagram of an electroencephalogram signal acquisition system of the present invention, an electroencephalogram signal acquisition system is established by adopting a double-computer and an electroencephalogram signal acquisition device, and electroencephalogram signal data acquisition comprises the following steps:
1) And (3) collecting resting state brain electrical signal data: the computer 1 does not display the function of the attention task, at the moment, the user keeps still and does not carry out any muscle tension and movement, the computer 2 carries out electroencephalogram signal acquisition and analysis processing under the resting state, and the acquisition time length is consistent with the task time length;
2) And acquiring electroencephalogram signal data under the attention task state: the computer 1 realizes the function of displaying the attention task, the user finishes the attention task, the computer 2 collects and analyzes the electroencephalogram signals, and the collecting time length of the electroencephalogram signals is consistent with the time length of the task. Attention tasks are exemplified as follows: the computer 1 displays a trolley on a screen, a user uses a mouse to control the trolley to move so as to avoid randomly-appearing obstacles, the higher the attention degree is when the user finishes a task, the more obstacles are avoided, the lower the attention degree is when the user finishes the task, and the fewer the number of obstacles are avoided.
The embodiment of the invention arranges the electroencephalogram acquisition electrode based on the 10-20 electrode lead positioning standard, and the electrode position accords with the international universal standard.
And 2, carrying out data integration and conversion on the acquired user brain electrical signal data according to the prediction analysis requirement by utilizing an improved Hilbert-Huang algorithm, and calculating the extracted user brain electrical signal characteristic value and the normal group deviation value so as to obtain a corresponding risk probability value.
Fig. 4 is a block diagram of a system for extracting attention features and calculating probability based on electroencephalogram signals according to an embodiment of the invention. Step 2 comprises the following steps:
2.1, respectively inputting the lead data of the electroencephalogram signals acquired in the step 1 into a power spectrum calculation module. The module adopts a Hilbert-Huang algorithm, and comprises two parts of classical mode decomposition (Empirical Mode Decomposition, EMD) and Hilbert transformation. The electroencephalogram signals are decomposed into a limited number of intrinsic mode functions (Intrinsic Mode Function, IMF) with different feature scales by an EMD module.
Taking x (t) as an original electroencephalogram signal, calculating all maximum value points and minimum value points of the signal x (t), and connecting the maximum value points by an interpolation method to form an upper envelope line x 1 (t) connecting minima points to form a lower envelope x 2 (t) obtaining x 1(t) and x2 The average curve of (t) is given by:
w 1 (t)＝(x 1 (t)+x 2 (t))/2
thereby obtaining a new signal s 1 (t) the formula:
s 1 (t)＝x(t)-w 1 (t)
judging the new signal s 1 Whether all local extremum points and zero crossing points of (t) are identical or differ by at most 1, or s 1 (t) the mean value of the upper and lower envelopes is 0, if not, let x (t) =s 1 (t) repeating the above steps to obtain a levelThe step of homography and new signal until the new signal s is obtained 1μ (t) all local extremum points and zero crossing points are identical or differ by at most 1, or s 1μ (t) the mean of the upper and lower envelopes is 0. To eliminate the riding wave, avoid the IMF component amplitude becoming constant, define the standard deviation threshold SD<0.25, the formula of which is:
wherein mu is the decomposition times when the step of calculating the average curve and the new signal is finished, and T is the signal period;
let c 1 (t)＝s 1μ (t), then c 1 (t) is the first IMF component of the original signal x (t);
solving for a first residual component r 1 (t) the formula:
r 1 (t)＝x(t)-c 1 (t)
judgment r 1 (t) whether it is a monotonic function, if so, ending the EMD decomposition, if not, letting x (t) =r 1 (t) repeating the above steps of obtaining the average curve, the new signal and the residual component to obtain n IMF components c i (t) until the residual component r obtained by the last decomposition n (t) ending the EMD decomposition when it is a monotonic function or less than 0.5;
thus, the original signal x (t) can be expressed as:
for each IMF component c obtained i (t) carrying out Hilbert transformation, wherein the formula is as follows:
wherein δ is the cauchy principal component;
thus, a Hilbert spectrum H (ω, t) is obtained with amplitude over time and instantaneous frequency, the formula of which is:
wherein ,
for instantaneous amplitude +.>
Is the instantaneous frequency;
integrating H (omega, t) with time to obtain a marginal spectrum function P (omega), wherein the formula is as follows:
thereby, the power spectrum value P on the fixed frequency band of each lead under the rest state and the attention task state is obtained 1,k (f)、P 2,k (f) The formula is as follows:
wherein ,P1,k (f) The power spectrum value of the kth lead in the resting state, P 2,k (f) The power spectrum value of the kth lead for the attention task state, f represents frequency, f ∑ Represents the set of all frequency points within a fixed frequency band, ω=2pi f.
