Method and device for measuring efficacy of a sportive activity

The invention relates to the field of physical activities. It relates, more particularly, to a method and a device for measuring the effectiveness of a sporting gesture.

For any sportsman, competitor or amateur, who seeks to progress, training aims to improve its efficiency, or yield, which can be defined by the ratio between the energy expended and the energy cost required by movement and / or displacement of the sportsman.

If we take the example of the cross-country skier, we understand, indeed, that on a given route, a sportsman must, to be more efficient, either increase the energy he spends, or improve his technique and its sliding quality.

Conventionally, training focuses on increasing the numerator of the ratio mentioned above, that is to say the energy expended. Many works on the physiology of the sport made it possible to develop theories and exercises to improve the physical capacities of the sportsman. They are based on an essential, easily accessible criterion: heart rate. However, this physical training involves significant efforts for generally slow progress.

Improving efficiency also means reducing the denominator of the “energy spent / energy cost” ratio. With regard to the energy cost of movement or displacement, the problem essentially rests on material aspects of the sportsman's equipment. On the other hand, the energy cost of the movement goes through an improvement in the technical control of the gesture in question. Its effectiveness is directly linked to parameters that are difficult to quantify (orientation of the supports, relaxation of the opposing muscles in the work phases and of the agonist muscles in the rest phases, parasitic movements, etc.).

The analysis of these parameters requires, most often, the presence of a knowledgeable outside person, capable of analyzing the gesture of the sportsman and providing adequate advice. But this is generally based more on the eye and the experience of the observer than on real scientific measurements.

The present invention aims to allow an athlete to assess the effectiveness of his technique, without the need for an outside gaze. The invention also provides, particularly for coaches, a rational means for quantifying the technical progress of the sportsman.

More specifically, the invention relates to a method for measuring the effectiveness of an athlete's gesture. It includes the following steps:<ul id="ul0001" list-style="dash" compact="compact"><li>retrieving at least one attribute <i>AT</i> the signal provided by an accelerometer placed on the athlete so as to capture the acceleration linked to this gesture,</li><li>comparison of this attribute with the average value <maths id="math0001" num=""><math display="inline"><mrow><mover accent="true"><mrow><mtext mathvariant="italic">AT</mtext></mrow><mo>¯</mo></mover></mrow></math><img file="EP1586353A1_D0001.tif" /></maths> of the same attribute, obtained for a reference population and stored in a memory, and</li><li>production of a representation of the effectiveness of the gesture.</li></ul>

According to the invention, the comparison is made between:<ul id="ul0002" list-style="dash" compact="compact"><li>the attribute <i>AT</i>, and</li><li>its average value <maths id="math0002" num=""><math display="inline"><mrow><mover accent="true"><mrow><mtext mathvariant="italic">AT</mtext></mrow><mo>¯</mo></mover></mrow></math><img file="EP1586353A1_D0002.tif" /></maths>, obtained for a reference population, corresponding to the percentage of the maximum heart rate supplied by a heart rate monitor worn by athletes.</li></ul>

Advantageously, the attribute <i>AT</i> is the standard deviation of the signal supplied by the accelerometer.

The invention also relates to a device for measuring the effectiveness of a sporting gesture. He understands :<ul id="ul0003" list-style="dash" compact="compact"><li>an accelerometer placed on an athlete so as to provide a signal representative of this gesture,</li><li>a processing circuit provided with a microprocessor and a memory, receiving the signal, the microprocessor being programmed to perform, at determined intervals, the following operations:<ul id="ul0004" list-style="none" compact="compact"><li>■ extraction of at least one attribute <i>AT</i> the accelerometer signal,</li><li>■ attribute comparison <i>AT</i> with the mean value <maths id="math0003" num=""><math display="inline"><mrow><mover accent="true"><mrow><mtext mathvariant="italic">AT</mtext></mrow><mo>¯</mo></mover></mrow></math><img file="EP1586353A1_D0003.tif" /></maths> of the same attribute, obtained for a reference population and stored in said memory, and</li><li>■ production of a representation of the effectiveness of the gesture, and</li></ul></li><li>signaling means for transmitting this representation to the sportsman.</li></ul>

Advantageously, the device further comprises a heart rate measuring device intended to supply the processing circuit with information relating to the athlete's heart rate. This processing circuit is programmed to perform the comparison operation between:<ul id="ul0005" list-style="dash" compact="compact"><li>the attribute <i>AT</i> , and</li><li>its average value <maths id="math0004" num=""><math display="inline"><mrow><mover accent="true"><mrow><mtext mathvariant="italic">AT</mtext></mrow><mo>¯</mo></mover></mrow></math><img file="EP1586353A1_D0004.tif" /></maths> , obtained for a reference population, corresponding to the percentage of the maximum heart rate supplied by a heart rate monitor worn by athletes.</li></ul>

