Instrumented Gait Analysis
Original Editor - Mariam Hashem
What is Normal Gait?[edit | edit source]
Normal gait is a range of typical gait patterns, found in a healthy population, presenting similar characteristics. People present a certain degree of variability that is called inter-subject variability and is due to differences in age, gender, muscular strength and anatomical differences.
For example. if we are to assess the gait of an 80-year-old man and a 20-year-old woman, both without any pathologies affecting gait, they might present completely different gait patterns. However, they will both be considered physiological or normal because their gait patterns will be within the range of normality corresponding to their own population. So when assessing the gait of a patient in physiotherapy, the idea is to confront the gait parameters we find with our own patient against the range of normality established for the corresponding population.
For the purpose of providing a non-biased interpretation of the results, it is important to acknowledge all sources of data variability. Once these sources of variability are acknowledged or considered negligible, results can be interpreted with higher confidence.
- Inter-subject variability: refers to the difference of gait parameters obtained with different subjects. It can be due to anatomical differences, or difference in muscular strength These differences often result in variation of spontaneous walking speed, an indicator of one’s own 2 adjustment to reach the lowest energy expenditure while walking.
- Within-subject variability: refers to the possibility of obtaining slightly different gait parameters on two different trials with the same person. This can be caused by small changes in one person's gait from one trial to another due to, for example, stress or apprehension or the desire of the patient to do his best performance. Also, it can be caused by the measurement techniques which give slightly different results under the same conditions
Visual Gait Analysis[edit | edit source]
Visual gait analysis is commonly used by physiotherapists to investigate gait problems in their patients. With increasing numbers of low-cost or freely available apps, more physiotherapists are using smartphones to assess their gait in clinical practice.
- Quick and easy method
- Allows peer-reviewing, by showing your colleagues the videos and discussing the presented case
- Reproducible, by taking multiple videos you can track your patient's progression
- Allows observing the gait from multiple angles to detect deviations in multiple plans
- The reliability is not very high and it depends on the clinical experience
- It doesn't allow observing high-velocity events, force in moments during walking
- Subjective - it depends on the observer and is, therefore, prone to error
Instrumented Gait Analysis[edit | edit source]
Three-dimensional instrumented gait analysis has helped to increase our knowledge of gait pathology and treatment. Literature suggests that instrumented gait analysis is a valuable tool in clinical practice for the diagnosis, assessment, and management of patients affected by conditions that alter their gait. In the research field, instrumented gait analysis is often used to evaluate the effectiveness of treatments such as foot orthoses and to explore the consequences of pathologies related to gait, like Rheumatoid Arthritis. Also, many clinical interventions and evaluations are based on the findings of instrumented gait analysis conducted in research settings, for example in children with cerebral palsy, or amputees.
- To apply in clinical practice
- To get familiar with the terminology that is often used in research publications and scientific articles
In gait analysis, reliability refers to the consistency of the results across multiple repetitions (within-subject or absolute reliability) and the consistency of the results within a cohort of subjects (inter-subject or relative reliability). It is a key feature for making correct interpretation of the results and generalizes them in other conditions than the ones of the experiment. For example, information regarding the reliability of the measurements should be clearly exposed by authors publishing a randomized controlled trial using instrumented gait analysis as the method of assessment.
Types of Instrumented Gait Analysis[edit | edit source]
Electromyography[edit | edit source]
Electromyography (EMG) is the measurement of the electrical activity of a muscle during its contraction by measuring the motor unit potential action(MUAP) . A motor unit is a group of muscular fibres innervated by the same motoneuron. The MUAP refers to the characteristic shape of depolarization voltage of the motor unit. The signal recorded by EMG is a summation of the electrical activity from a number of MUAP that are simultaneously activated during a muscle contraction.
