ES2415579B1 - PROCEDURE AND MULTIAGENT ELECTRONIC DEVICE WITH DIFFUSIVE CONTROL FOR THE DETECTION OF THE SIDE CEREBRAL ICTUS. - Google Patents

Abstract

Multiagent electronic procedure and device with diffuse control, which helps in the detection of stroke or stroke (CVA), specifically through the detection of one of its main symptoms: lack of mobility in the opposite body to the location stroke of the stroke.

Description

PROCEDURE AND MULTIAGENT ELECTRONIC DEVICE WITH

DIFFUSIVE CONTROL FOR THE DETECTION OF CEREBRAL ICTUS

SIDE

5

DESCRIPTION

1 o

It is a multi-agent electronic device with diffuse control, which helps in the detection of stroke or stroke (CVA), specifically through the detection of one of its main symptoms: the lack of mobility in the opposite body to the cerebral location of the stroke.

Object of the invention.

15 20 25

  The object of the invention is to provide a multiagent electronic device and method with diffuse control that allows the detection of lateral cerebral stroke or stroke (CVA), from a solution that takes into account the detection of one of the main symptoms, which It is the beginning of hemiplegia. This bodily disorder is characterized by presenting paralysis in the lateral half of the body, so that the detection of this body is adjusted to an analysis of the acceleration and angular velocity in the movements of each upper limb. For this, it proposes a procedure and a multi-agent electronic device with diffuse control for the detection of cerebral stroke on a person who can perform movements with their upper extremities, where the detection is carried out by means of two electronic means of acquisition, acceleration and speed angular movements in the extremities, adapted to be located in the upper extremities of the person. The analysis is carried out by means of two electronic processing means, of the primary evaluation that consists in obtaining the parameters of the maximum module and the average movement module in both extremities; where the modulus of motion in each limb is the difference between the position vectors of two consecutive movements at moments t and t + 1; where the position vectors are

5 1O

calculated from acceleration and angular velocity in the movements of each limb; and where the maximum module and the average movement module at each limb is calculated by sampling intervals that can be adjusted by the processing means. The two electronic processing means are adapted to be located in the upper extremities of the person. The generation of an alarm signal is performed as a result of the secondary evaluation on electronic or computer processing means using the maximum module and the average movement module at each limb; which is determined by two diffuse Mandami controllers that take as input into one of them the maximum module of movement in each limb and in the other, the average module of movement also in each limb, to output the percentage of alarm that an Ictus is being produced, calculated from the fact that the output on both controllers exceeds an adjustable threshold for a period of time also adjustable, by the processing means.

fifteen

Other aspects of the invention are defined in the dependent claims.

Background of the invention.

20 25

Stroke is a cerebrovascular disease that affects the blood vessels that supply blood to the brain. It occurs when a blood vessel that carries blood to the brain ruptures or is plugged by a clot or other particle. Due to this rupture or blockage, part of the brain does not get the blood flow it needs. The consequence is that nerve cells in the affected brain area do not receive oxygen, so they cannot function and die after a few minutes. In general, strokes are sudden onset and rapidly developing, and cause brain injury in minutes. Depending on the area of the brain affected, many different symptoms may occur: sudden numbness or weakness in the face, arm or leg, especially on one side of the body.

  Usually, the doctor can diagnose a stroke through the history of the facts and the physical examination. The latter helps the doctor to

Determine where the brain injury is located. Imaging tests such as a computerized tomography (CT) or magnetic resonance imaging (MRI) are also usually done to confirm the diagnosis, although these tests only detect the stroke after a few days have elapsed.

5

The methods that are commonly used for the detection of cardiovascular accidents are methods based on the analysis of certain markers such as proteins or molecules that the brain releases during an episode of ischemia or cerebral hemorrhage.