2.2 the power spectrum value P on the fixed frequency band of each lead obtained in the step 2.1 1,k (f)、P 2,k (f) Inputting the data to an optimal lead selection module, calculating the lead contribution degree by adopting an improved distance criterion, and defining a relative distance coefficient beta (k) by the difference between the rest state and the power spectrum value in the task state of the same frequency band during calculation, wherein the formula is as followsThe method comprises the following steps:
wherein m is the number of leads (i.e. the number of electrodes) of the electroencephalogram acquisition device, and alpha E [0,1] characterizes the attention degree of the same lead in different states, and is used for adjusting the individual specificity caused by a user in the process of calculating the relative distance coefficient. The larger β (k) indicates that the more the power spectrum values in the rest state and the task state of the lead differ, i.e., the higher the recognition contribution of the lead to the task.
Thus, the lead represented by the maximum value of beta (k) is selected as the optimal lead set k on the fixed frequency band zuiyou (j) (representing the serial number of the sampling electrode), j=1, 2,3.
2.3 the k obtained in step 2.2 zuiyou (j) Task state power spectrum value on optimal lead
The input deviation calculation module calculates a power spectrum mean value range matrix (I) under the same state as the normal crowd>
(m is the number of means on each lead) the deviation d, which is given by: />
Wherein a weight factor epsilon is introduced j ∈[0,1]Representing the attention degree of each optimal lead, and adjusting the data deviation caused by the deviation between the lead position and the standard lead position when the user collects the brain electricity. The power spectrum mean value range under the same state of the normal crowd in the embodiment of the invention is taken from the electroencephalogram signal power spectrum mean value range on all leads of the crowd with the highest degree of completion when the attention task in the embodiment of the invention is completed by the same age and sex as the user, and the characteristic extraction step is synchronous2.1。
Thus, the average distance value d is obtained, namely the deviation value d, and the absolute value |d| of different deviation values d corresponds to different risk probability values and is marked as F 1 The method comprises the steps of carrying out a first treatment on the surface of the Dividing |d| into different grades according to a step length of 2, wherein the absolute value of the |d| increases by 2,F each time 1 When the magnitude of (d) is increased by 0.2 and exceeds 10, the magnitude is constant to be 1, and the absolute value of (d) is equal to F 1 An example of the correspondence of (a) is shown in fig. 5.
According to one embodiment of the present invention, the electroencephalogram signals in steps 2,3 and 4 are based on frequency band θ waves, f ∑ Ranging from 4-7Hz, associated with brain nerve suppression.
Fig. 6 is a flowchart of the establishment of the fall risk assessment model based on the bayesian network according to the present invention. Step 3 comprises the following steps:
3.1, performing Bayesian network learning by using a Tabu Search algorithm Tabu Search (TS), determining the influence weights of all nodes of the Bayesian network (the influence factors of all conventional physiological data on falling), and establishing a Bayesian network model, namely searching a network structure which is best matched with a training sample set on the premise of giving one data sample set, wherein the method comprises the following steps of:
and 3.2, carrying out parameter learning on the probability of each node by utilizing a maximum likelihood estimation algorithm, namely searching the probability distribution of each node of the network on the basis of sample data. The conditional probability density at each node of the bayesian network model is determined using the network topology and the training sample set and prior knowledge, and is illustrated in this embodiment as: each item of conventional physiological information of the user is respectively marked as x 1 ，x 2 ，……x i ,F(x i ) To generate x i Probability of F (x) j |x i ) Represents x i Under the condition x occurs j Is a fall risk probability value F (x|x) 1 ,x 2 ,......x i ) The calculation is as follows:
F(x|x 1 ,x 2 ,......x i )＝∏F(x i |x 1 ,......x i-1 )......F(x 2 |x 1 )F(x 1 )
where i, j is the number of normal physiological information, i+.j.