The information provided by EMG is electrical but not mechanical: it does not give information about the nature of the contraction (concentric, eccentric, or isometric) or the force produced by the muscle contraction. However, EMG facilitates the understanding of the onset of different muscle contractions through recording of signal’s magnitude and frequency, respectively referring to the amount of electrical activity during a contraction and the range of firing rates from the motor units recorded.
Within each muscle, motor units are hierarchically activated in order of increasing size. The bigger the motor unit is, the higher is the activation threshold. When a motor unit is activated during a sustained contraction, it fires repeatedly, and the series of firing times can be represented with a train of bars over time (firing train). Firing rate = number of pulses occurring in a second at a motor unit level. The force produced by a muscle is modulated by the number of active motor units and the firing rate.
EMG is frequently included in research using instrumented gait analysis, for example when exploring gait abnormalities of patients with neurological disorders, as it provides valuable input on the timing of muscle activation, synchronization with antagonists.
Three methods are reported in the literature for EMG recording: fine wire, needle electrodes and surface electrodes (sEMG) which is the most commonly used technique for gait analysis
|Surface EMG (sEMG)||Non-invasive, operator does not need training to place electrodes, highly tolerated, wireless/telemetry, possibility to use a protocol for electrode placements, various designs, allows prolonged recordings from multiple sites, most frequently reported in the literature, low-cost.||Records superficial signal records a group of motor units, movement of electrodes in relation to the bony references, inadequate for overweight patients, cross talk, noise, artefacts.|
|Intramuscular myography (fine wire and needle)||Allow recording deep muscles and single motor unit, signal is stable, higher amplitude and frequency recording, less sensitive than sEMG to noise.||Invasive, damages muscle fibres, risk of infection, can modify patient’s movement because of discomfort/pain, mechanical artifacts, cables, requires operator training, wire can break.|
Before undertaking surface EMG, clinicians need to place electrodes on the patient’s skin surface. For this purpose, palpation combined with “testing” (active muscle contraction) is helpful to identify as precisely as possible the location of muscle’s bodies (Ewins & Collins, 2014). In order to favour reproducibility of measurements, Hermens et al published recommendations for the “surface EMG for a non-invasive assessment of muscles”
In order to favor reproducibility of measurements, Hermens et al., (2000) published recommendations for the “surface EMG for a non-invasive assessment of muscles” (SENIAM). Below is a summary of the recommendations from the SENIAM (Hermens et al., 2000):
|Electrode shape and size||Researchers undertaking sEMG should report the type, manufacture and shape of the electrodes. The size of the electrode (size of the conductive area of the electrode) should not exceed 10mm.|
|Inter electrode distance||Bipolar sEMG electrodes should respect an inter-electrode distance of 20mm. For small muscles, the inter-electrode distance should not exceed one-quarter of the muscle fibre length.|
|Electrode material||The electrode material in contact with the skin should provide good contact, low impedance and stable behaviour in time.||Vertical|
|Sensor construction||Authors recommend fixed electrode distances using light materials. Cables motion should be limited with tape for example in order to reduce mechanical artefacts.|
|Sensor placement procedure||The procedure contains 6 steps for which recommendations are the following:
1. Selection of EMG sensor: adapted electrode (shape, type, size) depending on the muscle studied. 2. Skin preparation: shave the area, clean the skin with alcohol and keep the electrode contact area dry. 3. Positioning the patient for manual muscle testing. 4. Determining sensor location: sensors should be placed in a stable location. The ideal location is defined in relation to the longitudinal transversal location of the sensor on the muscle. For the longitudinal location: halfway between distal motor endplate zone and the distal tendon for longitudinal. For the transversal location: away from the edge of the muscle belly divisions and as close as possible as the middle of the muscle belly. 5. Sensor placement and fixation: electrodes orientation should be placed parallel to the muscle fibres. Electrodes should be stabilized using tape or elastic band in order to limit mechanical artefacts. 6. Testing of the connection: subject produces a voluntary contraction and clinician can observe in real time that the signal is recorded
Despite being the most commonly used technique, sEMG presents certain limitations. The technique is based on the measurement of the voltage difference between two electrodes and the use of a grounding electrode (somewhere else on the body)The electrical signal that reaches the electrode on the skin surface is usually small because it has been attenuated by the layers of fascia, fat and skin underlying (noise effect). Therefore, the signal is amplified (1000 to 10 000 times) close to the electrodes and picked-up by electrodes as the sum of the muscle action potentials from many motor units within the most superficial muscle(s)EMG signal is influenced by intrinsic and extrinsic variables which can affect the data recorded and therefore the reliability of interpretations.