1 o

An early diagnosis of a stroke is essential in preventing the injuries caused by this lack of oxygenation in the brain to be more serious and irreversible.

fifteen

Several inventions have been located that seek to detect and / or alleviate the consequences of cerebral infarction, but all of them do so through chemical substances or proteins generated by the organism. For example, the PCT application ES2006000206 uses Hydroxytyrosol in the prevention and treatment of stroke. The invention W02009075863 detects stroke by identifying the presence or absence of one or more markers associated with the stroke. And the application US2008254495

diagnose stroke by using polyamines and acrolein content, polyamine oxidase or protein activity thereof.

twenty

Description of the drawings.

Figure 1 shows the block diagram of the transmitter system.

Figure 2 shows the block diagram of the receiving system.

25

Figure 3 computer. It represents he scheme from the license plate from communication with he means, medium

  Figure 4 schematically represents the communication between the transmitting system and the receiving system.

Figure 5 schematically represents the communication between the computing medium and the systems.

Figure 6 is a vector representation of the physical quantities used in the primary evaluation.

Figure 7 represents a diagram of diffuse reasoning, graphically explaining how uncertainty is assigned to each linguistic label of the output variable.

Figure 8 represents a diagram of the inputs of the diffuse control and the output obtained from it.

Description of the invention

The invention we propose is based on the detection of hemiplegia caused by the lack of blood supply to a part of the brain, based on the detection of the most common symptom in cerebral stroke, such as the lack of mobility in a part of the body.

This lack of mobility increases as time goes by, and we seek early detection and intervention to avoid major sequelae. So that a person who is intervened in the first hour since the stroke occurs, would be virtually without sequelae, and if it is intervened in the first three hours, the sequels are minimal.

With this procedure and device it is intended to detect all strokes that are lateral, since in the case that it was occipital or frontal, it does not generate motor difficulties in the opposite body to the stroke. However, the percentage of stroke cases, which are lateral, is in 70-80% of the cases, compared to the rest that are frontal or occipital.

The multi-agent electronic device with diffuse control for the detection of lateral cerebral stroke in a person who can perform movements with their upper extremities, will be composed of:

Two means of electronic acquisition, acceleration and angular velocity of

limb movements adapted to be placed on the extremities

superiors of the person.

Two electronic processing means, to obtain the maximum parameters

5

module and the middle movement module in both extremities; where the module

of movement in each limb is the difference between the position vectors of two

consecutive movements at moments t and t + 1; where the position vectors are

calculated from acceleration and angular velocity in the movements of each

end; and where the maximum module and the average movement module in each

1 o

Tip is calculated by sampling intervals that can be adjustable by

processing means The two electronic processing means are

adapted to be placed in the upper limbs of the person.

The programming means are electronic processing means or

computer, equipped with two diffuse Mandami controllers, which take as

fifteen

entry into one of them the maximum movement module in each limb and in the

another the middle module of the movement also in each limb to output

the percentage of alarm that an Stroke is occurring, calculated from which the

output on both controllers exceeds a threshold for a period of time also

adjustable, by processing means. Preferably, the threshold will be 80%

twenty

for a period of 15 minutes and samples will be taken every 30 seconds.

Thus the device will be composed of two systems that integrate the means of

Previous processing, one for each arm. For the evaluation they must be taken into

counts the signals of both systems equally, therefore the communication between them is

direct. With this, as a final goal, the activation of the alarm signal is established, of a

25

Reliable way, but with deactivation capability. These alarms must be

easily identifiable by any possible user.

The device has a design fully adaptable to the human body.

As the device must have continuous use for the detection to cover all

  your manifestation possibilities, preferably you will have an energy autonomy of

24 hours and a design adaptable to different situations of the user's daily life, from rest, to sports activities.

The communication will be obtained wirelessly, and the power supply will be done by rechargeable batteries, whose capacity allows the device to operate without interruption during a daily cycle of human activity.