3.3, calculating a user falling risk probability value, wherein the formula is as follows:
F(x)＝F 1 ×F(x|x 1 ,x 2 ,......x i )
and 4, performing risk early warning by using the fall risk probability value F (x) calculated in the step 3. The F (x) is classified into one to five levels according to the size, and the falling risk degree is defined according to the size of F (x), and the correspondence between the size of the falling probability of F (x) and the falling risk degree is shown in fig. 8. And (3) early warning is carried out on the user with F (x) exceeding 0.6 and high falling risk.
Claims (3)
1. A fall risk assessment and early warning method comprising: step 1, collecting physiological information data of a user; step 2, integrating and converting the data of the electroencephalogram signals acquired in the step 1 according to the prediction analysis requirement, and calculating the extracted characteristic values of the electroencephalogram signals of the user and the deviation values of the normal group of people so as to obtain corresponding risk probability values; step 3, fall risk assessment; step 4, early warning; the method is characterized in that: the user physiological information data comprises conventional physiological information and electroencephalogram signal data;
the step 2 comprises the following steps:
2.1, respectively inputting the lead data of the electroencephalogram signals acquired in the step 1 into a power spectrum calculation module to perform classical mode decomposition and Hilbert transformation:
taking x (t) as an original electroencephalogram signal, calculating all maximum value points and minimum value points of the signal x (t), and connecting the maximum value points by an interpolation method to form an upper envelope line x 1 (t) connecting minima points to form a lower envelope x 2 (t) obtaining x 1(t) and x2 The average curve of (t) is given by:
calculating a new signal s 1 (t) the formula:
s 1 (t)＝x(t)-w 1 (t)
judging the new signal s 1 Whether all local extremum points and zero crossing points of (t) are identical or differ by at most 1, or s 1 (t) the mean value of the upper and lower envelopes is 0, if not, let x (t) =s 1 (t) repeating the above steps of obtaining the average curve and the new signal until a new signal s is obtained 1μ (t) all local extremum points and zero crossing points are identical or differ by at most 1, or s 1μ (t) the mean of the upper and lower envelopes is 0; to eliminate the riding wave, avoid the IMF component amplitude becoming constant, define the standard deviation threshold SD<0.25, the formula of which is:
wherein mu is the decomposition times when the step of calculating the average curve and the new signal is finished, and T is the signal period;
let c 1 (t)＝s 1μ (t), then c 1 (t) is the first IMF component of the original signal x (t);
solving for a first residual component r 1 (t) the formula:
r 1 (t)＝x(t)-c 1 (t)
judgment r 1 (t) whether it is a monotonic function, if so, ending the EMD decomposition, if not, letting x (t) =r 1 (t) repeating the above steps of obtaining the average curve, the new signal and the residual component to obtain n IMF components c i (t) until the residual component r obtained by the last decomposition n (t) ending the EMD decomposition when it is a monotonic function or less than 0.5;
thus, the original signal x (t) is expressed as:
hilbert transformation is carried out on each obtained IMF component, and the formula is as follows:
wherein δ is the cauchy principal component;
thus, a Hilbert spectrum H (ω, t) is obtained with amplitude over time and instantaneous frequency, the formula of which is:
wherein ,
for instantaneous amplitude +.>
Is the instantaneous frequency;
integrating H (omega, t) with time to obtain a marginal spectrum function P (omega), wherein the formula is as follows:
thereby, the power spectrum value P on the fixed frequency band of each lead under the rest state and the attention task state is obtained 1,k (f)、P 2,k (f) The formula is as follows:
wherein ,P1,k (f) The power spectrum value of the kth lead in the resting state, P 2,k (f) The power spectrum value of the kth lead for the attention task state, f represents frequency, f Σ Representation ofA set of all frequency points within a fixed frequency band, ω=2pi f;
2.2 the total power spectrum value P on the fixed frequency band of each lead obtained in the step 2.1 1,k (f)、P 2,k (f) Inputting the power spectrum value to an optimal lead selection module, calculating the lead contribution degree by adopting an improved distance criterion, and defining a relative distance coefficient beta (k) by using the difference between the power spectrum value in the rest state and the task state of the same frequency band during calculation, wherein the formula is as follows:
wherein m is the number of leads of the electroencephalogram acquisition equipment, and alpha epsilon [0,1] represents the attention degree of the same lead in different states and is used for adjusting the individual specificity caused by the difference of users in the process of calculating the relative distance coefficient;
thus, the lead represented by the maximum value of beta (k) is selected as the optimal lead set k on the fixed frequency band zuiyou (j) J=1, 2,3, … … s, s being the number of β (k) maxima;
2.3 the k obtained in step 2.2 zuiyou (j) Task state power spectrum value on optimal lead
The input deviation calculation module calculates a power spectrum mean value range matrix (I) under the same state as the normal crowd>
The deviation d between the two is expressed as:
wherein m is the lead number of the electroencephalogram acquisition equipment, and a weight factor epsilon is introduced j ∈[0 , 1]Representing the degree of interest for each optimal lead;
thus, the average distance value d, i.e., the deviation value d, is obtained withoutThe absolute value |d| of the same deviation value d is marked as F according to different risk probability values 1 ；
The step 3 comprises the following steps:
3.1, performing Bayesian network learning by utilizing a tabu search algorithm, determining the influence weight of each node of the Bayesian network, and establishing a Bayesian network model, namely searching a network structure which is best matched with a training sample set on the premise of giving one data sample set;
3.2, carrying out parameter learning on the probability of each node by utilizing a maximum likelihood estimation algorithm, namely searching the probability distribution of each node of the network on the basis of sample data, and determining the conditional probability density at each node of the Bayesian network model by utilizing a network topological structure, a training sample set and priori knowledge: each item of conventional physiological information of the user is respectively marked as x 1 ，x 2 ，……x i ,F(x i ) To generate x i Probability of F (x) j |x i ) Represents x i Under the condition x occurs j Is a fall risk probability value F (x|x) 1 ,x 2 ,......x i ) The calculation is as follows:
F(x|x 1 ,x 2 ,......x i )＝∏F(x i |x 1 ,......x i-1 )......F(x 2 |x 1 )F(x 1 )
wherein i, j is the number of normal physiological information, i+.j
3.3, calculating a user falling risk probability value F (x), wherein the formula is as follows:
F(x)＝F 1 ×F(x|x 1 ,x 2 ,......x i )
the fall risk assessment and early warning method comprises the following steps: and (3) performing risk early warning by using the fall risk probability value F (x) calculated in the step (3).
2. A fall risk assessment and early warning method according to claim 1, characterized in that: the step 1 comprises the following steps: (1) The user provides various conventional physiological information including age, sex, height, weight, blood pressure, quantity of motion and sleeping state through the terminal equipment; (2) And acquiring electroencephalogram signal data in a resting state and an attention task state through an electroencephalogram acquisition system.
3. A fall risk assessment and early warning system for performing a fall risk assessment and early warning method as claimed in any one of claims 1 to 2, the system comprising a data acquisition module, a data analysis module, an assessment model module and an early warning module, characterised in that: the data acquisition module is used for collecting physiological information data of the user; the data analysis module is used for integrating and converting data according to the electroencephalogram signal data acquired by the data acquisition module and the prediction analysis requirement, and calculating a risk probability value; the evaluation model module is used for establishing a fall risk evaluation model and calculating a fall risk value; the early warning module is used for early warning.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202010303476.8A CN111481193B (en) | 2020-04-17 | 2020-04-17 | Fall risk assessment and early warning method and system |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202010303476.8A CN111481193B (en) | 2020-04-17 | 2020-04-17 | Fall risk assessment and early warning method and system |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN111481193A CN111481193A (en) | 2020-08-04 |
| CN111481193B true CN111481193B (en) | 2023-05-19 |
Family
ID=71789924
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202010303476.8A Active CN111481193B (en) | 2020-04-17 | 2020-04-17 | Fall risk assessment and early warning method and system |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN111481193B (en) |
