We use cookies! Read Confidentiality, Terms & Conditions!. Close (auto)
EFSMA european federation of sports medicine associations
Newsletter Medicina Sportiva
Romanian Society for Sports Medicine

Home       • Contact        • Sport Medicine Journal      • International Events Calendar     • Information

"Sport Medicine Journal" No.6 - 2006

The study on the efficiency of
electromyographic biofeedback in the posttraumatic recuperation of the knee

Oraviţan Mihaela
University of West Timişoara, Sport and Physical Education Faculty

Abstract. The electromyographic biofeedback has been associated to the classical protocol of recovery (consisting of physiotherapy and physical therapy) on an experimental lotof 60 patients. The efficiency of this method that is relatively new in our country has been pointed out by comparing the development of the patients part of the experimental group with the development of the witness lot (66 patients) which has been treated only by means of a kinetic program and physiotherapy procedures.
The comparison criteria were joint mobility (flexion and knee extension), muscular force (of the quadriceps femoris) and the parameter of the KOOS scale measured each month at all 126 patients monitored during the 6 months.
The statiscal analysis of the results shows that by combining the electromyographic biofeedback with the classical recovery method improves significantly most of the parameters monitored during posttraumatic rehabilitation (joint mobility, muscle force, recovery period).

Key words: electromyographic biofeedback, rehabilitation, posttraumatic knee.


The knee is the joint that is mostly exposed during sports activities. The treatment and the recovery of the posttraumatic injury of this joint is submitted to complex protocols which have as final objective the patient’s quick and best integration into the activities performed before the trauma. This demand is obviously most important for sportsmen. One of the quite new methods that have been applied for some years in our country too is the one of the electromyographic biofeedback which completes the classical recovery protocols consisting of kinetic programs and physiotherapy procedures adapted to the injury type, its severeness as well as to the individual particularities of the patient (age, sports activity before the trauma etc.) [2,3,10-14]. In this study I have tried to highlight the advantages of the electromyographic biofeedback in comparison to the classical recovery regarding the quality and the rapidity of the recovery which crucial for sportsmen.

The recovery after a posttraumatic injury of the knee has as main objective the reduction or even elimination of posttraumatic after-effects regardless of the applied orthopedic-surgical treatment (ligament reconstruction, immobilization by means of plaster cast etc.). Great emphasis is given to the recovery of the quadriceps femoris. However, very often the specific exercises are performed with great difficulties due to the inflammations or to the post-surgical effects which occur within the first days/weeks (edema, tumefaction, pain) and due to a possible affection of the joints proprioceptors [7,10]. The feedback loop which determines the proper functioning of the muscle is out-of-tune as well as the facilitating or inhibiting influences which act physiologically upon the musculature of the knee and the patterns of the muscle contraction become less efficient. All these phenomena may diminish the efficiency of the exercises applied within the recovery program for the improvement of muscle control and muscle force [2,3]. Up to now, the electromyographic biofeedback has been applied with great success in neurology [4], in the posttraumatic rehabilitation of the hand, in shoulder instability [1], in case of posttraumatic spine injuries and cerebral vascular injuries [4], after meniscus injuries [11] and after the reconstruction of the cruciate ligaments[2,3].

Material and method
For this study I have chosen by random 126 patients which have been classified into two lots: the experimental lot (EL: 60 patients) and the whiteness lot (WL: 66 patients). The structure of both studied lots is illustrated in table I and, as it can be noticed, it is similar in what concerns the age, gender and injuries of the patients). On both lots there have been applied classical recovery protocols (consisting of specific kinetic programs, hidrokinetotherapy, masage and physiotherapy).
For the experimental lot I have applied additionally the electromyographic biofeedback. I have used a biofeedback and electrical stimulation apparatus Myomed 134 produced. Like most of the high-performance biofeedback apparatuses, the Myomed 134 apparatus enables the setting of the recovery protocols which will be selected by the therapist and applied afterwards on various patients, depending on the status of the injured knee, on the recovery stage and on the individual particularities of each patient (this recovery method being known also for its adaptability to the previously mentioned factors).
Table I – studied lots

Parameter Experimental Lot Whiteness Lot
The structure of the lot
  • Total number of patients
  • Patients of feminine gender
  • Patients of masculine gender
  • Medium ages

60 cases

30 cases

30 cases

29,2 years

66 cases

31 cases

35 cases

30,8 years
Traumatic injuries
  • III degree sprains
  • Meniscus injury
  • II degree sprains