The device can operate autonomously or in a supervised way; In the latter way, the alarm signal will be transmitted to a personal computer, portable device, tablet, etc., so that different users can be monitored and manage their respective alarm signals.

Autonomous operation

The elements that make up the device are presented in Figures 1 and 2 respectively:

Transmitter system (1): it is a printed circuit board (PCB) (1 O) made specifically to measure the motor of one of the upper body extremities. Specifically, being placed on the wrist, its movement is measured. It is included in a case (15) with a strap that acts as a bracelet. This includes a rechargeable battery (13) with the PCB (10).

The printed circuit board (1 O) has as main elements the microcontroller (9) acting as a brain; an inertial measurement unit (8), consisting of two sensors: an accelerometer (6) and a gyroscope (7); and a radio frequency transmission module (3). This module will be the direct communication channel with the receiving system (2). Other elements that make up this PCB are the battery charger (11), the display element (14) (LED), and the micro USB power connector (12), the latter 2 being the only externally visible elements.

Receiver system (2): This element with receiver functions, is a device identical to the transmitter system (1), again composed of a PCB (18), a battery

rechargeable (13) and container housing (15). The PCB (18) includes many of the same sub-elements and functions as the sender PCB (9), added to others of the receiver.

5

The system has an inertial measurement unit (8), to measure the 2nd arm in the same way, a radio frequency reception module (4) replacing the previous transmitter, both modules being in operation, and a microcontroller (9) that for its part, is responsible for the evaluation of the motor movements of both transmitter (1) and receiver (2) elements.

1 o

In the microcontroller (9) of the receiving system (2) is where the Mandami type control is implemented, which evaluates the alarm index, from which the ultimate goal is achieved, the activation and deactivation of the alarm. The elements arranged for the alarm signal are included in its own PCB (18), a sound transducer (17) (which acts as a buzzer) and an LED-type indicator (14), or the like.

fifteen

Supervised operation

20 25

  It has the same elements and functions as autonomous operation, except that the receiving system (2) acts as a transceiver with the computer medium (5) (personal computer, portable device, tablet, etc.) and also has a monitoring software and alarm management with diffuse control models, for each user. Within this system we can find, figure 3, a PCB (21) with Zigbee communication (19) that allows, through its UART-USB interface (20) and the USB connector (12), to receive in a computer medium (5 ) the data of the first evaluation - primary evaluation - of each of the limbs of each user connected to the system. Each user will have a diffuse model associated with it, so that when the alarm occurs it will be sent to the receiving system (2), which will act as a transceiver with the computer medium (5). Both systems in this case, are identical to those of autonomous operation unless there is no communication between them and both are connected to the computer medium (5), by means of Zigbee communication. The system called sender (1) only has communication

unidirectional and the system called receiver (2) will act as a transceiver with the computer medium, which can be a PC, or another system (5) - two-way communication.

Description of a preferred embodiment.

5 1 O

The physical magnitude acceleration and angular velocity are captured through 2 sensors, accelerometer (6), and gyroscope (7), respectively. Both use MEMS (Microelectromechanical Systems) technology in their operation, with which we have in the same integrated circuit a great capacity to detect instrumentation variables. Thus, the accelerometer (6) obtains the acceleration in 3 axes, X, Y, Z, and the gyroscope (7) the angular speed in 2 axes, X, Y.

fifteen

The set of both forms what is called an Inertial Measurement Unit (IMU) (8) of 5 degrees of freedom. These spatial references in axes, it should be noted that they always occur with respect to the same integrated circuit (IC) that is evaluating the magnitude. The IMU set (8) of each system is placed at each end with the axes of the device in the same direction.

twenty

For the control of said signals, the microcontroller (9) is available. The modules included in the microcontroller will be key when it comes to achieving it, such as the digital analog converter (ADC) (16) present, with which the controller will directly obtain digital signals. With the voltage obtained Vout of each output of the sensors, we can know the real physical magnitude, according to their calibration curves.

