DE102012004506A1 - Measuring- and evaluation system for evaluating mobility and performance of subject, has determined sensors which generate amount of data which is assigned to certain sensor and is transferred to processing unit - Google Patents

Abstract

The measuring- and evaluation system has determined multiple sensors which generate an amount of data which is assigned to certain sensor and is transferred to a processing unit to generate another amount of data from the former amount of data by using measures. The latter amount of data is substantially lesser than the former amount of data. The amount of data consist of indices. The low-pass filtering is a Savitzky-Golay-filtering. An independent claim is included for a measuring- and evaluation method.

Description

The present invention is a measurement and assessment system for assessing the mobility and performance of a subject.

The measurement and assessment system according to the invention is based on a known system for measuring the mobility of the human spine, as shown in the DE 10 2008 052 406 is described. A comparable method is not known in the prior art.

However, the known system for measuring the mobility of the human spine measures only the movement of the spine, so that statements about mobility and performance is not possible. But judging the mobility and efficiency of the spine is for some sports, such as gymnastics, athletics or football essential criteria.

The object of the present invention is therefore to provide a system and method by means of which it is possible to assess the spinal column and to be able to make statements about its mobility and performance.

This object is achieved by the measurement and assessment system or method according to the invention with the features according to claim 1 or claim 10. Advantageous developments are characterized in the subclaims 2 to 9. The measurement and evaluation method according to the invention uses the measurement and assessment system according to the invention. The following description of the system according to the invention and of the individual features, elements, components and functions, in particular the sequence of functions, therefore applies mutatis mutandis expressly also for the inventive method, without express reference to the respective use as a process or functional step below.

The measurement and evaluation system according to the invention consists of a certain number of sensors that generate a first set of data and that data are assigned to a particular sensor, which data is transmitted to a processing unit to obtain from the first set of data by applying measures generating a second set of data, wherein the second set of data is substantially smaller than the first set of data.

Advantageously, the second set of data consists of indices.

• – Anwendung einer Tiefpassfilterung auf die erste Menge von Daten;
• – Berechnung von spezifischen Merkmalen;
• – Anwendung einer Regressionsanalyse und/oder Faktorenanalyse.

Advantageously, the measures consist of the following steps:

• Applying low pass filtering to the first set of data;
• - calculation of specific characteristics;
• - Application of a regression analysis and / or factor analysis.

Advantageously, the low pass filtering is Savitzky-Golay filtering.

Advantageously, in computing the specific features, the mean and standard deviation are modeled for all features.

Advantageously, other indices are taken into account in averaging.

Advantageously, the index (s) is taken from the group of D-MCI and / or D-DCI and / or D-AI and / or D-MCUI.

Advantageously, features consist of raw data describing the mobility of a subject.

• • Gesamtindex (ELDI)
• • Gesamtindex Kurzzeit (ELCI)
• • Beweglichkeit-Index Kurzzeit (D-MCI/C-MCI)
• • Dynamik-Index Kurzzeit (D-DCI/C-MCI)
• • Aktivität-Index Langzeit (D-AI)
• • Beweglichkeit-Index Langzeit (D-MCI)
• • Dynamik-Index Langzeit (D-DCI)
• • Vergleich-Index KZM/LZM Index (D-MCUI)

Advantageously, the following indices are formed:

• • Overall Index (ELDI)
• • Overall Short-Term Index (ELCI)
• • Agility Index Short Time (D-MCI / C-MCI)
• • Dynamic Index Short Time (D-DCI / C-MCI)
• • Activity Index Long Term (D-AI)
• • Flexibility index long-term (D-MCI)
• • Dynamic index long-term (D-DCI)
• • Comparison Index KZM / LZM Index (D-MCUI)

The measurement and assessment system according to the invention serves, for example, for detecting the functional capacity of the lumbar spine. It consists of two flexible, elongated known from the prior art sensors, which are attached by means of special plaster parallel to the spine. The sensors are operatively connected to a processing unit. The measures can be carried out in the processing unit and / or in a computer. According to the invention, the processing unit can be a computer.

The distance between the two pilasters to the spine is advantageously between 3 cm and 7 cm, preferably about 5 cm. A reference line on the patches or sensors is based on the height of the spina ilaca posterior superior. Thus, the first sensitive segment is at height S1, so that the range of movement of the spine is measured from the transition S1-L5 in the cranial direction. The position of the upper segment is dependent on the height and deflection and therefore variable.

The sensors each detect the deflections of the sensor in the sagittal plane in 12 equivalent areas. In addition, 3-axis acceleration sensors are positioned at both ends of both sensors to sense the alignments to the Earth's gravity field.

If deviations in certain measurement segments occur in a measurement compared to the norm values of healthy subjects, the question arises as to which spine segment corresponds to it. By means of a developed model it is possible to assign the sensor segmental measured movements and dynamic parameters, such as speed and acceleration, to individual lumbar vertebrae.

For movements in the sagittal plane (flexion and extension) both sensors are deflected evenly and the deflections can be calculated from the bending sensors. For movements outside the sagittal plane, the dimensions of lateral flexions can be determined from the acceleration sensors and those from rotations to the asymmetric deflections of both sensor straps.

1 Data to be collected

In order to be able to derive the position of the individual vertebral bodies from one another, the sensor bands of the measurement and assessment system according to the invention are to be projected onto the spinal column. Since uniform, equally long sensors are used for all subjects of different sizes, the distance between the spinous process of the vertebra prominens (C7-7 cervical vertebrae) and the posterior superior spine (SIPS) in the upright position is measured in all subjects. The flexible measuring tape must rest against the back surface over the entire length. The same distance is additionally recorded in the same way with maximum flexion and extension of the spine. These values are a measure of the extent of possible extension and compression of the spine.

Subsequently, the test person with the applied sensor performs a uniform short-term choreography (optionally while standing, sitting or lying down), which is transmitted by Bluetooth (R) through a processing unit to a computer. The choreography consists of repeated flexions, extensions, rotations (right and left) and lateral flexions (right and left).

2 projection in the sagittal plane

Since the sensor segments move relative to the spinal column, the projection is first applied to the straight line and then to the maximum deflections of flexion and extension. Since both sensors are glued symmetrically, it can be assumed that both sensors are deflected equally during movements in the sagittal plane. Therefore, the measurement results of the left and right sensors are averaged.

Based on morphological studies, as out Panjabi MM, Goel V, Oxland T. Takata K, Duranceau J, Krag M. Price M. Human Lumbar Vertebrae-Quantitative three-dimensional anatomy. Spine. 17 (3): 299-306, March 1992 ; Panjabi MM, Takata K, Goel V, Federic D, Oxland T, Duranceau J. Krag M. Thoracic Human Vertebrae-Quantitative Three-Dimensional Anatomy. Spine. 16 (8): 888-901, Aug 1991 ; Pooni JS, Hukin's DWL, Harris PF, Hilton RC Davies KE. Camparison of the structure of human intervertebral discs in the cervical, thoracic and lumbar regions of the spine. Surgical Radiologic Anatomy. 8: 175-182, 1986 ; Amman C, Naderi S, Kiray A, Aksu FT, Yilmaz HS, tetik S, Korman E. The human sacrum and screwdriving approaches for screw placement. Journal of Clinical Neuroscience. 16: 1046-1049, 2009 Assumptions were made for the percentage distribution of vertebral and disc heights between C7 and SIPS.