|Variables||Cause||Consequence / Possible way to limit variable|
|Intrinsic||Tissue characteristic (type of muscle, motor unit size, presence of fat, skin temperature, etc.)||Affects electrical conductivity|
|Crosstalk (undesired recording of the activity of adjacent muscles)||Interference with EMG signal.
Selecting a group rather than a particular muscle
|Mechanical artefact (change in the electrical baseline happening when subject moves)||Contaminate EMG signal with “fake” activity.|
|Internal noise (thermal noise)||Contaminate EMG signal with “fake” activity.
Locating amplifiers very close to the recording electrodes, therefore reducing the length of wire
|Skin (preparation, thickness, sweat, etc.)||Impedance of the signal and signal amplitude|
|Extrinsic||Changes in muscle geometry between muscle belly and electrode site||Modifies EMG reading|
|External noise (amplifiers)||Contaminate EMG signal with “fake” activity|
|Electrodes (placement regarding the motor point, muscle fiber orientation, etc.)||Affects conduction velocity|
|Environmental conditions (temperature, humidity, interferences: electromagnetic disturbances coming from surrounding electrical equipment, etc.)||Signal amplitude|
|Contraction conditions (type of contraction, muscle length, etc.)||Signal amplitude|
|Signal processing (filters, etc.)||Signal amplitude|
Processing EMG Data: The first signal obtained after recording is called “Raw EMG” from which it is possible to see if a muscle is active or not. “Raw” EMG presents a wide range of frequency (from 5 to 500 Hz), artifacts, many data points, and positive and negative values which make the interpretation of the results difficult. In order to address this challenge, different techniques of data processing are available such as filtering, Root mean square, smoothing, Rectification, integration, and amplitude.
|Filtering||Reduces/eliminates unwanted signal associated with noise/artefact||Makes data interpretation easier||Not acknowledging all the data can bias the interpretation of the results, not all studies use the same filters
Selecting the correct filter is important to no lose valuable data
|Root mean square||Squaring all the values, taking the mean and finally the square roots||Removes negatives values
Linked with electrical power Preferred method in literature
|Details on the data use for the mean calculation should be provided|
|Smoothing||Average calculated for a specific window||Reproducible||Data details are lost because of the averaging|
|Rectification||Reflection of negative values to convert data in “all positive”||Makes statistical manipulation and reading easier|
|Integration||Mathematical calculation of the “area” under the EMG graphic function.||Useful for relatively long duration segments|
|Amplitude||Peak amplitude average of the first 10 highest peaks||Useful for comparison analysis
True mathematical integral under EMG amplitude
|Does not reflect force production or effect of force on the joint|
Normalization of Data: EMG data is often normalized (rescaled to a reference value) using various techniques in order to facilitate interpretation, for example when comparing EMG recordings between different days, subjects or muscles. Authors suggest that the technique for normalization should be chosen according to the possibilities of the patients, for example in producing a “true” maximal voluntary isometric contraction
|Method for normalization||Advantages||Limitations|
|Maximal voluntary isometric contraction||Frequently used in studies, limits the variability of EMG data, provides information about muscle activation||Not all patients are able to produce Maximal voluntary isometric contraction due to fatigue, participation from the patients (pediatrics), repetition in time|
|Sub maximal voluntary contraction (% of maximal voluntary contraction)||Useful in case the patient can not produce maximal voluntary contraction, easier to maintain||Still needs to maximal contraction values to normalize from|
|Instrumentally evoked contraction (stimulation)||Activates all motor units to reach “tetanus” level without any participation of the patient||Discomfort, doesn’t give information about patient’s activation pattern|
|Reference contraction: one movement is repeated by the subject and the middle period is taken as the reference contraction||Provides a stable reference value, closer to lower activation levels for muscles which do not physiologically contract at the maximal level (ex: tibial posterior during gait)||Discomfort, doesn’t give information about patient’s activation pattern|