The comparison is made according to the equation:<maths id="math0005" num=""><img file="EP1586353A1_D0005.tif" /></maths> in which :<ul id="ul0006" list-style="dash" compact="compact"><li><i>d</i>_(<i>ƒc</i>) is the difference between the runner and the reference population for a given percentage <i>ƒc</i> maximum heart rate,</li><li><maths id="math0006" num=""><math display="inline"><mrow><mfrac linethickness="0"><mrow><mtext>ρ</mtext></mrow><mrow><mtext>θ</mtext></mrow></mfrac></mrow></math><img file="EP1586353A1_D0006.tif" /></maths> is the value of the attribute determined for the runner or, in the general case, a vector whose components are the different attributes measured for the latter,</li><li><maths id="math0007" num=""><math display="inline"><mrow><mfrac linethickness="0" numalign="left" denomalign="left"><mrow><mtext>ρ</mtext></mrow><mrow><msub><mrow><mtext>µ</mtext></mrow><mrow><mtext mathvariant="italic">reƒ</mtext></mrow></msub></mrow></mfrac></mrow></math><img file="EP1586353A1_D0007.tif" /></maths> is the average of the attribute determined for the reference population or, in the general case, the vector whose components are the average values ​​of the attributes for said reference population,</li><li><i>VS</i>_<i>ƒc</i> is the covariance matrix of the mean values ​​of the attributes of the reference population, and</li><li><i>T</i> indicates that the corresponding matrix is ​​transposed.</li></ul>

The device can also include one or other of the following characteristics:<ul id="ul0007" list-style="dash" compact="compact"><li>the processing circuit and the signaling means are housed in a watch worn by the sportsman,</li><li>the signaling means are audible and are arranged so as to produce an intermittent sound whose frequency is correlated with the representation of the effectiveness of the gesture.</li></ul>

Other characteristics of the invention will appear more clearly on reading the description which follows, made with reference to the appended drawing in which:<ul id="ul0008" list-style="dash" compact="compact"><li>Figures 1a and 1b are schematic representations of the device according to the invention; and</li><li>Figure 2 illustrates the data processing method used in such a device.</li></ul>

To describe the invention, we will take the example of a runner, represented in FIG. 1, which we seek to evaluate the efficiency of the stride.

According to running theories, running efficiency is determined by the following points. The runner must:<ul id="ul0009" list-style="dash" compact="compact"><li>stand almost perpendicular to the ground;</li><li>focus its movements in the direction of movement;</li><li>choose an appropriate landing angle so that the energy of the support is recycled as much as possible;</li><li>have a high stride frequency;</li><li>coordinate the movements of his arms and legs;</li><li>keep his hips in line with his shoulders and head; and</li><li>have a quick swivel of the ankle after the foot has been placed so as to have better thrust.</li></ul>

All these parameters, or attributes, can be evaluated or quantified, but they are complex to combine in a single equation. In what follows, for the sake of simplification, the stride is modeled through the center of gravity of the runner, which has a sinusoidal movement. Stride efficiency results in constant speed and minimized acceleration of the center of gravity in a vertical direction.

As illustrated in FIG. 1, the sportsman is equipped with a device according to the invention, comprising an apparatus for measuring the heart rate or heart rate monitor 10 of a type known as, for example, that marketed by the firm Polar®. The measurement is made by means of a probe incorporated in a chest strap 12 placed on the athlete's chest. It emits a signal received by a wrist watch 13 equipped to process it and display the instant heart rate, including as a percentage of the athlete's maximum heart rate.

The device according to the invention further comprises an accelerometer 14 secured to the runner's chest. It is, for example, attached to the belt 12 of the heart rate monitor. Advantageously, the sensor used is of the ADXL 202E type marketed by the firm Analog Devices and provides a signal representative of the acceleration that the center of gravity of the runner undergoes in the three directions of space.

A processing circuit 15 receives the signal from the accelerometer 14 routed by transmission means, consisting of a transmitter 16a and a receiver 17, respectively associated with the accelerometer 14 and the processing circuit 15. To avoid any perturbation induced by background noise or by an artefact, the signal is corrected by a filter 18 of the bandpass type before arriving at the processing circuit 15. The latter also receives from a transmitter 16b information from the heart rate monitor 10 relating to the instantaneous heart rate and the percentage of maximum heart rate.

Circuit 15 processes this information in order to assess the effectiveness of the athlete's gesture, as will be described below.

Advantageously, the processing circuit 15 is housed in the watch 13 and the transmission means 16 and 17 are wireless. Of course, in this case, the emission frequencies of the signals from the probe of the heart rate monitor 10 and the accelerometer 14 are protected from one another to avoid any disturbance.