Under the assumption that the distance between sensor tapes and spinal column is constant, a parallel curve can thus be constructed parallel to the shape of the sensor tapes. This has the property that the vertical distance to the sensor curve is always constant - for example - 4.5 cm and forms in the spine model, the back of the actual vertebral body (without spinous processes). From this parallel curve go perpendicular to the base and cover plates of the individual vertebra. The distances of the vertical lines on the parallel curve to one another result from the assumed percentage distribution of the vertebrae and intervertebral discs between C7 and the SIPS. For this purpose, a vertical line is removed at the lower end of the parallel curve, which forms the base plate of the vortex S1. The next line, perpendicular to the base plate of S1, forms the top plate of said vortex, the distance (arc length on the parallel curve) corresponding to the percentage of S1 (6.25%) of the distance C7-SIPS in the straight state.

Subsequently, the base and cover plates of the other vortices are constructed accordingly. The distance between the vertebrae corresponds to the percentage of each disc between the vertebrae. Thus, the location of the individual sensor segments with respect to the spinal column segments is determined.

Due to the relative movement between the spinal column and the sensor bands due to the extension and compression of the spinal column during movement, the projection of the upright posture during movement is no longer consistent.

Since the movement at maximum flexion and extension mainly comes from the lumbar spine, it is assumed that the extension or compression of the spine at maximum flexion or extension distributed to the intervertebral discs in the lumbar column area. This means that a parallel curve is constructed for the maximum deflections, as in the straight position. The difference is then only at the distance of the vertical lines to each other. While the height of the vertebral bodies remains constant, the height of the intervertebral discs varies as a function of the deflection and the percentage distribution (see , left and right).

The information about the tilt of the first sensor segment to the earth's gravity field is determined from the lower acceleration sensors.

From the position of the individual vertebrae to each other, the angle between the individual vertebrae can be determined.

The distance between C7 and the SIPS can be interpolated for deflections between the even state and the maximum displacements for each time point. Thus, the projection of the sensor tapes can also be implemented outside of the three described positions and the deflection of the vertebrae at each measurement time take place.

3 rotation

The determination of rotational movements is based on the asymmetrical deflections of the right and left sensor in a lateral view, since they can detect the deflection only in the sagittal plane. In this case, the extent of the angle differences serves to determine the extent of the achieved rotation. The sign of the difference provides information about the direction of the rotation.

Since the approximate position of the sensor segments to the vertebral bodies is known from the sagittal projection, the extent of the rotation in the individual sensor segments can be transmitted to the individual vertebrae by interpolation. The calculation is possible as with flexion and extension at all measuring times.

4 lateral flexion

The maximum angle of inclination, which is achieved in the Lateralflexion is determined via the deflections of the upper acceleration sensors in the respective lateral direction. It means that the information about the lateral inclination refer to the total length of the sensor and not to the individual measurement segments.

In order to eliminate compensatory movements with the pelvis, the lateral deflections from the lower acceleration sensors are determined and subtracted from the lateral inclination, which is determined with the upper acceleration sensors. The approximate position of the upper acceleration sensors with respect to the vortex segments is determined from the above-described projection in the sagitta plane. An approximately vortex accurate determination of the range of motion of individual vertebrae can then be based on Grieve (see , right) are determined by a distribution of the total lateral inclination percentage of the covered vertebral bodies. Thus, even with lateral flexions, the extent of movement for each vertebral body is approximately determined at each measurement time.

The present invention will be explained in more detail with reference to a preferred embodiment with reference to figures and diagrams. In particular, the measures according to the invention are explained in detail. Show it:

: Lateral view of the sensor bands and the spine (lowest vertebra S1) in maximum flexion, upright posture and maximum extension;

: Range of movement of the individual vertebrae during rotation and lateral flexion (after Grieve, 1981 );

: Schematic representation of the sensor tapes (top view on the back);

: Structure of the indices

: Schematic representation for the calculation of the factor matrix and the regression coefficients

: Schematic representation of the calculation of the short-term index ELCI

: Schematic representation of the calculation of the long-term ELDI index

: Schematic representation of the calculation of the index D-AI

: Schematic representation of the calculation of the index D-MCI from standardized features (see and )

: Schematic representation for the calculation of standardized range-of-motion long-term features of the index D-MCI

: Schematic representation for the calculation of standardized range-of-motion short-term features of the index D-MCI

: Schematic representation of the calculation of the index C-MCI from standardized features (see )

: Schematic representation of the calculation of the index D-DCI from standardized features (see and

: Schematic representation for the calculation of standardized range-of-dynamic short-term characteristics of the index D-DCI

: Schematic representation for the calculation of standardized range-of-dynamic long-term characteristics of the index D-DCI

: Schematic representation of the calculation of the index D-DCI from standardized features (see )

: Schematic representation of the calculation of the index D-MCUI

: Schematic representation of the kyphosis angle, lordosis angle and kyphosis-lordosis transition.

5 dynamics

Based on the determination of the vertebral body deflections at each measurement time, conclusions about the achieved movement velocities α 'and accelerations α''of the individual vertebral bodies in all three spatial directions are also possible from the derivation over the measurement time t: α '= Δα / Δt α '' = Δα '/ Δt

6 Benefits of the spine model

The spine model of the measurement and assessment system according to the invention makes it possible to determine the position and movements of the individual lumbar vertebrae both in real time and for up to 24 hours without restriction and in a non-invasive manner. This makes it possible to differentiate the movements of the lumbar spine and to make statements about the duration, frequency and extent of the various movements in the long-term measurement. The extent of the movements can be related to the results in the short-term measurements as a percentage or specified in absolute terms.

7 sensors

With the measurement evaluation system according to the invention, it is possible to generate a small number of indices from a large number of data of different informative value and quality, which allow a specific significance for certain processes in the musculoskeletal system of humans.

However, these indices are not diagnostic and are not tools for providing a diagnosis, or are intended to precede the diagnosis. Rather, the indices allow the operator to judge the spinal movement only. Neither to operate nor to evaluate the measurement and assessment system according to the invention, a medical education or a medical certification is necessary.

One possible application of these indices is in the performance assessment of athletes, as the development of the indices can be considered over a period of time, thus providing the basis for an objective assessment of mobility and performance.

Nevertheless, it is also possible to use the indices for medical assessments. In terms of patent law, the scope of protection should therefore not be limited to non-medical utilization options.