|Mean or peak value of EMG set during a full gait cycle||Limits the variability of EMG data, reduces inter participants difference, allow to detect normal/abnormal muscle activation during gait||Expose data to confoundable variable: force-velocity relationship and change of muscle mass located under the electrode
Modifies EMG data amplitude
|Muscle activity at rest||Interesting for patients with neurological disorders||No information on the muscle and possible confusion between various muscle groups.|
Interpreting EMG Data: Once processed and normalized, clinicians can make various interpretations from EMG data such as the motor unit firing trains, which refer to the moments when the motor unit fires (MUAP): it informs clinicians about the activation/nonactivation of the muscle. In addition, the recruitment / de-recruitment threshold of a motor unit which can be interpreted as an indication of the force level at which the motor unit starts and stops firing during a contraction. Within the same muscle, different thresholds can be observed and be unlocked as the need for force increases. The shape of MUAP provides information about the morphology of the motor unit and the state of the motor fibers. EMG data can also be used to calculate the motor unit’s mean firing rate (number of pulses per unit of time) and the synchronization of a motor unit with another.
Reliability of surface EMG: EMG data, such as other types of data collected during gait analysis is quantitative and can therefore be evaluated in term of reliability using statistical manipulation. Smith suggest that the reliability of EMG data is an important consideration as the quality of the results depends directly on it. For example, inter-participant reliability is important for making valid comparisons between subjects from a control and intervention group, as it ensures that the differences observed are due to the intervention. Different features can be assessed for reliability such as the way the muscle activity is produced (evoked contraction / natural contraction), the interval between measurements, the comparability within-subjects, and environmental conditions, the material used for recording, For example, Kollmitzer et al  suggest that long-term intervals for measurement (6 weeks) were correlated with lower reliability; whereas high reliability could be obtained in short interval measurements (within 90min). These results suggest that EMG could be a valuable technique to undertake an evaluation of direct effect interventions such as orthotics for example (with/without trial). However, these results are questioning the quality of evidence brought to clinicians by EMG measurements, as they are often used to assess the effectiveness of training and rehabilitation programs showing a positive impact on patients after a long period of time (4 weeks and more).
Regarding normalization, Smith suggests MVIC has the best inter-participant reliability, followed by peak dynamic amplitude and single leg stance measurements. In addition, the choices made by researchers/ clinicians regarding data collection depend on the abilities of the patient to comply with rigorous research protocol, but they should be enounced explicitly so readers can be critical on the reliability of the results.
Kinematics[edit | edit source]
Kinematic data refers to the characterization of movement using the geometric description of bodies or segments over time . It encompasses linear and angular kinematics, respectively characterizing trajectories and angular position of body segments from one to another over time. In order to make mathematical analysis easier, patient body’s representation is simplified to a series of rigid bodies or segments moving in 3D (6 degrees of movement) from one to another. The clinical meanings of the data recorded are the following: motion of anatomical joints across the different planes (sagittal, horizontal and frontal) at the foot, ankle, knee and hip level. Cadence, walking speed, duration of stance phase, step length and size of the sustentation polygon are also part of kinematic data. Kinematic data can be particularly useful to link changes in motion with underlying forces and physiological events during gait, for example, heel strike and ankle dorsiflexion angle. It can also be used to characterize overall gait parameters such as mean walking velocity for comparison between healthy and pathological subjects.