25

Thus, for the 3 outputs, corresponding to the 3 axes of the accelerometer, we have the equation (1), and for the 2 outputs of the gyroscope the equation (2); where S is the sensitivity of the sensor, a is the acceleration, w the angular velocity, VouT the output voltage of the different sensors, VZEROg the gyro output voltage with angular velocity O and VzERO the accelerometer output voltage with acceleration O .

VouT = a * S + VzEROa

(one)

  VouT = 1ll * S + VzEROg

(2)

equation (4). For this, the Rgiro vector, obtained from Rest (n -1), will be calculated beforehand.

and angular velocities on the corresponding axis Wx and Wy.

(Racc + Rgiro · fJ)

R () = -'------- = ----'--- '

e st n (4) (1 + fJ)

5 As an example, the calculations to obtain Rxgiro are presented, which are deduced from Figure 6, and are presented from equations (5) to (9), where T

It is the sampling period.

R R

tan (A xz) = __x_ => Axz = arctan (__ x_) (5) Rz Rz

_ 1) - (Rxest (n-1))

Axz (n -arctan - ::: ======= :: (6)

Rzest (n-1)

(7)

(8)

. one

Rxgrro = ---; :: =========== (9) ~ 1 + coe (Axz (n)) · sec \ Ayz (n))

To perform the fuzzy control evaluations, the device will need a wireless communication - Figure 4 - between the transmitter system (1) and the receiver system (2). A radio frequency (RF) communication is established through a pair of transmitter-receiver modules (3) and (4) with FSK (frequency jump modulation) functions. The transmitter module (3) will be in charge of continuously sending information about its own primary motor assessment of the corresponding arm. Through SPI communication between the microcontroller (9) and the 20 RF modules (3) and (4) (master-slave in that order), you can configure the

FSK function parameters of the modules. The established communication is therefore of the unidirectional SPI type in the Transmitting System (1), and in the Receiving System (2).

image 1

The primary evaluation is carried out in each system separately and consists of

obtain the maximum value of the movement module and the average movement module that is calculated as the average value of the movement. These calculations will be carried out at intervals, which in the preferred embodiment are set to 30 seconds, however this interval is parameterizable and can be adjusted depending on the sensitivity that is to be obtained therein. The moving module M is calculated according to equation (1 0), where Rest (n) and Rest (n-1) are normalized, with respect to themselves.

2 2 2

M = ~ (Rest x (n) -Rest x (n -1)) + (Rest y (n) -Rest y (n -1)) + (Rest z (n) -Rest z (n -1))

(10)

The secondary evaluation is carried out using a diffuse method of Mandami, figure 8, which is used twice over two sets of inputs: maximum module and average module of movement in both extremities. The fuzzy inference process consists of several stages: fuzzification, background aggregation, involvement, aggregation and defuzzification.

The first step is to fuzzify the input variables, that is, calculate for the degree of belonging to each fuzzy set defined for that variable. Thus, it can be observed -figure 7-, that if the variable x has value xO and the variable y has value yO, then we can affirm that: the degree of belonging of xO to Rl.l is O, the degree of belonging of xO to R2.1 is O, the degree of membership of xO to R3.1 is 0.37, the degree of membership of yO to Rl.2 is 0.66, the degree of membership of yO to R2.2 is O, The degree of membership of yO to R3.2 is 0.66 and so on.

The second step, the aggregation of the antecedents, is carried out as a rule. An operator is applied to the degrees of belonging corresponding to the background. Typically, the operator can be the minimum -of the variables-, the maximum -of the variables-, etc. In the embodiment, figure 7, the minimum has been used. Then the output for rule R3 has the minimum value between 0.37 and 0.66, in our case this minimum is 0.37. In the same way we proceed with the other rules.

The third step is the implication, represented in the figure as the operator ® the minimum operator - and a weight W. In the embodiment all weights have the same value; It is important to state that the result of the implication is a diffuse set with a new membership function.