R1 bis R12:

1. bis 12. Messzone des rechten Sensorbandes

L1 bis L12:

1. bis 12. Messzone des linken Sensorbandes

RX:

X-Achse des Beschleunigungssensors des rechten Sensorbandes

RY:

Y-Achse des Beschleunigungssensors des rechten Sensorbandes

RZ:

Z-Achse des Beschleunigungssensors des rechten Sensorbandes

LX:

X-Achse des Beschleunigungssensors des linken Sensorbandes

LY:

Y-Achse des Beschleunigungssensors des linken Sensorbandes

LZ:

Z-Achse des Beschleunigungssensors des linken Sensorbandes

The sensors on the sensor bands are named according to the following nomenclature.

R1 to R12:

1st to 12th measurement zone of the right sensor band

L1 to L12:

1st to 12th measurement zone of the left sensor band

RX:

X-axis of the acceleration sensor of the right sensor band

RY:

Y-axis of the acceleration sensor of the right sensor band

RZ:

Z axis of the acceleration sensor of the right sensor band

LX:

X-axis of the acceleration sensor of the left sensor band

LY:

Y-axis of the acceleration sensor of the left sensor band

LZ:

Z axis of the acceleration sensor of the left sensor band

In addition to the variables, there is a time component in many calculations. This is given as a subindex. An exemplary embodiment of the sensors is shown. The processing unit according to the invention is in Not shown.
L1 _j = jth measuring point of the 1st measuring zone of the left sensor band since the beginning of the file.

7.1

The obtained data of the individual short-term measurements are constructed in the form of two-dimensional matrices. These consist of 30 columns and varying numbers of rows. Since the measurement and evaluation system according to the invention measures at 50 Hz, 50 lines each describe one second. In the 30 columns, the first 12 columns describe the measurement results of the right sensor (lowest segment in the first column, etc. [R1 to R12]). In columns 13 to 24 are the segments of the left sensor, where the arrangement corresponds to that of the right, so that the lowest segment and the 13th column are, etc. [L1 to L12]. In columns 25 to 27 and 28 to 30 are the X, Y and Z components (in this order) of the right and left acceleration sensors.

As a rule, the zero measurement is repeated 6 times and the flexion, extension, right rotation, left rotation, lateral flexion right and lateral flexion left 5 times each.

8 Index overview

Various indexes can be formed by means of the measurement and assessment system according to the invention, wherein a distinction is made between short-term data and long-term data.

Advantageously, 5 indices are calculated (total index, range of motion, dynamics, activity and comparison between short term and long term). However, it is also possible to form further indices with the measurement and assessment system according to the invention. Since some indices also contain long-term parameters, which only exist if a long-term measurement was made, some case distinctions must be made.

• • wird der Gesamtindex aus Kurzzeit- und Langzeitdaten berechnet. (Modell Gesamtindex LZM KZM) sonst nur aus Kurzzeitdaten (Modell Gesamtindex KZM)
• • wird der Beweglichkeit aus Kurzzeit- und Langzeitdaten berechnet und heißt D-MCI sonst nur aus den Kurzzeitdaten (C-MCI)
• • wird der Index Dynamik aus Kurzzeit und Langzeitdaten berechnet und heißt D-DCI sonst nur aus den Kurzzeitdaten (C-DCI)

If there is a long-term measurement, then

• • the total index is calculated from short-term and long-term data. (Model total index LZM KZM) otherwise only from short-term data (model total index KZM)
• • the mobility is calculated from short-term and long-term data and is otherwise called D-MCI only from the short-term data (C-MCI)
• • the index dynamics is calculated from short-term and long-term data and is otherwise called D-DCI only from the short-term data (C-DCI)

9

The basis for the formation of the indices according to the invention by the measurement and assessment system according to the invention is the recognition that a large number of individual data do not permit any assessment of the overall state of a system. As a first step, training data is calculated based on the standard collective. For each training subject, the characteristics for each index are calculated. For each feature, the standard deviation and the mean over each subject are determined. Subsequently, the factor matrix and the regression coefficients are calculated by statistical methods.

The selection of the features to be calculated and the filtering of the raw data is individual for each index.

In the factor analyzes, the 10 most significant factors are chosen.

10

• 1. Alle Merkmale müssen berechnet werden.
• 2. Die berechneten Merkmale müssen standardisiert werden.
  

  • x

  = Durchschnitt aller Werte eines Merkmals X.

  Der Nenner gibt die Standardabweichung an. (beides wird aus den Trainingsdaten berechnet)
• 3. Faktorenmatrix ermitteln, die für den Index benötigt werden. Faktorenmatrix errechnet sich aus den Trainingsdaten.
• 4. Die Faktoren für die logistische Regression müssen berechnet werden.
  
• 5. Aus den Faktoren den Zwischenindex berechnen.
  
• 6. Den Zwischenindex auf eine möglichst lineare Skala von 0 bis 10 projizieren. Dazu wird die Sigmoid-Funktion aus Schritt 5 umgekehrt. Index = logit(Zwischenindex) F / 10-4 Index > 10 wird auf 10 gesetzt Index < 0 wird auf 0 gesetzt
• 7. Aus den Trainingsdaten werden Kurven für die Wahrscheinlichkeitsdichte von Probanden berechnet. Im Bereich in dem beide Kurven unter 70% liegen wird der unbestimmte Bereich des Index (blauer Bereich definiert).

In the measurement and assessment system according to the invention, all indices are calculated and formed according to a uniform scheme. The following applies:

• 1. All features must be calculated.
• 2. The calculated features must be standardized.
  

  • x

  = Average of all values of a feature X.

  The denominator indicates the standard deviation. (both are calculated from the training data)
• 3. Identify the factor matrix needed for the index. Factor matrix is calculated from the training data.
• 4. The factors for the logistic regression must be calculated.
  
• 5. Calculate the intermediate index from the factors.
  
• 6. Project the intermediate index to the nearest possible linear scale from 0 to 10. To do this, the sigmoid function of step 5 is reversed. Index = logit (intermediate index) F / 10-4 Index> 10 is set to 10 Index <0 is set to 0
• 7. From the training data curves for the probability density of subjects are calculated. In the area where both curves are below 70%, the indefinite area of the index (blue area) is defined.

11 Calculation of indices

The calculation of the indices according to the invention is carried out according to the steps described below with the aid of the diagrams shown schematically in the figures.

11.1 Short-term index ELCI

The Epionics SPINE Lumbar Choreography Index (ELCI) of the measurement and assessment system of the invention is used to evaluate short-term choreography.

11.2 Overall ELDI index

The Epionics SPINE Lumbar Dynamography Index of the measurement and assessment system according to the invention serves the function evaluation of a subject in only one number.

11.3 Activity Evaluation Index: D-AI

The Epionics SPINE Activity Index (D-AI) of the measurement and evaluation system according to the invention describes the activity during a long-term measurement. It incorporates all the traits that suggest energy, activity or movement.