20 cases

18 cases

22 cases

22 cases

20 cases

24 cases

Figura 1.a

Figura 1.b

Figure 1, a and b
a- Usage of the Myomed 134 apparatus connected to the PC; on the PC monitor one may observe a contraction-relaxation-stimulation protocol ;
b – Electrodes on

Table II The parameters of the protocols used during the study

Protocol Therapy Submenu
1 EMG Contraction 5”/ relaxation 5”
2 EMG Contraction 7”/ relaxation5”
3 EMG Contraction 10”/relaxation5”
4 EMG Pattern
5 EMG Continuous contraction
6 EMG Contraction 5” / Relaxation5” on 2 channels
7 EMG Pattern on two channels
8 EMG + Stimulation Relaxation5”/ contraction 5” / stimulation 5”
9 EMG + Stimulation Relaxation5”/contraction5”/stimulation 5” on 2 channels
10 EMG + Stimulation Stimulation under the lower threshold value
11 EMG + Stimulation Stimulation over the lower threshold value

When the apparatus is connected to the computer, the protocols may be graphically illustrated showing clearly the different stages of the protocol (contraction, relaxation, electrical-stimulation) (figure 1).
The protocols selected for the patients recovery who suffered knee traumas are illustrated in table II.
The initiation of the recovery program by using the EMG biofeedback will be made in dependence on the diagnostic, the patient’s state and its recovery stage. The first protocol from table (protocol 1) will be used for most of the patients in the preoperatory, immediate posoperatory or post-immobilization stage. As the patient will carry out a succession of contraction and relaxations of 5” with the EMG approaching normal parameter it will be switched to a more difficult protocol (protocols 2 and 3).
If the patient doesn’t show contraindications regarding electrotherapy, then the protocols 8-9 can be used in parallel to the protocols 1-3. Thus in the initial stage of the recovery protocol 8 (figure 3) and afterwards protocol 9. The comparison with the healthy lower membre is very important in the estimation of the recovery level and may be carried out by means of EMG by selecting 2 channels. This type of examination has been included in the protocols 6 and 7 (figure 2).

Protocols 10 and 11 are meant exclusively for the electrical-stimulation via TENS.

The patient will be first trained about the protocol to which he well be submitted and will follow the execution of the therapist’s indications respectively of the biofeedback apparatus.The pattern type protocols (as in case of protocol 4 and 7) require the adjustment of the apparatus in such way as to allow the patient to make the contractions and relaxations which will be carried out within certain limits set by the therapist. Some bars will be displayed on the monitor (figure 2). Their widths (amplitude in microvolts) are defined by the therapist according to the recovery stage. Thus, when first using this type of protocol,



Figure 2. The monitor of the PC in case of a pattern protocol (protocol 7)

Figure 3 The monitor of the PC in case of a contraction / relaxation / stimulation protocol (protocol 8)

he used ranges will be larger becoming more and more narrower once the patient’s control over the contraction and relaxation processes increases. Generally, this type of protocol is used in later stages of recuperation when the patient has already improved his neuromotor control and is able to graduate the contraction process. The protocols 8 and 9 combine the biofeedback process with the electrical stimulation process via TENS; protocol 8 so as it is represented on the PC’s monitor consists of 5 cycles of contraction5”, relaxation 5” and electrical stimulation 5”. The electrical stimulation is performed by means of low frequency currents respectively through subcutaneous nerve stimulation (SNET or TENS according to the internationally approved terminology).
Assessment methods
The elements that have been monitored on the patients were joint mobility (flexion and knee extension though goniometry), the strength of the thigh muscles (of the quadriceps femoris and of the hamstrings through manual examination), the electrical potentials in contraction and relaxation at the quadriceps femoris (assessed electromyographically by means of the Myomed 134 apparatus). Figures 4 and 5 illustrate two electromyographic pictures of a patient in different recovery stages.

Figure 4. EMG form for protocol No. 1 (contraction- relaxation) in Microsoft Word
(Patient S.C., 23 years, WL 35, 1 month postoperative)


Statistics global Channel 1

Work min 0 µV
Work max 16 µV
Work avr 8 µV
Rest min 0 µV
Rest max 13 µV
Rest avr 6 µV
Work Stand. Dev. 8,23 µV
Rest Stand. Dev. 6,38 µV
Rising-time 0,49 s
Drop-time 0,77 s
Figure 5. EMG form for protocol No. 1 (contraction- relaxation) in Microsoft Word
(Patient S.C.., 23 years, WL 35 Status after anterior cruciate ligament reconstruction, 3 month postoperative)


Statistics global Channel 1

Work min 0 µV
Work max 58 µV
Work avr 37 µV
Rest 0 µV
Rest max 14 µV
Rest avr 8 µV
Work Stand. Dev. 9,92 µV
Rest Stand. Dev. 3,27 µV
Rising-time 0,60 s
Drop-time 1,10 s

The patients have been assessed each month also through the KOOS scale Knee Injury and Osteoarthritis Outcome Score) by means of a set of 5 questions which were meant to asses the P-pain, S-other symptoms, ADL-daily activities, PS- sport and recreational activities, QOL-life quality.