Recording kinematic data:
The gold standard for kinematic data collection is the optoelectronic system (OS) that recognizes the spatial coordinates of reference points placed on the patient. Several infrared cameras are placed along the perimeter of a laboratory in order to obtain data from different angles coming from markers placed initially on point of interest, for example, bony prominences. The recording is based on infra-red light produced by the camera which is reflected by the markers and finally captured by the cameras. As the ambient light is not recorded, images obtained have a high contrast which makes the analysis easier. Positions of markers obtained is computed in reference to calibrated position of the cameras (often called “global coordinate system”), which is translated into 3 coordinates (often called “local coordinate system) in space. The use of multiple cameras also allows to prevent markers to be hidden during motion and increases the accuracy of the spatial reconstruction . In order to calculate accurate joint kinetics kinematics, it is necessary to place markers close to joints’ center of rotation which can be obtained by computation. Before starting data collection, researchers usually undertake a static and dynamic trial in order to proceed to verifications on the good placement of the markers and effective recognition from the cameras. Later, the subject is invited to follow a protocol for data collection: “walk at self-selected speed”, “work increasingly faster”, etc. Depending on the reason for referral, various protocols can be applied for markers placement. For example, if the focus of the analysis if the foot motion, a multi-segmented foot model will be used with multiple markers on different parts of the foot. If the focus in knee motion, then more markers will be placed on the knee, etc.
An alternative to optoelectronic system is the combination of accelerometers, gyroscopes and magnetometers which are wearable sensors placed on the subject (often reported as inertial measurement units IMUs). They have the advantage to be less expensive and easier to use than the OS, as they do not require cameras and as the data is acquired through a computed algorithm such as G-WALK for example.
Data is collected in time intervals called “frames” which varies between 50 and 200Hz depending on the system used for acquisition. Errors in markers position are relatively small (up to 1mm) but increases when position is used to calculate velocity and acceleration because of the use of mathematical differentiation “magnifying” errors. This bias can be addressed using accelerometers which calculate directly acceleration and therefore provide “clean” data. Another source of error is the movement of markers on patient’s skin, which authors believe to be little for movements calculated in the sagittal plane, whereas it can be considerable for movements in the horizontal plane. This problem can be addressed by increasing the number of markers and protocolized marker placement. Different methods are reported in the literature for kinematic data processing such as direct kinematics (DK) which is based on the displacement of markers initially placed on segments in order to calculate joint angles and inverse kinematics (IK) which uses a skeletal-joint model with embedded markers and transduces joint kinematics through adjustment of model’s angles and matching between model and experimental marker position.
Reliability of kinematic data:
Literature suggests high reliability for lower-limb and trunk kinematic obtained in the sagittal plane. A possible explanation is that the movements in the sagittal plane have greater amplitude and therefore are easier to detect. Some publications reported precision up to 1° or 1mm in joint motion, which can be considered clinically relevant . Correct marker placement is crucial to obtain reliable kinematic data. In addition, kinematic data may suffer from mechanical artefacts due to the motion of soft tissues while walking (“wobbling”), which makes the analysis significantly more difficult for overweight patients. In order to address this limitation, some authors recommend the use of elastic tape previously applied on patient’s limbs to ensure maximal stability of the markers.
kinetic data[edit | edit source]
Kinetic data describe forces and moments applied on the patient while walking such as the ground force reaction (GRF), the joint muscle forces and moments and the discrete pressure analysis of patient’s feet during stance phase. From each reaction studied, different values with clinical meaning can be isolated. For example, while undertaking “pressure measurement”, features such as the peak pressure (PP) applied on the foot during stance phase, the time during the peak pressure is applied called pressure time integral (PTI) and the motion of the center of pressure (instantaneous point of application of the GRF vertical component) can show “typical” patterns related to a known condition. Another example is the combination of kinetic, kinematic and anthropometric data which allows calculation of joint moments and powers (often called “inverse dynamics”), which can result increased in conditions such as osteoarthritis. This process involves calculations often called “inverse dynamics” which refers to the separation of the body in multiple segments and the treatment of each one of them as a free body, with multiple forces acting on them such as gravity, GRF, muscle and ligaments forces.