11.4 Index for assessing mobility including long-term measurement data: D-MCI

The Epionics SPINE Dynamography Motion Capability Index (D-MCI) of the measurement and assessment system of the present invention describes the range of motion of a subject. It incorporates short and long-term measurement data. All characteristics are used for the calculation which describe the mobility of the subject (eg maximum and minimum motion dimensions).

11.5 Index for the assessment of mobility from short-term data: C-MCI

The Epionics SPINE Choreography Motion Capability Index (C-MCI) of the measurement and evaluation system according to the invention, like the D-MCI, describes the range of motion of the subject. However, only short-term measured data are included in his calculation.

11.6 Index for assessing dynamics including long-term measurement data: D-DCI

The Epionics SPINE Dynamography Dynamic Capability Index (D-DCI) of the measurement and assessment system of the present invention describes the dynamics in the subject's movements. It is calculated from short and long-term measurement data. For the calculation, features from velocities are used.

11.7 Index for the assessment of dynamics from short-term measured data: C-DCI

The Epionics SPINE Choreography Dynamic Capability Index (C-DCI) of the measurement and assessment system of the invention, like the D-DCI, describes the dynamics in the subject's movements. However, it is only calculated from short-term measured data.

11.8 Index for comparing short-term and long-term measurement data: D-MCUI

The Epionics SPINE Dynamography Motion Capability Utilization Index (D-MCUI) of the measurement and assessment system according to the invention describes how well the subject uses his functional potential shown in the short-term measurement in everyday life. For this purpose, different short-term and long-term characteristics are set in relation.

12 General calculations for the practice of the invention

12.1 Low pass filtering using Savitzky-Golay filtering

The long-term data is smoothed with a low-pass filter, for example with a Savitzky-Golay filter (window width k = 31, order n = 3). The smoothing takes place individually for each segment and each position sensor axis. Other low pass filters are equally possible. However, in the following, only the use of Savitzky-Golay filtering in the present invention will be discussed in detail, since the Savitzky-Golay filtering has been found to be advantageous. However, this does not limit Savitzky-Golay filtering.

13 long-term calculations

13.7 Angular parameters segmental

For each measurement zone, the angles are filtered using the Savitzky-Golay filter. Then the following parameters are calculated for all angle segments.

13.1.1 Mean

13.1.2 Standard deviation:

13.13 minimum:

Returns the minimum value found in the vector. The parameter is called LZ_WinkSeg_MIN_ ?? With ?? = R1 to L12.

13.1.4 Maximum:

Returns the maximum value in the vector. The parameter is called LZ_WinkSeg_MAX_ ?? With ?? = R1 to L12.

13.1.5 Lordosis-kyphosis transition:

To identify the lordosis-kyphosis transition, the sensor band is viewed at a time from the lowest segment upwards. All negative angles still count as lordosis. The first segment with an angle greater than zero is defined as a lordosis-kyphosis transition.
Ex .: Seg1 Seg2 Seg3 seg4 seg5 SEG6 seg7 SEG8 Seg9 Seg10 Seg11 Seg12 angle -1 -1 -1 -1 -1 -1 1 1 1 1 1 1 Table 1: Example for the calculation of the lordosis-kyphosis transition

In the example, the value 7 is output as lordosis-kyphosis transition, since the 7 segment is the first with a positive angle. The determination of the transition takes place for each time point of the long-term measurement.

13.1.6 Lordosis angle:

After determining the lordosis-kyphosis transition, the lordosis angle can be determined. For this, the angle values of the segments are added up to the lordosis-kyphosis transition segment.

Subsequently, the mean value is formed over all measurement times.

Similarly, the calculation for LZ_WinkSeg_Mean_Lord_Winkel_L.

13.1.7 Kyphosis angle:

All angles that were not added to the lordosis angle are summed up for the kyphosis angle.

Analogously, here too the mean value is formed over all measurement times.

Similarly, the calculation for LZ_WinkSeg_Mean_KyphWinkel_L

13.2 angle sums

For each measurement zone, the angles are filtered using the Savitzky-Golay filter. Subsequently, all 24 angle values are added up for each time point.

On this basis, all further calculations of this subchapter take place.

13.2.1 Mean

13.2.2 Standard deviation

13.2.3 Minimum:

Returns the minireal value found in LZ_WinkSum_Mean. The parameter is called LZ_WinkSum_MIN.

13.2.4 Maximum:

Returns the maximum value found in LZ_WinkSum_Mean. The parameter is called LZ_WinkSum_MAX.

13.2.5 Percentile Calculations

• 1. Die Vektorelemente werden sortiert.
• 2. Jedem Vektorelement wird ein Prozentwert zugewiesen:
  
• 3. Wenn ein Prozentwert gesucht wird, der nicht bereits zugewiesen wurde, wird zwischen 2 zugewiesenen Werten linear interpoliert.
  
• 4. Wenn der gesuchte Wert außerhalb der berechneten Prozentgrenzen liegt, wird das Maximum bzw. Minimum angenommen.

For a vector X with n elements, the percentiles are calculated as follows.

• 1. The vector elements are sorted.
• 2. Each vector element is assigned a percentage value:
  
• 3. When searching for a percentage that has not already been assigned, it interpolates between 2 assigned values linearly.
  
• 4. If the requested value is outside the calculated percent limits, the maximum or minimum is assumed.

13.3 Middle rankings of the angle segments

As under 13.1, the long-term data of the angles are also smoothed for this parameter with the Savitzky-Golay filter. Subsequently, the calculation of the ranking places for each time point and each sensor band starts individually.

Algorithm Ranking:

• 1. i = 1
• 2. Find the segment with the smallest value
• 3. Count how often this value occurs and in which segments
  
• 4. All segments with this value are ranked
• 5. i = i + n
• 6. For all values without rank, the procedure starts again at point two. Then the average rank is determined for each segment via the ranking places of all points in time.
  

13.4 angular velocities segmental

For the evaluation of the angular velocity, each individual angular segment is filtered first with the Savitzky-Golay filter. Subsequently, by means of the formula: Angular velocity _j = 5 · (angle _{j + 10} - angle _j ) F / 13-14 the angular velocity is calculated. This segmental angular velocity forms the basis for the further calculations of this chapter. The following calculations follow the formulas already known from segmental angle calculations.

13.4.1 Mean

13.4.2 Standard deviation:

13.4.3 Minimum:

Returns the minimum value found in the vector. The parameter is called LZ_Vseg_MIN_ ?? With ?? = R1 to L12.

13.4.4 Maximum:

Returns the maximum value in the vector. The parameter is called LZ_Vseg_MAX_ ?? With ?? = R1 to L12.

13.5 angular velocity sums

To calculate the speed sums, the data was first smoothed using the Savitzky-Golay filter. Then, according to the formula F 13-14, the angular velocity of each segment was calculated. Thereafter, the sum was calculated for each time over all angle segments. This sum forms the basis for the following calculations.

13.5.1 Mean

13.5.2 Standard deviation

13.5.3 Minimum:

Returns the minimum value found in LZ_Vsum_Mean. The parameter is called LZ_Vsum_MIN.