In this study I have chosen the difference between the electrical potential as parameter to be monitored because in my opinion it comprises the progression of the contraction potential (the average value of the electrical potentials during the contraction cycles from within a protocol) as well as the potential from within the relaxation (from within the same protocol).

The greater this difference is the grater is the neuromotor control of the physiological processes specific to the monitored muscle as well as the developed force. Furthermore, the average value of the contraction potentials depends to some extent also on the reaction time (rising-time), respectively on the time required starting from the transmission of the visual or acoustic signal up to the accomplishment of the contraction potential. The average value of the electrical potentials from the relaxation period is also influenced by the rapidness with which the muscle is able to relax (drop-time). Generally, the general rule concerning the electrical potentials collected from a joint is that in case of a muscular dysfunction occurred at a certain movement, the spent energy is much greater than in case of a healthy muscle, even if that movement is carried out with a minimum of effort. In other words, in 95% of cases, the amplitude of electrical potentials is greater in case of the symptomatic muscles [365]. The same phenomenon occurs in case of electrical potentials of pause. For this reason I have chosen to estimate de difference of potential that includes both values.


I have established a statistical analysis of the quantifiable data of all tests and assessments made within a time period of 6 months (time period in which each patient has been monitored) and illustrated it in form of the graphics 1-3. I have shown the parameters sensible to the rehabilitation programs which are the objective translation of the recovery stage of each patient, respectively of each lot.

On the 6 main axes of the diagrams I have shown the progression of the followings parameter:

  • Active flexion in the knee joint (FA active flexion);

  • Quadriceps femoris muscle force (FQ – quadriceps force);

  • The difference between the medium electrical potential during the contraction of the vast medial muscle and the electrical potential during relaxation within a contraction/relaxation protocol (protocol 1) (DP potential difference);

  • Capacity to carry out daily activities (ADL - activities of daily living);

  • Capacity to carry out sport and recreational activities (SPsport and recreational activities);

  • Quality of life (QOLquality of life).

In order to have a total evidence of the progression of these values, they have been expressed in percentage compared to the physiological values considered as usual for the respective parameter. The data considered to be 100% for each monitored parameter are:
  • Active flexion 140°

  • Quadriceps femoris muscle force 5

  • The difference between the electrical potential 10 μV

  • The values ADL, SP and QOL are computed from the beginning of the assessment as being related to a knee that didn’t suffer any injury (these are P=100, S=100, ADL=100, SP=100 and QOL=100) so that they will be graphically represented with their actual values.

In the cases in which the potential difference has shown a negative result (the average electrical potential between the relaxation periods was greater than the one from the contraction periods), which may occur very often at the beginning of the rehabilitation, the values have been considered void in order to be illustrated in the upper diagrams and shown in the respective graphics.

Considering the fact that the values from the initial phases were alike, I found it necessary, from a statistic point of view, to asses the progression of the two lots WL and EL at an interval of 1 month, 3 months respectively 6 months

Graphic 1

Graphic 2

Graphic 3

Graphics 1-3

The compared progression of the parameters monitored as 1, 3, respectively 6 months posttraumatic
Experimental lot
Whitness lot
  • ADL: capacity to carry out every day activities
  • SP: capacity to carry out sport and recreational activities
  • QOL: life quality
  • FA: active flexion of the knee
  • FQ: quadriceps femoris force
  • DP: the difference of electrical potential between the contraction and relaxation of quadriceps femoris


The results obtained from repeated assessments have been statistically analyzed by using the test t impar and finally the correlation index Pearson. The statistical analysis made between the progression of the 6 previously mentioned parameters are presented in the section that follows.

The above statistical analysis has allowed the following assessments: in what concerns the evolution of the capacity to perform daily physical activities, there are unsignificant differences (p>0,05) between the two lots; in other words, classical recovery ensures enough rehabilitation for such activities within the normal period limits of the patient (tabel III).

Pursuing the evolution of active mobility in the knee articulation (active flexion) has shown the fact that there are significant differences(p<0,05)) between the two lots after one month of starting the recovery and unsignificant differences after 3 and 6 months of starting it (p>0,05); classical recovery thus determines enough recuperation to recover active knee mobility, except for the initial recovery period when electromyographic biofeedback has caused major changes (tabel IV).