Recording kinetic data:
The different components GRF can be recorded with a “force platform” that calculates the force and moment (vector exchanged between the patient and the ground) and the coordinates of the center of pressure from which can be deduced speed and acceleration, PP and PTI. Other alternatives to force platform are insoles with embedded sensors or pressure mats that can be useful in the establishment of patient’s foot pressure profiles . Gait analysis protocols often recommend the force platform to be placed in the walking track used for gait analysis (usually a 10m long flat surface with homogeneous surface). Authors suggest that the platform should be as “hidden” as possible to prevent the patient to “aim” for it while walking (patients are susceptible to unconsciously modify their gait in order to step right on the platform). An alternative to the platform is an embedded system which is similar to an insole that goes inside the patient’s shoeing.
Kinetic data interpretation:
Variations in the GRF can be interpreted to identify pathological gait patterns. In their study, have shown that diabetic patients have increased antero-posterior component of the GRF during stance compared to healthy subjects. The horizontal component of the GRF is a shear force acting in a parallel direction to the plantar foot surface and can to initiate the development of a foot ulcer.
Furthermore, authors have shown that patients with established foot ulcer tend to transfer their weight on the intact limb, which can be interpreted as a protective compensation to prevent excessive loading which could exacerbate pain and ulcer development. This research has therefore clinical implications: patients with increased horizontal component of the GRF may develop plantar ulcers, and interventions should aim to reduce this feature with foot orthotics for example.
Another example of altered kinetics is found with children with cerebral palsy. Children with cerebral palsy may present “crouch gait” which is a gait pattern in triple flexion (hip-knee-ankle) during stance phase. During the stance phase, authors have demonstrated that the typical curve of the vertical component of the GRF is lost. This may be correlated to further bone deformities, soft tissue injury, poor balance and gait speed. Strengthening of lower limb extensors may therefore be a valid approach in this case.
Reliability of kinetic and kinematic parameters:
One limitation of kinetic data recording is that it does not take into account friction forces and visco-elastic properties of tissues, which result in a possible error in calculating muscle’s contribution to movement, for example.
Integration of kinetic, kinematic, and electromyographic data:
Once collected, data can be integrated to other features (kinematics, EMG) and 3D virtual representation of the patient's gait can be built using specialized software and allows specific and precise analysis of gait’s different point of interest, for example, the orientation of the GRF during heel strike. The interpretation of the data can be improved by replacing the patient's performance in the context of the analysis, including the patient's walking aids, fatigue, confidence while walking, motivation, pain, visual ability, balance, etc
The interpretation of data collected during instrumented gait analysis allows identifying impairments in walking by confronting patient’s data with data obtained with a similar healthy subject. From data can be deduced what compensations or strategies the patient develops to overcome a disorder and what are the implications of these compensations on the patient’s posture. Usually, instrumented gait analysis is broken down in the analysis of multiple groups such as foot, ankle, knee, hip, and pelvis; which make kinematic and kinetic analysis easier. Protocol for gait analysis is the following: patient walks as “naturally” as possible along a pathway within the lab, force platforms are hidden in the floor so the patient does not modify his gait, and recording is made for several repetitions in order to reduce error and increase the reliability of the results.
Combining kinetic and kinematic analysis allows us to calculate the moments of force and the power generated or absorbed across the different joints of the lower limb. For this purpose, information about masses, moments of inertia, and centers of gravity are necessary.
Data collected are used for a computed construction of a 3D model of the subject walking
Usually, the motion of anatomical angles is presented in a graph and normalized to a full gait cycle (0% representing heel strike on one limb and 100% representing the next heel strike of the following cycle). Events such as the “toe-off” usually occurring at 60% should be included as a reference. Data can also be normalized with the step length and the length of the person.