13.5.4 Maximum:

Returns the maximum value found in LZ_Vsum_Mean. The parameter is called LZ_Vsum_MAX.

13.6 Center ranks of angular velocities

The calculation of the average ranking of the angular velocities is analogous to the algorithm in 13.3. However, the angular velocities of the individual segments are defined as the input variable.

13.7 angular accelerations segmental

For the evaluation of the angular acceleration, each individual angular segment is filtered first with the Savitzky-Golay filter. Then the angular velocity is calculated by means of the formula F 13-14 and with Angular acceleration _j = 5 · (angular velocity _{j + 10} - angular velocity) F / 13-21 the angular acceleration. The angular acceleration again forms the basis for all subsequent calculations.

13.7.1 Mean

13.7.2 Standard deviation:

13.7.3 Minimum:

Returns the minimum value found in the vector. The parameter is called LZ_Aseg_MIN_ ?? With ?? = R1 to L12.

13.7.4 Maximum:

Returns the maximum value in the vector. The parameter is called LZ_Aseg_MAX_ ?? With ?? = R1 to L12.

13.8 Angular acceleration sums

Analogous to the calculation of the angular velocity sums.

13.8.1 Mean

13.8.2 Standard deviation

13.8.3 Minimum:

Returns the minimum value in LZ_Asum_Mean. The parameter heals LZ_Asum_MIN.

13.8.4 Maximum:

Returns the maximum value in LZ_Asum_Mean. The parameter is called LZ_Asum_MAX.

13.9 Energy approaches

13.9.1 Activity

To calculate the activity, each measurement channel, including the position sensors, was smoothed using the Savitzky-Golay filter.

13.9.2 Energy Approach 11

As a data basis, the angular velocity is determined segmentally via F 13-14 from filtered angle data. So that positive and negative speed totals are not canceled on average, each time point is squared and the sum is formed over all 24 segments.

This will perform various calculations.

Average

Median

For the calculation of the median the vector Velocity square is sorted first.

standard deviation

Integral of the speed square

Size-scaled integral

13.10 Comparison of short-term measurements with long-term measurements

To compare short-term measurements with long-term measurements, all raw data is filtered using the Savitzky-Golay filter. For the calculation of the angle variables, the angle values of each short-term measurement are attached to each other. For the velocity variables, the velocities for each short-term measurement are calculated individually and also hung together.

13.11 Symmetry

For the parameters of symmetry considerations, all measurement channels were filtered using the Savitzky-Golay filter. Thereafter, the data of the right was subtracted from those of the left segments. From the result, the following angle parameters were calculated. parameter name description LZ_Symm_Mean_LR ?? Mean of difference right LR LZ_Symm_STD_LR ?? Standard deviation of the difference LR LZ_Symm_Min_LR ?? Minimum difference LR LZ_Symm_MAX_LR ?? Maximum of the difference LR Table 2: Angular parameters of symmetry considerations Each with ?? = Segment 1 to 12.

13.11.1 Pearson Correlation

In addition, it is advantageous to determine the Pearson correlation coefficients of different measured variables in the case of the symmetry parameters. In each case, the results of the measurements of the right sensor band are compared with those of the left sensor band. parameter name description LZ_Symm_Corr_Pears_R_L ?? Correlation coefficient between right and left angle segments LZ_Symm_Corr_Pears_RX_LX Correlation coefficient between right and left X axis of position sensor LZ_Symm_Corr_Pears_RY_LY Correlation coefficient between right and left he Y axis of position sensor LZ_Symm_Corr_Pears_RZ_LZ Correlation coefficient between right and left Z axis of position sensor

14 short-term calculations

14.1 zero measurement

In the case of the zero measurements, the 50 quietest consecutive frames are identified according to the invention for further evaluation. However, this only happens if the NM_i is longer than 52 frames. The identification is done as follows:
A variable is set to a very high value (std_buf = 1000). Then frame 1 starts to calculate the standard deviation over the next 50 frames. If this is less than the set value, it will be replaced by the calculated standard deviation. This procedure is performed frame by frame up to final frame-50. The frames with the lowest standard deviation (buf) are used for further calculation, resulting in the variables NM_1 (50), ..., NM_6 (50).

All zero measurements together are defined as follows:

14.1.1 Mean for 24 segments

14.1.2 Standard deviation for 24 segments

mit n = Anzahl der Messpunkte, k = Mittelwert und ?? = 24 Spalten (R1 bis L12).Calculate the standard deviation over adjacent zero measurements NM_all:

with n = number of measuring points, k = average and ?? = 24 columns (R1 to L12).

mit n = Anzahl der Wiederholungen, NM_i = verschiedene Wiederholungen der Nullmessungen, n = Anzahl der Messwerte von NM_i und ?? = 24 Spalten (R1 bis L12).Calculate the standard deviation over the mean values of the individual zero measurements NM_1 (50) to NM_6 (50):

with n = number of repetitions, NM_i = different repetitions of the zero measurements, n = number of readings of NM_i and ?? = 24 columns (R1 to L12).

14.1.3 Median for 24 segments

The median should be calculated from all zero measurements attached to each other. All values of a column NM_alle_ ?? should be sorted according to their size. The value in the middle (central value) is the median. For an even number of values, the mean is calculated from the arithmetic mean of the two mean values: NM_Median_Nulla_ ?? = Median (NM_all_ ??)

14.1.4 Middle rank for 2 × 12 segments

For the left and right sensors, the average rankings of the individual segments should be calculated. This means that the rankings must be calculated on the one hand for columns 1-12 and on the other hand 13-24. The matrix NM_Mean_alle must be traversed line by line. The values of columns 1-12 are to be ordered line by line in each column in size. The smallest value is then assigned the 1, the second the 2, etc. and the largest the 12. If several values are equal, the arithmetic mean of the consecutive values to be assigned is assigned to the same-sized values. The arithmetic mean value is then to be formed over the individual columns of these assigned values.

Outcome variables:

• Mittl_Rang_rechts,
• Mittl_Rang_links

14.1.5 Difference between mean maxima and mean minima

mit k = Anzahl der vorliegenden Nullmessungen, ?? = 24 Spalten (R1 bis L12).The parameter NM_Diff_mean_Min_Max_ describes the difference between the mean values of the maxima and the minima of the individual repetitions:

with k = number of zero measurements, ?? = 24 columns (R1 to L12).

14.1.6 Difference mean, median, rank

Average:

• NM_Diff_mean_L_R_S? = Mean (NM_all_ [column12 +?]) - Mean (NM_all_ [column?]) F / 14-5 With ? = 1 to 12.

Median:

• NM_Diff_median_L_R_S? = Median (NM_all [column12 +?]) - median (NM_all_ [column?]) F / 14-7 With ? = 1 to 12.