In regards of the evolution at the quadriceps femoris force there are significant differences(p<0,05) between the two lots after only one month and after 3 months from starting the recovery and extremely major differences 6 months after having started this (p<0,001) (tabel V); classical recovery determines a less efficient rehabilitation like the one where electromyographic biofeedback has been associated as a complementary recovery method.

This difference is significantly bigger on a long term, considering that the recovery of motor patternsthrough biofeedback plays an important role in muscle force recovery, along with the patient’s consciousness and direct implication in the rehabilitation program.

In regards of the electrical potential difference at the quadriceps femoris (table VI)there are extremely major differences (p<0,001)between the two lots, at one, three and six months from starting the recovery.

Monitoring the electrical potential difference at the quadriceps femoris has allowed us a more precise quantification than the manual testing of muscle force and has also allowed to consider the muscle relaxation process, which is also affected in the initial recovery phases; the certainty of a valid result is also given by making the average of five contraction-relaxation cycles and including the latency times necessary for these two muscle phenomena (almost of same importance in the recovery process) to appear, into the calculation.

Statistical analysis of SP evolution (table VII) has allowed us the following assessments: in regards of the capacity to perform sports activities or other recreational physical activities, emerges the fact that there are significant differences(p<0,05) between the two lots, at 3 and 6 months from starting the rehabilitation program and unsignificant differences (p>0,05) after one month; in other words, the differences are minor, because the posttraumatic knee status only allows performing a very short number of physical activities after one month (and only the ones from the rehabilitation program), while after 3 and 6 months, the differences between the two lots becomesignificant, the EL showing higher values in KOOS score for SP.

In table VIII the values of linear correlation Pearson factors are shown together with the p probability values which prove the significance of linear correlation. Studies have been performed upon the correlations of DP values to FQ, SP and ADL for the EL lot at one month, 3 months, 6 months and the average values obtained for these time periods.

Atone month
At3 months
At6 months
Witness lot75,44±9,99 94,19±4,53 99,28±2,45
Experimental lot73,67±10,54 94,1±4,96 99,53±1,05
p 0,33660,9179 0,4664

Table III. Statistical analysis of ADL evolution between WL and EL
Atone month
At3 months
At6 months
Witness lot98,09±8,69 127,74±9,02 135,7±8,27
Experimental lot101,02±7,42 126,87±10,21 137,4±8,53
p0,0437 0,61230,2582
Table IV. Statistical analysis of range of motion evolution(active flexion) for WL and EL

Atone month
At3 months
At6 months
Witness lot 4,29±0,53 4,62±0,38 4,89±0,21
Experimental lot4,5±0,56 4,77±0,35 5±0,03
p 0,03170,0192 0,0002
Table V. Statistical analysis of muscular force evolution for WL and EL
Atone month
At3 months
At6 months
Witness lot-0,11±3,7 2,86±2,73 6,19±2,17
Experimental lot3,32±2,39 6,66±2,04 9,58±1,71
p <0,001<0,001 <0,001

Table VI. Statistical analysis of DP evolution for EL and WL

At one month
At 3 months
At 6 months
Witness lot 37,42±17,83 62,58±27,22 97,58±4,14
Experimental lot 37,33±17,41 72,50±21,30 99,5±5,65
p 0,977 0,0238 0,0328

Table VII. Statistical analysis of SP evolution for EL and WL

Coeficient PearsonAt
one month
3 months

DP cu FQ0,630 0,054 -0,218 0,437
DP cu SP0,110 0,514 0,523 0,102
DP cu ADL0,050 0,183 -0,085 0,110
p 0,702 0,158 0,515 0,217

Table VIII.
Display of the correlation between followed parameters


The conclusion of these results is that the average values of DP (determined at the three time periods) with the average values FQ are in a direct and significant average linear correlation (r=0,437, p=0,0254). At one month, the correlation is direct and strong (r=0,63 cu p<0,001).

DP with SP at 3 and 6 months are correlated directly, significant and average (r=0,514 at 3 months, respectively r=0,523 at 6 months, in both cases p<0,05).

DP with ADL show a unsignificant and very weak linear correlation.

The statistical analysis of data obtained by articular, muscle, electromyographic testing and by making the KOOS scale at different intervals within the posttraumatic knee rehabilitation process has allowed us to point out the fact that there is a significant correlation between muscle testing evolution (realized by manual testing of the quadriceps femoris and the electrical potential difference measured by electromyography); this one is strong at one month (confirming the hypothesis according to which the recovery of neuro-motoric patterns, muscle coordination and muscle force as such represent essential targets in posttraumatic knee recovery; this correlation is even stronger for cases where surgical intervention has been performed and where the posttraumatic disturbance has obviously been greater.