Once data is collected, it is possible to “score” patient’s gait patterns with validated scales such as the Gait deviation index, using for example multiple angles of videography. Other scales exist and clinicians should choose their assessment tool regarding the patient’s characteristics and reason for referral in order to ensure the best reliability.
Energy Expenditure[edit | edit source]
Ewins & Collins, suggest that oxygen uptake / carbon dioxide production are appropriate methods to evaluate energy expenditure while walking as they reflect the patient’s metabolism chemistry. Data collected is often normalized to the patient’s body mass function of time (mL/kg/min). Oxygen uptake/ carbon dioxide production can be recorded using a face mask, but one limitation is discomfort. As an alternative, the patient’s heart rate can be used as it is linearly supposed to be well correlated to oxygen consumption
Gait analysis in practice[edit | edit source]
Gait assessment protocol:
|Before appointment||Referral from specialist and multidisciplinary review of the case||Patient details and medical history, reason for referral, types of assessment required.|
|Set-up of appointment and laboratory||Equipment set up and calibrated.|
|During appointment||Patient review||Self-reported gait assessment in order to investigate patient’s ability to comply with data collection protocol.|
|Data collection and record||Data collection undertaken as close as possible as pre-established protocol depending on patient’s abilities.|
|After appointment||Data processing and interpretation, record and communication||Analysis undertaken by different reviewers and report of the conclusions of the examination communicated to the patient and the team.|
Challenges when undertaking instrumented gait analysis:
|Limiting influence from external parameters||Laboratory environment influences patient’s gait parameters and should therefore be taken into consideration when interpreting the results|
|Repeatability||The reliability of measurements undertaken in routine clinical use is limited. For example, the same patient evaluated in 2 different laboratories may show different results. A majority of reliability studies have been done with intact subjects|
|Anthropometry||This challenge encompasses the difficulty to place markers in an accurate/repeatable way and determining joint centers in relation to the markers.|
|Soft tissue artefact||This refers to the degree of movement of the soft tissues such as the skin, muscles, fat layer, etc. in relation to the skeleton structures while walking.|
References[edit | edit source]
- Levine D, Richards J, Whittle MW. Whittle's Gait Analysis-E-Book. Elsevier Health Sciences; 2012 Jul 13.
- Ewins D, Collins T. Clinical Gait Analysis. InClinical Engineering 2014 Jan 1 (pp. 389-406). Academic Press.
- Haentjens M. Instrumented Gait Analysis Course. Physioplus 2020
- Ancillao A. Modern functional evaluation methods for muscle strength and gait analysis. Cham, Switzerland: Springer International Publishing; 2018.
- Finkbiner MJ, Gaina KM, McRandall MC, Wolf MM, Pardo VM, Reid K et al. Video Movement Analysis Using Smartphones (ViMAS): A Pilot Study. J Vis Exp. 2017;(121):54659.
- Brunnekreef JJ, Van Uden CJ, van Moorsel S, Kooloos JG. Reliability of videotaped observational gait analysis in patients with orthopedic impairments. BMC musculoskeletal disorders. 2005 Dec 1;6(1):17.
- Roberts M, Mongeon D, Prince F. Biomechanical parameters for gait analysis: a systematic review of healthy human gait. Physical Therapy and Rehabilitation. 2017;4(1):6.
- Cervical Motor Control Example . Available from: https://www.youtube.com/watch?v=HoLbn1NhUGk[last accessed 08/06/2020]
- Winter DA. Biomechanics and motor control of human gait: normal, elderly and pathological. 1991.
- Wren TAL, Tucker CA, Rethlefsen SA, Gorton GE 3rd, Õunpuu S. Clinical efficacy of instrumented gait analysis: Systematic review 2020 update. Gait Posture. 2020;80:274-9.
- Robertson DG, Caldwell GE, Hamill J, Kamen G, Whittlesey S. Research methods in biomechanics. Human kinetics; 2013 Nov 1.