Rank Place:

• NM_Diff_mittl_Rangplatz_S? = Mittl_Rang_links - Mittl_Rang_rechts F / 14-8

Mittl_Ran_links / right see above! With ? = 1 to 12. Kyphosis-lordosis transition

Right:

• • KyphLord_rechts = 12;
• • Lordose_rechts = sum(Mean_Nulla(1,1:12));
• • Kyphose_rechts = 0;

Determination of the parameters KyphLord_right, Lordose_right and Kyphose_rechts. First, the position of all segments which have a positive or = 0 (pos_Wink_right) and negative (neg_Wink_right) angles is determined in columns 1-12. If there is no positive entry, then:

• • KyphLord_right = 12;
• • Lordose_right = sum (Mean_Nulla (1,1: 12));
• • kyphosis_right = 0;

• 1. Es gibt keine Differenz nebeneinanderliegender Einträge >1 sowohl in diff_pos_Segm_rechts als auch in diff_neg_Segm_links →
• • Kyph_Lord_rechts = pos_Wink_rechts(1) [(1) = kleinster Eintrag, auch im Folgenden!
• • Lordose rechts = sum(mean(Null_alle(1,neg_Wink_rechts(1):neg_Wink_rechts(end)))); [Summe aller Segmente von neg_Wink_rechts(1) bis neg_Wink_rechts(end)Hend) = größter Eintrag, summ(...) = Summe der in der Klammer stehenden Einträge, (1:2) = bedeutet von 1 bis 2; dies gilt auch im Folgenden!]]
• • Kyphose_rechts = sum(Mean_Nulla(1,pos_Wink_rechts(1):pos_Wink_rechts(end)));
• 2. Es gibt keine Differenz nebeneinanderliegender Einträge >1 in diff_neg_Segm_rechts →
• • KyphLord_rechts = (pos_Wink_rechts(diff_pos_Segm_rechts(end)+1))
• • Lordose_rechts = sum(Mean_Nulla(1,neg_Wink_rechts(1):neg_Wink_rechts(end)))
• • Kyphose_rechts = sum(Mean_Nulla(1,pos_Wink_rechts(diff_pos_Segm_rechts(end)+1):pos_Wink_rechts(end)))
• 3. Es gibt keine Differenz nebeneinanderliegender Einträge >1 in diff_pos_Segm_rechts →
• • KyphLord_rechts = (pos_Wink_rechts(1))
• • Lordose_rechts = sum(Mean_Nulla(1,neg_Wink_rechts(1):neg_Wink_rechts(diff_neg_Segm_rechts(1)–1)))
• • Kyphose_rechts = sum(Mean_Nulla(1,pos_Wink_rechts(1):pos_Wink_rechts(end)))
• 4. Ansonsten (es gibt bei diff_pos_Segm_rechts oder bei diff_pos_Segm_links eine Differenz nebeneinanderliegender Einträge >1 →
• • KyphLord_rechts = pos_Wink_rechts(diff_pos_Segm_rechts(end)+1)
• • Lordose_rechts = sum(Mean_Nulla(1,neg_Wink_rechts(1):diff_neg_Segm_rechts(1)–1))
• • Kyphose_rechts = sum(Mean_Nulla(1,pos_Wink_rechts(diff_pos_Segm_rechts(end))+1:pos_Wink_rechts(end)))

The variables pos_Wink_right and neg_Wink_right are also checked to see if the difference between two adjacent entries is greater than 1. Depending on the result (diff_pos_Segm_right and diff_neg_Segm_links), there is a case distinction:

• 1. There is no difference of adjacent entries> 1 in diff_pos_Segm_right as well as in diff_neg_Segm_links →
• • Kyph_Lord_rechts = pos_Wink_rechts (1) [(1) = smallest entry, also in the following!
• • Lordose right = sum (mean (zero_all (1, neg_Wink_right (1): neg_Wink_right (end)))); [Sum of all segments from neg_Wink_right (1) to neg_Wink_right (end) Hend) = largest entry, summ (...) = sum of the entries in the bracket, (1: 2) = means from 1 to 2; this also applies in the following!]]
• • kyphose_right = sum (Mean_Nulla (1, pos_Wink_right (1): pos_Wink_right (end)));
• 2. There is no difference of adjacent entries> 1 in diff_neg_Segm_rechts →
• • KyphLord_right = (pos_Wink_right (diff_pos_Segm_right (end) +1))
• • Lordose_right = sum (Mean_Nulla (1, neg_Wink_right (1): neg_Wink_right (end)))
• • kyphose_right = sum (Mean_Nulla (1, pos_Wink_right (diff_pos_Segm_right (end) +1): pos_Wink_right (end)))
• 3. There is no difference of adjacent entries> 1 in diff_pos_Segm_rechts →
• • KyphLord_right = (pos_Wink_right (1))
• • Lordose_right = sum (Mean_Nulla (1, neg_Wink_right (1): neg_Wink_right (diff_neg_Segm_right (1) -1)))
• • kyphosis_right = sum (Mean_Nulla (1, pos_Wink_right (1): pos_Wink_right (end)))
• 4. Otherwise (diff_pos_Segm_right or diff_pos_Segm_links have a difference of juxtaposed entries> 1 →
• • KyphLord_right = pos_Wink_right (diff_pos_Segm_right (end) +1)
• • Lordose_right = sum (Mean_Nulla (1, neg_Wink_right (1): diff_neg_Segm_rechts (1) -1))
• • kyphose_right = sum (Mean_Nulla (1, pos_Wink_right (diff_pos_Segm_right (end)) + 1: pos_Wink_right (end)))

14.1.7 Lateral and Sagittal Angles

Sagittal and lateral angles describe the tilt of the sensors in a lateral or sagittal plane. The angles can be calculated from the measured values of the acceleration sensors. The individual components are each indicated in the order X, Y, Z. Then the searched angles from the values of the position sensor are calculated as follows: x _{right / left} = mean (zero_all (:, m) F / 14-9 y _{right / left} = mean (zero_all (:, n) F / 14-10 z _{right / left} = mean (zero_all (:, m) F / 14-11

mit z = z-Wert des Lagesensors des Sensorbands und x = x-Wert des Lagesensors des Sensorbandes

mit y = y-Wert des Lagesensors des Sensorbands und x = x-Wert des Lagesensors des SensorbandesWith mean = mean and ":" = over all lines, m = 25 and 28, n = 26 and 29, o = 27 and 30

with z = z value of the position sensor of the sensor band and x = x value of the position sensor of the sensor band

with y = y value of the position sensor of the sensor band and x = x value of the position sensor of the sensor band

From this the parameters follow:
NM_Latwink_right, NM_Latwink_links, NM_Sagwink_right, NM_Sagwink_links

14.1.8 Difference between lateral and sagittal angles

• NM_Diff_Sagwink = NM_Sagwink_links - NM_Sagwink_rechts F / 14-14
• NM_Diff_Latwink = NM_Latwink_links - NM_Latwink_rechts F / 14-15

14.1.9 Angle sums

The angle sums are calculated from the sum of the mean values of all rows of Null_alle over the columns 1 to 12 or 13 to 24: NM_WinkSum_right = sum (mean (zero_all (:, 1:12)) F / 14-16 NM_WinkSum_links = sum (mean (zero_all (:, 13:24) F / 14-17

14.1.10 difference angle sums

• NM_Diff_WinkSum = NM_WinkSum_links - NM_WinkSum_right F / 14-18

14.2 Flexion

14.2.1 Maximum Flexion

For each repetition of flexion, it is necessary to identify the time of maximum flexion. This results from the arithmetic mean (rounded down to nearest integer) from the time points of the maxima of the left and right sensor, the time being defined as the moment when the sum over the lower 6 segments (about the most mobile range) is the maximum , The angles at this moment refer to the maximum angles, standard deviations, lordosis and kyphosis angles, sagittal and lateral angles, etc.