Finally, the device also comprises, placed at the outlet of the processing circuit 15, signaling means 19 to inform the sportsman about the effectiveness of his gesture. The signaling can be audible and the means 19 then consist of a loudspeaker also housed in the watch 13.

As shown in FIG. 2, the processing circuit 15 is conventionally provided with a microprocessor 20 and a memory 21. The microprocessor 20 is programmed to process the heart rate information supplied by the heart rate monitor 10 and the acceleration signal supplied by the accelerometer 14, performing the following operations at specified intervals:<ul id="ul0010" list-style="dash" compact="compact"><li>in 22, extraction of at least one attribute <i>AT</i> the signal from the accelerometer 14,</li><li>in 23, normalization of the attribute in order to overcome the influence of inter-subject variability,</li><li>in 24, comparison of the standard attribute <i>AT</i> with a reference value <maths id="math0008" num=""><math display="inline"><mrow><mover accent="true"><mrow><mtext mathvariant="italic">AT</mtext></mrow><mo>¯</mo></mover></mrow></math><img file="EP1586353A1_D0008.tif" /></maths> stored in memory 21 and corresponding to the heart rate measurement delivered by the heart rate monitor 10, and</li><li>at 25, providing the signaling means 19 with a representation of the effectiveness of the gesture.</li></ul>

More specifically, the attribute <i>AT</i> extracted at 22 is the standard deviation of the torso acceleration signal delivered by the accelerometer 14. Every 5 seconds, this standard deviation is estimated on the basis of the measurements made during the last 10 seconds.

Standardization 23 of the attribute <i>AT</i> is required for further comparison. The frequency of the stride having a direct impact on the acceleration undergone by the runner's torso, this normalization is carried out with respect to this frequency, directly extracted from the signal provided by the accelerometer 14, advantageously by the method of zero crossings of the signal. Indeed, according to the sinusoidal modeling of the displacement of the center of gravity of the runner, there are two zero crossings of the signal in each of the periods.

The reference value used for the comparison operation 24, contained in the memory 21, is the average <maths id="math0009" num=""><math display="inline"><mrow><mover accent="true"><mrow><mtext mathvariant="italic">AT</mtext></mrow><mo>¯</mo></mover></mrow></math><img file="EP1586353A1_D0009.tif" /></maths> of the attribute, obtained for a reference population. However, it is obvious that the effectiveness of the stride is not the same during a simple jogging or a competition. We must therefore compare the attributes of the subject and the reference runners to a similar intensity of effort.

In practice, the average <maths id="math0010" num=""><math display="inline"><mrow><mover accent="true"><mrow><mtext mathvariant="italic">AT</mtext></mrow><mo>¯</mo></mover></mrow></math><img file="EP1586353A1_D0010.tif" /></maths> of the attribute is saved in the memory, at the factory, for different values ​​of a force parameter P_i, which is a percentage of the maximum heart rate. The attributes of the reference population are obtained by running its members at stable threshold intensities, typically 60%, 70%, 80%, 85% of their maximum heart rate. These percentages, given as an example, correspond to an audience of endurance runners. For each of the thresholds, the average of the values ​​obtained,<maths id="math0011" num=""><math display="inline"><mrow><mover accent="true"><mrow><mtext mathvariant="italic">AT</mtext></mrow><mo>¯</mo></mover></mrow></math><img file="EP1586353A1_D0011.tif" /></maths>P₁= <maths id="math0012" num=""><math display="inline"><mrow><mover accent="true"><mrow><mtext mathvariant="italic">AT</mtext></mrow><mo>¯</mo></mover></mrow></math><img file="EP1586353A1_D0012.tif" /></maths>60%, <maths id="math0013" num=""><math display="inline"><mrow><mover accent="true"><mrow><mtext mathvariant="italic">AT</mtext></mrow><mo>¯</mo></mover></mrow></math><img file="EP1586353A1_D0013.tif" /></maths>P₂= <maths id="math0014" num=""><math display="inline"><mrow><mover accent="true"><mrow><mtext mathvariant="italic">AT</mtext></mrow><mo>¯</mo></mover></mrow></math><img file="EP1586353A1_D0014.tif" /></maths>70%, <maths id="math0015" num=""><math display="inline"><mrow><mover accent="true"><mrow><mtext mathvariant="italic">AT</mtext></mrow><mo>¯</mo></mover></mrow></math><img file="EP1586353A1_D0015.tif" /></maths>P₃= <maths id="math0016" num=""><math display="inline"><mrow><mover accent="true"><mrow><mtext mathvariant="italic">AT</mtext></mrow><mo>¯</mo></mover></mrow></math><img file="EP1586353A1_D0016.tif" /></maths>80% and <maths id="math0017" num=""><math display="inline"><mrow><mover accent="true"><mrow><mtext mathvariant="italic">AT</mtext></mrow><mo>¯</mo></mover></mrow></math><img file="EP1586353A1_D0017.tif" /></maths>P₄= <maths id="math0018" num=""><math display="inline"><mrow><mover accent="true"><mrow><mtext mathvariant="italic">AT</mtext></mrow><mo>¯</mo></mover></mrow></math><img file="EP1586353A1_D0018.tif" /></maths>85%, is saved in memory 21.