- AVEYARD, 2018. Doing a Literature Review in Health and Social Care: A Practical Guide. Fourth edition. Edition: Open University Press. UK. [online] [viewed 12/10/2019]. Available from: https://www.dawsonera.com/abstract/9780335248018
- Fernandes R, Armada-da-Silva P, Pool-Goudaazward A, Moniz-Pereira V, Veloso AP. Test–retest reliability and minimal detectable change of three-dimensional gait analysis in chronic low back pain patients. Gait & posture. 2015 Oct 1;42(4):491-7..
- Richards J. The Comprehensive Textbook of Biomechanics [no Access to Course] E-Book:[formerly Biomechanics in Clinic and Research]. Elsevier; 2018 Mar 29.
- SimiSystems. Gait Analysis with synchronized electromyography (EMG). Available from: https://www.youtube.com/watch?v=tFGHggMQ7IQ[last accessed 12/01/2021]
- Semple R, Murley GS, Woodburn J, Turner DE. Tibialis posterior in health and disease: a review of structure and function with specific reference to electromyographic studies. Journal of Foot and Ankle Research. 2009 Dec;2(1):1-8.
- Hermens HJ, Freriks B, Disselhorst-Klug C, Rau G. Development of recommendations for SEMG sensors and sensor placement procedures. Journal of electromyography and Kinesiology. 2000 Oct 1;10(5):361-74.
- Soderberg GL, Knutson LM. A guide for use and interpretation of kinesiologic electromyographic data. Physical therapy. 2000 May 1;80(5):485-98.
- Smith SL. Neuromuscular control in knee osteoarthritis (NEKO) (Doctoral dissertation, Glasgow Caledonian University).
- Graham JE, Karmarkar AM, Ottenbacher KJ. Small sample research designs for evidence-based rehabilitation: issues and methods. Archives of physical medicine and rehabilitation. 2012 Aug 1;93(8):S111-6.
- Kollmitzer J, Ebenbichler GR, Kopf A. Reliability of surface electromyographic measurements. Clinical Neurophysiology. 1999 Apr 1;110(4):725-34.
- Kerr A. Introductory Biomechanics.
- SimiSystems. Gait analysis software - inverse kinematics. Available from: https://www.youtube.com/watch?v=704OYuJpKqg[last accessed 12/01/2021]
- Bravi M, Gallotta E, Morrone M, Maselli M, Santacaterina F, Toglia R, Foti C, Sterzi S, Bressi F, Miccinilli S. Concurrent validity and inter trial reliability of a single inertial measurement unit for spatial-temporal gait parameter analysis in patients with recent total hip or total knee arthroplasty. Gait & Posture. 2020 Feb 1;76:175-81.
- Kainz H, Modenese L, Lloyd DG, Maine S, Walsh HP, Carty CP. Joint kinematic calculation based on clinical direct kinematic versus inverse kinematic gait models. Journal of biomechanics. 2016 Jun 14;49(9):1658-69.
- Barn R, Rafferty D, Turner DE, Woodburn J. Reliability study of tibialis posterior and selected leg muscle EMG and multi-segment foot kinematics in rheumatoid arthritis associated pes planovalgus. Gait & posture. 2012 Jul 1;36(3):567-71.
- BERNARD, 2006. La marche de l’infirme moteur cérébral, enfant et adulte. Edition Springer.
- Wren TA, Rethlefsen SA, Healy BS, Do KP, Dennis SW, Kay RM. Reliability and validity of visual assessments of gait using a modified physician rating scale for crouch and foot contact. Journal of Pediatric Orthopaedics. 2005 Sep 1;25(5):646-50.
- Perry J. Gait analysis: technology and the clinician. Journal of rehabilitation research and development. 1994;31(1):vii.
- Baker R. Gait analysis methods in rehabilitation. Journal of neuroengineering and rehabilitation. 2006 Dec 1;3(1):4.