The fourth step is the aggregation of consequents represented by the EB symbol in the figure. This step results from accumulating by means of an operator as the maximum or the arithmetic sum the implications of each of the rules, and also represents a fuzzy set. In the embodiment, figure 7, is the aggregation of the sets defined by the membership functions associated with rules R3 and R6 by the maximum operator.

Finally, in the fifth step, the defuzzification operation is performed. This operation consists in obtaining a measure of the diffuse set obtained by the aggregation of consequents, this measurement being a real value. Typically, measures such as the center of mass, or the average of the maximums, are applied to obtain the final output value of the diffuse inference. In this embodiment the masaz center has been used.

The fuzzy controller of the present invention uses symmetric Gaussian membership functions, background aggregation through the minimum operator, involvement with the minimum operator and with unit value weights, consequent aggregation, defined by the maximum operator and defuzzification through the center of masses

The secondary evaluation uses two fuzzy controllers, in the number the input variables to the fuzzy controller are the maximum movement module of the right hand (22) and the maximum module of movement of the left hand (23); while in the second, the input variables to the fuzzy controller are the average right hand movement module (22) and the middle left hand movement module (23).

For the realization, the following fuzzy sets have been chosen for each

5

one of the input variables -symmetrical Gaussian-: Low Value = gaussmf (0.75.0), MEDIUM = gaussmf (1.5.5) and HIGH = gaussmf (2.10). The variable, only, of exit is the percentage of alarm of suffering cerebral stroke (24), for the next 20 seconds. It is a variable that as an example in the embodiment can be characterized by the following fuzzy sets of triangular type, figure 7:

LOW value = trimf (0.15,30), MEDIUM = trap (20, 55, 90), HIGH = trap (80, 100, 120).

1O

It is important to state that these values parameterize the membership functions and acquire relevance in training and the final adjustment of the controllers depending on the type of user.

fifteen

The set of rules to be used in the embodiment are those indicated in the double entry table 1, where the content in each cell corresponds to the fuzzy set of the output variable that each rule proposes.

Right hand movement module

LOW

MEANS, MEDIUM TALL

Right hand movement module

LOW Rl.O = LOW R5.0 = MEDIUM R2.0 = High

MEANS, MEDIUM

R6.0 = MEDIUM R4.0 = Low

TALL

R3.0 = High

Table 1

The establishment of an alarm based on the evaluation and detection of hemiplegia is

20 controlled and implemented in the receiving system (2), this is where the fuzzy controller is implemented and where the alarm is diagnosed. In the preferred embodiment, the alarm will occur when the system recognizes that there is a possibility of a stroke alarm, for a time exceeding 15 minutes, which implies an evaluation of 30 cycles with alarm capacity. The reset of this alarm possibility is

25 will produce if once this is detected, there are two cycles with a low level of

alarm within 30 necessary to set the stroke alarm. Alarm level

in each cycle it is determined when the two fuzzy controllers used determine

an alarm level greater than a setpoint value -threshold-, which in the embodiment has been

considered 80%. It is necessary to state that both the alarm level and threshold

5

how The number of cycles is configurable and can be adjusted according to the

sensitivity that you want to get in it.

In addition to autonomous operation

operation can be set

supervised, this part of operation requires both systems to communicate

directly

with the middle Person who is dedicated to computer science (5). This operation no modify the

1 o

characteristics of data acquisition and preprocessing of the information it performs

every

system in the evaluation primary. But be difference of the functioning

autonomous in that the secondary evaluation is carried out in the computer medium (5). So

Each user will have associated a diffuse model, customized to their needs. This

It is suitable when users to monitor may have lacks in movement

fifteen

normal of some limb and the model to be obtained has to be trained from

The data itself. The establishment of the alarm and the way in which it is carried out, is

identical to

in autonomous operation, with the difference that this management be

performed in the computer medium (5).

twenty