The weak linear correlation between DP and ADL confirms, if still necessary, the idea that electromyographic biofeedback does not influence in a significant way the obtaining of a “street knee”, being similarly performed for both lots, EL and WL; direct, significant and average correlation between DP and SP confirm though the hypothesis that electromyographic biofeedback contributes to a faster reintegration into the previous sport activity.


1.Electromyographic biofeedback majorly changes muscle force recovery and, implicitly, the electrical potential difference between contraction and relaxation of the quadriceps femoris as compared to classical rehabilitation in all the recuperation period phases, even if biofeedback training has only been performed in the first two months.

2.Associating electromyographic biofeedback to classical rehabilitation programs significantly changes the capacity to perform sport and recreational activities (at 3 and 6 months from starting rehabilitation), as well as the knee mobility recovery speed in the first posttraumatic period.

3.Regain muscle force, articular mobility and recuperation of neuro-motor coordination are the base of sport activity reintegration as before the accident.

4.The advantages of this method as opposed to other muscle force assessment techniques are:

  • objectivity: obtaining mathematical values transforms muscular contraction or relaxation into perfectly quantifiable elements, which turns this method into a very good monitoring means in posttraumatic recovery;
  • the response to this investigation is spontaneous, clear and easy to see, surface electromyography becomes the patient’s immediate response to the recuperation program;
  • interpretation is relatively easy, being sustained by the presence of an interpretation program which can be used through PC connection;
  • the possibility to apply it in any recuperation phase, starting with the preoperative phase (when necessary) and even postoperative or posttraumatic phases (in case of conservative treatment);
  • relatively high accessibility;
  • continuous possibility to compare with the counterlateral limb, which is considered to be a reference value;
  • this method ensures the progressive, quantified increase of the threshold value and thus the control upon the progressiveness principle in the rehabilitation process;
  • the method is untraumatizing; it can be applied ambulatory, at the patient’s home.

1.BEALL, M.S., DIEFENBACH, G., ALLEN, A.(1987) -Electromyiographic biofeedback in the treatment of volunteer posterior instability of the shoulder, Am. J. Sports Med15: 175-178;

2. DRAPPER ,V. ŞI COLAB.(1991) Electrical stimulation versus electromyographic biofeedback in the recovery of quadriceps femoris muscle function following anterior cruciate ligament surgery, Phis Ther; 71 (6); 445-64;

3. DRAPPER, V. (1990) Electromiographic biofeedback and recovery of quadriceps femoris muscle function following anterior cruciate ligament reconstruction, Phis Ther, 70(1):11-17;

4. NACHT, M.B., WOLF, S.L., COOGLER, C.E. (1982) Use of electromyoographic biofeedback during the acute phase of spinal cord injury, Phys Ther 62: 290-294;

5. SELLA, G.E. (2002) Muscles in motion: Surface EMG Analysis of the Human Body Range od Motion,vol. I, 3rd Edition Revised, Martins Ferry, OH: Genmed Publishing, pp:231-238;

6. SELLA, G.E.(2002) Muscular Dynamics: Electromyography Assesment of Energy and Motion, GenMed Publishing, Martins Ferry, OH, pp.320:345;

7. SELLA, G.E.(2005) Muscular dysfunction: an SEMG view of investigation and rehabilitation of the lower limb muscles, OH, Genmed Publishing, 2005;

8. SELLA, G.E.(1999) SEMG of the Hip ROM Protocol: A study of Consistency and Repeatability of Electrical Activity of 19 muscles, Europa MedicoPhysica, vol.35, no.2. pag.: 83-92;

9. SELLA, G.E. (2001) SEMG: Muscular Assesment Reference Manual, OH, Genmed Publishing, pp:35-56;

10. SODERBERG, G.L., COOK, T.M. (1983) An electromiographic analysis of quadriceps femori muscle setting and straight leg raising, Phys Ther, 1983, 63: 1434-1438;

11.SPRENGER, C., CARLSON K., WESSMAN, H. (1979) Applications of electromyographic feedback following medial meniscectomy, Phys Ther 59: 167-169

Web creator - Gabriel Panait M.D. Information belongs to S.Ro.M.S -  contact person Mirela Vasilescu M.D.

S.Ro.M.S 's web section is freely created and hosted on   www.medicinasportiva.ro  -  according to Terms & Conditions

Online since
March, 23 2006
www  www.medicinasportiva.ro

Website Statistics
Copyright© MedicinaSportiva.Ro & Partners. All rights reserved.