During the comparison operation 24, the microprocessor 20 searches the memory 21 for the reference value <maths id="math0019" num=""><math display="inline"><mrow><mover accent="true"><mrow><mtext mathvariant="italic">AT</mtext></mrow><mo>¯</mo></mover></mrow></math><img file="EP1586353A1_D0019.tif" /></maths> corresponding to the percentage of the maximum heart rate indicated by the heart rate monitor 10. This value is either directly extracted from memory if the percentage corresponds to a threshold intensity, or calculated by interpolation if the percentage is between two values ​​of the parameter P_i.

The comparison between the attribute measured <i>AT</i> and its average reference value <maths id="math0020" num=""><math display="inline"><mrow><mover accent="true"><mrow><mtext mathvariant="italic">AT</mtext></mrow><mo>¯</mo></mover></mrow></math><img file="EP1586353A1_D0020.tif" /></maths> is advantageously carried out, typically every five seconds, using the Mahalanobis formula. This makes it possible, in the case where several attributes are taken into account in the comparison, to weight the calculation by the influence of each attribute according to its reliability. To do this, it involves normalization by the covariance matrix of the average attributes of the reference population.

The formula is written:<maths id="math0021" num=""><img file="EP1586353A1_D0021.tif" /></maths><ul id="ul0011" list-style="dash" compact="compact"><li>in which: <i>d</i>_(<i>ƒc</i>) is the difference between the runner and the reference population for a given percentage <i>ƒc</i> maximum heart rate,</li><li><maths id="math0022" num=""><math display="inline"><mrow><mfrac linethickness="0"><mrow><mtext>ρ</mtext></mrow><mrow><mtext>θ</mtext></mrow></mfrac></mrow></math><img file="EP1586353A1_D0022.tif" /></maths> is the value of the attribute determined for the runner or, in the general case, the vector whose components are the various attributes measured,</li><li><maths id="math0023" num=""><math display="inline"><mrow><mfrac linethickness="0" numalign="left" denomalign="left"><mrow><mtext>ρ</mtext></mrow><mrow><msub><mrow><mtext>µ</mtext></mrow><mrow><mtext mathvariant="italic">reƒ</mtext></mrow></msub></mrow></mfrac></mrow></math><img file="EP1586353A1_D0023.tif" /></maths> is the average of the attribute determined for the reference population or, in the general case, the vector whose components are the average values ​​of the attributes for said reference population,</li><li><i>VS</i>_<i>ƒc</i> is the covariance matrix of the mean values ​​of the attributes of the reference population,</li><li><i>T</i> indicates that the corresponding matrix is ​​transposed.</li></ul>

In the embodiment described, only one attribute, the standard deviation of the signal supplied by the accelerometer, is taken into account, but the formula makes it possible to use several attributes.

The gap <i>d</i> , from comparison 24, is an index of stride efficiency. It is transcribed, at 25, to the runner by the signaling means 19. A correlation scale is stored for this purpose in the memory 21 to make the index calculated significant by associating it with a frequency of an intermittent sound signal. Thus, for example, the greater the difference calculated using the Mahalanobis formula, the higher the frequency of the signal. The index can also be displayed digitally on the watch screen 13.

Of course, the runners constituting the reference population were chosen according to their performance and the quality of their stride, judged by experts of the discipline according to traditional criteria. A runner with a less academic style will not be taken into account in this sampling, so as not to distort the reference values.

The example which has just been given is in no way limiting and other embodiments can be proposed, without departing from the scope of the invention. For example, to examine the various points specific to running mentioned above, one can use several accelerometers providing several attributes integrated into the Mahalanobis formula. Furthermore, sensors other than accelerometers can be envisaged to give signals from which the attributes representative of the efficiency are measured.

The device can also be adapted to other sports than running by modifying the position of the accelerometer. In particular, in cycling, it is interesting to analyze pedaling by having a sensor on the foot or on the ankle of the cyclist. Gymnasts can also use a device according to the invention to study their rotation during acrobatics. In this case, the use of a heart rate monitor and taking into account the intensity of the effort to make the comparison is not compulsory.

The heart rate monitor used in the device can also operate on the basis of any other measurement system, such as, for example, an optical measurement described in document EP 1 297 784 A1.