14.2.2 Mean of maxima:

• 
  with n = number of repetitions and ?? = 24 columns (R1 to L12)

14.2.3 Standard deviation of maxima

• 
  with n = number of repetitions and ?? = 24 columns (R1 to L12).

14.2.4 Median of maxima

• Flex_Winkel_max_median ?? = median (Values_Flex_max_ _?? ) F / 14-22 With ?? = 24 columns (R1 to L12)

14.2.5 Normalized Mean and Median

Flex_Winkel_max_median_norm_?? = median(Werte_Flex_max_??) – NM_Mean_alle_?? F1 / 4-24 Mit n = Anzahl der Wiederholungen und ?? = 24 Spalten (R1 bis L12)Mean values and medians are segmented by segment, ie subtraction of the zero measurement:

Flex_Winkel_max_median_norm_ ?? = median (values_flex_max _?? ) - NM_Mean_alle_ ?? F1 / 4-24 With n = number of repetitions and ?? = 24 columns (R1 to L12)

14.2.6 Maxima of the individual and individual normalized angles

• Flex_Winkel_max_einzel_NrX_ ?? = Values_Flex_max _{NrX, ??} F / 14-25 With NrX = repetition 1 to 5 and ?? = 24 columns (R1 to L12)
• Flex_Winkel_max_einzel_norm_NrX_ ?? = Flex_Winkel_max_einzel_NrX_ ?? - NM_Mean_alle_ ?? F / 14-26 With NrX = repetition 1 to 5 and ?? = 24 columns (R1 to L12)

14.2.7 Differences in angles and standard deviations

• Flex_Diff_Wink_max_einzel_NrX_S? = Winkel_Flex_max_einzel_ [Column12 +?] - Winkel_Flex_max_einzel_ [Column?] F / 14-27 With ? = 1 to 12, X = repetition 1 to 5.
  
  With ? = 1 to 12 and n = number of repetitions
• Flex_Diff_STD_S? = Flex_STD_max_Winkel_ [Column12 +?] - Flex_STD_max_Winkel_ [Col??] F / 14-29 With ? = 1 to 12

14.2.8 Kyphosis / Lorodan transition

The kyphosis / lordosis transition is expediently calculated individually for the right and left sensors. The transition occurs at the segment where the first value (seen from the lowest segment) is greater than or equal to zero.

Emerging variables:

• Flex_Kyph_Lord_rechts, Flex_Kyph_Lord_links

14.2.9 Sinus fit

• 1. Repetitions of the flexion movement in the time domain are contiguous.
• 2. Using a nonlinear regression fit a sine in the curve as well as possible.
• Third
  
  With ?? = L1 to L12 and R1 to R12 Explanation Variance derived from the fitted sinusoidal function Total variance stems from the flexion curves attached to each other.

14.2.10 Sum of all positive / negative angles

Mit ?? = 12 Spalten (R1 bis R12)

Mit ?? = 12 Spalten (L1 bis L12)

Mit ?? = 12 Spalten (R1 bis R12)

Mit ?? = 12 Spalten (L1 bis L12)The sum of the positive and negative angles is also calculated separately from left and right sensors. The summation depends on the decision whether the value is greater or less than 0:

With ?? = 12 columns (R1 to R12)

With ?? = 12 columns (L1 to L12)

With ?? = 12 columns (R1 to R12)

With ?? = 12 columns (L1 to L12)

14.2.11 Speed

Mit ?? = 24 Spalten (R1 bis L12), n = Anzahl der Wiederholungen

Mit n = Anzahl der Messpunkte, ?? = 24 Spalten (R1 bis L12)

Mit n = Anzahl der Wiederholungen, ?? = 24 Spalten (R1 bis L12)

Mit n = Anzahl der Wiederholungen, ?? = 24 Spalten (R1 bis L12)To determine the speed values, the formula F 13-14 is used to calculate the segmental speeds of the individual exercises (speed _{j, ??} ) at each time.

With ?? = 24 columns (R1 to L12), n = number of repetitions

With n = number of measuring points, ?? = 24 columns (R1 to L12)

With n = number of repetitions, ?? = 24 columns (R1 to L12)

With n = number of repetitions, ?? = 24 columns (R1 to L12)

14.2.12 Middle rank

Mit ?? = 24 Spalten (R1 bis L12), n = Anzahl der Wiederholungen, m = Anzahl der MesspunkteThe middle rank is calculated as described in chapter 14.1.4.

With ?? = 24 columns (R1 to L12), n = number of repetitions, m = number of measuring points

14.2.13 Lateral and Sagittal Angles

Lateral and sagittal angles are calculated as described in Chapter 14.1.7. They are calculated according to the invention in each case at the time of maximum flexion.

From this, the parameters lateral angle flex right / left and sagittal angle flex right / left follow

14.2.14 Difference between lateral and sagittal angles

• Diff Lateralw Flex = Lateral Angle Flex Left - Lateral Angle Flex Right F / 14-41
• Diff Sagittalw Flex = sagittal angle flex left - sagittal angle flex right F / 14-42

14.2.15 angle sums

Mit n = Anzahl der Wiederholungen und 1–12 und 13–24 = zu summierende SpaltenThe angle sums are calculated from the sum over the columns 1 to 12 and 13 to 24 at the time of maximum flexion, averaged over all repetitions:

With n = number of repetitions and 1-12 and 13-24 = columns to sum

14.2.16 Difference angle sums

• Flex_Diff_WinkSum = Flex_WinkSum_links - Flex_WinkSum_right F / 14-45

14.3 Extension

• 1. Bei der Variablenbezeichnung wird aus Flex -> Ext.
• 2. Der Zeitpunkt der maximalen Extension berechnet sich nicht aus dem Maximum der 6 unteren Segmente (siehe Kapitel 14.2.1), sondern aus dem Minimum der 6 unteren Segmente.

With the extension, all values are calculated analogous to the inflection. There are, however, two things to consider:

• 1. For the variable name, Flex -> Ext.
• 2. The time of the maximum extension is not calculated from the maximum of the 6 lower segments (see chapter 14.2.1), but from the minimum of the 6 lower segments.

14.4 Right lateral flexion

To evaluate the lateral flexion, the mean zero value is subtracted from each time of a lateral flexion.

This formula applies not only to the angle segments but also to the values of the sagittal and lateral pelvic angles.

14.4.1 Maximum values of the exercises

14.4.1.1 Angle values

und das Maximum gesucht. Das Maximum markiert den Zeitpunkt der maximalen Lateralflexion. Da jede Übung 5-mal wiederholt wird können Mittelwert und Standardabweichungen für die Maximalwerte der Winkel bestimmt werden. Parametername Beschreibung LatR_maxLat_Mean_?? Winkel bei maximaler Lateralflexion; davon den Mittelwert aus den 5 Wiederholungen LatR_maxLat_STD_?? Winkel bei maximaler Lateralflexion; davon die Standardabweichung aus den 5 Wiederholungen Tabelle 3: Parameternamen der Winkelmaximalwerte der Lateralflexion ?? = R1 bis L12First, the time of maximum lateral flexion is determined. For this purpose, the amounts of the measuring zones of the right sensor band are added up

and searched for the maximum. The maximum marks the time of maximum lateral flexion. As each exercise is repeated 5 times, the mean and standard deviations for the maximum values of the angles can be determined. parameter name description LatR_maxLat_Mean_ ?? Angle at maximum lateral flexion; of which the average of the 5 repetitions LatR_maxLat_STD_ ?? Angle at maximum lateral flexion; of which the standard deviation from the 5 repetitions Table 3: Parameter names of the angle maximum values of the lateral flexion ?? = R1 to L12

14.4.1.2 Pelvic angle

At the time of maximum lateral flexion, the pelvic angles are also determined in the sagittal and lateral planes.

mit z = z-Wert des Lagesensors des Sensorbands und x = x-Wert des Lagesensors des Sensorbandes

mit y = y-Wert des Lagesensors des Sensorbands und x = x-Wert des Lagesensors des SensorbandesAfter that, the searched angles from the values of the position sensor are calculated as follows.

with z = z value of the position sensor of the sensor band and x = x value of the position sensor of the sensor band

with y = y value of the position sensor of the sensor band and x = x value of the position sensor of the sensor band

Subsequently, the sagittal or lateral angle of the zero measurement is subtracted.

14.4.1.3 Speed values

To determine the speed values, the formula F 13-14 is used to calculate the speeds of the individual exercises. For each exercise and each segment, the maximum positive and maximum negative speed are determined. From this, the parameters shown in Table 4 are calculated. parameter name description LatR_maxLat_vmax_mean_ ?? maximum positive speed of the exercise; of which the average of the 5 repetitions LatR_maxLat_vmax_std_ ?? maximum positive speed of the exercise; of which the standard deviation from the 5 repetitions LatR_maxLat_vmin_mean_ ?? maximum negative speed of the exercise; of which the average of the 5 repetitions LatR_maxLat_vmin_std_ ?? maximum negative speed of the exercise; of which the standard deviation from the 5 repetitions Table 4: Parameter names of the maximum velocity values of the lateral flexion ?? = R1 to L12

14.4.2 All times of the exercises

After calculating the speed for each exercise, the speeds of each exercise are linked together. From the resulting field, averages, standard deviations, medians, and middle rankings are determined for each segment. parameter name description LatR_alle_v_mean_ ?? Mean of all time points (5 repetitions hanged together) LatR_alle_v_std_ ?? Standard deviation of all time points (5 repetitions hanged together) LatR_alle_v_median_ ?? Median of all time points (5 repeats hanged together) LatR_v_rang_ ?? middle rank of all time points (5 repetitions hanged together) Table 5: Parameter names of the speed parameters for all times ?? = R1 to L12

14.5 Lateral flexion on the left

Apart from two differences, the parameters of the lateral flexion on the left are calculated in the same way as the parameters of the lateral flexion on the right. First, not the right but the left sensor band is used to determine the maximum value, and second, the parameter name "LatR" is replaced by "LatL".

QUOTES INCLUDE IN THE DESCRIPTION

This list of the documents listed by the applicant has been generated automatically and is included solely for the better information of the reader. The list is not part of the German patent or utility model application. The DPMA assumes no liability for any errors or omissions.

Cited patent literature

• DE 102008052406 [0002]

Cited non-patent literature

• Panjabi MM, Goel V, Oxland T. Takata K, Duranceau J, Krag M. Price M. Human Lumbar Vertebrae-Quantitative three-dimensional anatomy. Spine. 17 (3): 299-306, March 1992 [0023]
• Panjabi MM, Takata K, Goel V, Federic D, Oxland T, Duranceau J. Krag M. Thoracic Human Vertebrae-Quantitative Three-Dimensional Anatomy. Spine. 16 (8): 888-901, Aug 1991 [0023]
• Pooni JS, Hukin's DWL, Harris PF, Hilton RC Davies KE. Camparison of the structure of human intervertebral discs in the cervical, thoracic and lumbar regions of the spine. Surgical Radiologic Anatomy. 8: 175-182, 1986 [0023]
• Amman C, Naderi S, Kiray A, Aksu FT, Yilmaz HS, tetik S, Korman E. The human sacrum and screwdriving approaches for screw placement. Journal of Clinical Neuroscience. 16: 1046-1049, 2009 [0023]
• Grieve, 1981 [0037]

Claims (10)

Measurement and assessment system consisting of a specific number of sensors that generate a first set of data and that data are assigned to a particular sensor, which data are transmitted to a processing unit to from the first set of data by applying a second Amount of data, wherein the second amount of data is substantially smaller than the first amount of data. The measurement and assessment system of claim 1, wherein the second set of data consists of indices. Measurement and assessment system according to claim 1 or 2, wherein the measures consist of the following steps: Applying low pass filtering to the first set of data; - calculation of specific characteristics; - Application of a regression analysis and / or factor analysis. The measurement and assessment system of claim 3, wherein the low pass filtering is Savitzky-Golay filtering. Measurement and assessment system according to claim 3 or 4, wherein in the calculation of the specific features of the average and the standard deviation for all features is illustrated. Measurement and assessment system according to claim 5, wherein the averaging also takes into account other indices. Measurement and assessment system according to claim 6, wherein the index (s) is taken from the group of D-MCI and / or D-DCI and / or D-AI and / or D-MCUI. A measurement and assessment system according to any one of the preceding claims, wherein features consist of raw data describing the mobility of a subject. Measurement and assessment system according to one of the preceding claims, wherein the following indices are formed: • Overall Index (ELDI) • Overall Short-Term Index (ELCI) • Agility Index Short Time (D-MCI / C-MCI) • Dynamic Index Short Time (D-DCI / C-MCI) • Activity Index Long Term (D-AI) • Flexibility index long-term (D-MCI) • Dynamic index long-term (D-DCI) • Comparison Index KZM / LZM Index (D-MCUI) Measuring and assessment method, characterized in that a measuring and assessment system according to one of the forthcoming claims is used.

Images