Families Citing this family (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| TWI809612B (en) * | 2021-12-21 | 2023-07-21 | 臺中榮民總醫院 | Fall Assessment Risk Warning System |
| CN117743810B (en) * | 2024-02-21 | 2024-04-30 | 莱凯医疗器械(北京)有限公司 | Intended understanding analysis method for electroencephalogram signal monitoring |
Citations (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN104586398A (en) * | 2013-10-30 | 2015-05-06 | 广州华久信息科技有限公司 | Old man falling detecting method and system based on multi-sensor fusion |
Family Cites Families (10)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US7612681B2 (en) * | 2007-02-06 | 2009-11-03 | General Electric Company | System and method for predicting fall risk for a resident |
| US10709376B2 (en) * | 2015-06-02 | 2020-07-14 | Koninklijke Philips N.V. | System for supporting an elderly, frail and/or diseased person |
| EP3317630A4 (en) * | 2015-06-30 | 2019-02-13 | ISHOE, Inc | Identifying fall risk using machine learning algorithms |
| WO2017118610A1 (en) * | 2016-01-07 | 2017-07-13 | WOLFGANG, Müller-Adam | Method and device for detecting a fall |
| CN105933080B (en) * | 2016-01-20 | 2020-11-03 | 北京大学 | Fall detection method and system |
| US20170258390A1 (en) * | 2016-02-12 | 2017-09-14 | Newton Howard | Early Detection Of Neurodegenerative Disease |
| US10319209B2 (en) * | 2016-06-03 | 2019-06-11 | John Carlton-Foss | Method and system for motion analysis and fall prevention |
| EP3346402A1 (en) * | 2017-01-04 | 2018-07-11 | Fraunhofer Portugal Research | Apparatus and method for triggering a fall risk alert to a person |
| US11158179B2 (en) * | 2017-07-27 | 2021-10-26 | NXT-ID, Inc. | Method and system to improve accuracy of fall detection using multi-sensor fusion |
| CN109864745A (en) * | 2019-01-08 | 2019-06-11 | 国家康复辅具研究中心 | A kind of novel risk of stroke appraisal procedure and system |
-
2020
- 2020-04-17 CN CN202010303476.8A patent/CN111481193B/en active Active
Patent Citations (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN104586398A (en) * | 2013-10-30 | 2015-05-06 | 广州华久信息科技有限公司 | Old man falling detecting method and system based on multi-sensor fusion |
Also Published As
| Publication number | Publication date |
|---|---|
| CN111481193A (en) | 2020-08-04 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN108416367B (en) | Sleep staging method based on multi-sensor data decision-level fusion | |
| CN103294199B (en) | A kind of unvoiced information identifying system based on face's muscle signals | |
| Bairy et al. | Automated diagnosis of depression electroencephalograph signals using linear prediction coding and higher order spectra features | |
| CN111481193B (en) | Fall risk assessment and early warning method and system | |
| CN109009102B (en) | Electroencephalogram deep learning-based auxiliary diagnosis method and system | |
| CN112426160A (en) | Electrocardiosignal type identification method and device | |
| CN113017645B (en) | P300 signal detection method based on void convolutional neural network | |
| CN111407243A (en) | Pulse signal pressure identification method based on deep learning | |
| CN111436929A (en) | Method for generating and identifying neurophysiological signals | |
| CN113069117A (en) | Electroencephalogram emotion recognition method and system based on time convolution neural network | |
| CN116058800A (en) | Automatic sleep stage system based on deep neural network and brain-computer interface | |
| CN111053552B (en) | QRS wave detection method based on deep learning | |
| CN106845348B (en) | Gesture recognition method based on arm surface electromyographic signals | |
| CN113951868B (en) | Method and device for detecting man-machine asynchronism of mechanical ventilation patient | |
| CN117457229B (en) | Anesthesia depth monitoring system and method based on artificial intelligence | |
| CN110916672A (en) | Old people daily activity monitoring method based on one-dimensional convolutional neural network | |
| CN113974655A (en) | Epileptic seizure prediction method based on electroencephalogram signals | |
| CN110547806B (en) | Gesture action online recognition method and system based on surface electromyographic signals | |
| CN112617833A (en) | Device for detecting depression based on resting brain waves | |
| CN210697629U (en) | Signal acquisition device, mobile terminal and signal analysis system | |
| CN111783719A (en) | Myoelectric control method and device | |
| CN116019467A (en) | Feature extraction method of facial nerve electromyographic signals | |
| CN116030937A (en) | Method for generating running exercise prescription | |
| CN112084935B (en) | Emotion recognition method based on expansion of high-quality electroencephalogram sample | |
| CN114469041A (en) | Heart rate change data characteristic analysis method in exercise process |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |