Antagonist Activation Measurement in Triceps Surae Using High-Density and Bipolar Surface EMG in Chronic Hemiparesis.

After a stroke, antagonist muscle activation during agonist command impedes movement. This study compared measurements of antagonist muscle activation using surface bipolar EMG in the gastrocnemius medialis (GM) and high-density (HD) EMG in the GM and soleus (SO) during isometric submaximal and maximal dorsiflexion efforts, with knee flexed and extended, in 12 subjects with chronic hemiparesis. The coefficients of antagonist activation (CAN) of GM and SO were calculated according to the ratio of the RMS amplitude during dorsiflexion effort to the maximal agonist effort for the same muscle. Bipolar CAN (BipCAN) was compared to CAN from channel-specific (CsCAN) and overall (OvCAN) normalizations of HD-EMG. The location of the CAN centroid was explored in GM, and CAN was compared between the medial and lateral portions of SO. Between-EMG system differences in GM were observed in maximal efforts only, between BipCAN and CsCAN with lower values in BipCAN (p < 0.001), and between BipCAN and OvCAN with lower values in OvCAN (p < 0.05). The CAN centroid is located mid-height and medially in GM, while the CAN was similar in medial and lateral SO. In chronic hemiparesis, the estimates of GM hyperactivity differ between bipolar and HD-EMGs, with channel-specific and overall normalizations yielding, respectively, higher and lower CAN values than bipolar EMG. HD-EMG would be the way to develop personalized rehabilitation programs based on individual antagonist activations.

Myoelectric manifestations of fatigue of the finger flexor muscles and endurance capacity in experienced versus intermediate climbers during suspension tasks.

Climbing is a physically demanding discipline, placing significant loads on the finger flexors. Notwithstanding the documented greater endurance capacity of experienced climbers, the mechanisms explaining these training-induced adaptations remain unknown. We therefore investigate whether two non-competing strategies - muscle adaptation and alternate muscle recruitment - may explain the disparity in endurance capacity in participants with different climbing experience. We analysed high-density surface electromyograms (EMGs) from 38 Advanced and Intermediate climbers, during suspension exercises over three different depths (15, 20, 30 mm) using a half-crimp grip position. From the spatial distribution of changes in MeDian Frequency and Root Mean Square values until failure, we assessed how much and how diffusely the myoelectric manifestations of fatigue took place. Advanced climbers exhibited greater endurance, as evidenced by significantly longer failure time (p < 0.009) and lower changes in MDF values (p < 0.013) for the three grip depths. These changes were confined to a small skin region (nearly 25% of the grid size), centred at variable locations across participants. Moreover, lower MDF changes were significantly associated with longer suspension times. Collectively, our results suggest that muscle adaptation rather than load sharing between and within muscles is more likely to explain the improved endurance in experienced climbers.

Volitional muscle activation intensifies neuronal processing of proprioceptive afference in the primary sensorimotor cortex: an EEG study.

Proprioception refers to the ability to perceive the position and movement of body segments in space. The cortical aspects of the proprioceptive afference from the body can be investigated using corticokinematic coherence (CKC). CKC accurately quantifies the degree of coupling between cortical activity and limb kinematics, especially if precise proprioceptive stimulation of evoked movements is used. However, there is no evidence on how volitional muscle activation during proprioceptive stimulation affects CKC strength. Twenty-five healthy volunteers (28.8 ± 7 yr, 11 females) participated in the experiment, which included electroencephalographic (EEG), electromyographic (EMG), and kinematic recordings. Ankle-joint rotations (2-Hz) were elicited through a movement actuator in two conditions: passive condition with relaxed ankle and active condition with constant 5-Nm plantar flexion exerted during the stimulation. In total, 6 min of data were recorded per condition. CKC strength was defined as the maximum coherence value among all the EEG channels at the 2-Hz movement frequency for each condition separately. Both conditions resulted in significant CKC peaking at the Cz electrode over the foot area of the primary sensorimotor (SM1) cortex. Stronger CKC was found for the active (0.13 ± 0.14) than the passive (0.03 ± 0.04) condition (P < 0.01). The results indicated that volitional activation of the muscles intensifies the neuronal proprioceptive processing in the SM1 cortex. This finding could be explained both by peripheral sensitization of the ankle joint proprioceptors and central modulation of the neuronal proprioceptive processing at the spinal and cortical levels.NEW & NOTEWORTHY The current study is the first to investigate the effect of volitional muscle activation on CKC-based assessment of cortical proprioception of the ankle joint. Results show that the motor efference intensifies the neuronal processing of proprioceptive afference of the ankle joint. This is a significant finding as it may extend the use of CKC method during active tasks to further evaluate the motor efference-proprioceptive afference relationship and the related adaptations to exercise, rehabilitation, and disease.

Developing RPC-Net: Leveraging High-Density Electromyography and Machine Learning for Improved Hand Position Estimation.

OBJECTIVE: The purpose of this study was to develop and evaluate the performance of RPC-Net (Recursive Prosthetic Control Network), a novel method using simple neural network architectures to translate electromyographic activity into hand position with high accuracy and computational efficiency. METHODS: RPC-Net uses a regression-based approach to convert forearm electromyographic signals into hand kinematics. We tested the adaptability of the algorithm to different conditions and compared its performance with that of solutions from the academic literature. RESULTS: RPC-Net demonstrated a high degree of accuracy in predicting hand position from electromyographic activity, outperforming other solutions with the same computational cost. Including previous position data consistently improved results across subjects and conditions. RPC-Net showed robustness against a reduction in the number of electromyography electrodes used and shorter input signals, indicating potential for further reduction in computational cost. CONCLUSION: The results demonstrate that RPC-Net is capable of accurately translating forearm electromyographic activity into hand position, offering a practical and adaptable tool that may be accessible in clinical settings. SIGNIFICANCE: The development of RPC-Net represents a significant advancement. In clinical settings, its application could enable prosthetic devices to be controlled in a way that feels more natural, improving the quality of life for individuals with limb loss.

The Sensitivity of Bipolar Electromyograms to Muscle Excitation Scales With the Inter-Electrode Distance.

The value of surface electromyograms (EMGs) lies in their potential to non-invasively probe the neuromuscular system. Whether muscle excitation may be accurately inferred from bipolar EMGs depends on how much the detected signal is both sensitive and specific to the excitation of the target muscle. While both are known to be a function of the inter-electrode distance (IED), specificity has been of long concern in the physiological literature. In contrast, sensitivity, at best, has been implicitly assumed. Here we provide evidence that the IED imposes a biophysical constraint on the sensitivity of surface EMG. From 20 healthy subjects, we tested the hypothesis that excessively reducing the IED limits EMGs' physiological content. We detected bipolar EMGs with IEDs varying from 5 mm to 50 mm from two skeletal muscles with distinct architectures, gastrocnemius and biceps brachii. Non-parametric statistics and Bayesian hierarchical modelling were used to evaluate the dependence of the onset of muscle excitation and signal-to-noise ratio (SNR) on the IED. Experimental results revealed that IED critically affects the sensitivity of bipolar EMGs for both muscles-indeliberately reducing the IED yields EMGs that are not representative of the whole muscle, hampering validity. Simulation results substantiate the generalization of experimental results to small and large electrodes. Based on current and previous findings, we discuss a potentially valid procedure for defining the most appropriate IED for a single bipolar, surface recording-i.e., the distance from the electrode to the target muscle boundary may heuristically serve as a lower bound when choosing an IED.

Modulations in motor unit discharge are related to changes in fascicle length during isometric contractions.

The integration of electromyography (EMG) and ultrasound imaging has provided important information about the mechanisms of muscle activation and contraction. Unfortunately, conventional bipolar EMG does not allow an accurate assessment of the interplay between the neural drive received by muscles, changes in fascicle length and torque. We aimed to assess the relationship between modulations in tibialis anterior muscle (TA) motor unit (MU) discharge, fascicle length, and dorsiflexion torque using ultrasound-transparent high-density EMG electrodes. EMG and ultrasound images were recorded simultaneously from TA using a 32-electrode silicon matrix while performing isometric dorsiflexion contractions at two ankle joint positions (0° or 30° plantar flexion) and torques (20% or 40% of maximum). EMG signals were decomposed into MUs and changes in fascicle length were assessed with a fascicle-tracking algorithm. MU firings were converted into a cumulative spike train (CST) that was cross-correlated with torque (CST-torque) and fascicle length (CST-length). High cross-correlations were found for CST-length (0.60, range: 0.31-0.85) and CST-torque (0.71, range: 0.31-0.88). Cross-correlation delays revealed that the delay between CST-fascicle length (∼75 ms) was smaller than CST-torque (∼150 ms, P < 0.001). These delays affected MU recruitment and de-recruitment thresholds since the fascicle length at which MUs were recruited and de-recruited was similar but MU recruitment-de-recruitment torque varied. This study demonstrates that changes in TA fascicle length are related to modulations in MU firing and dorsiflexion torque. These relationships allow assessment of the interplay between neural drive, muscle contraction and torque, enabling the time required to convert neural activity into movement to be quantified.NEW & NOTEWORTHY By employing ultrasound-transparent high-density EMG electrodes, we show that modulations in tibialis anterior muscle motor unit discharge rate were related to both changes in fascicle length and resultant torque. These relationships permitted the quantification of the relative delays between fluctuations in neural drive, muscle contraction, and transfer of torque via the tendon during sustained isometric dorsiflexion contractions, providing information on the conversion of neural activity into muscle force during a contraction.

Are the forearm muscles excited equally in different, professional piano players?

BACKGROUND AND OBJECTIVES: Professional pianists tend to develop playing-related musculoskeletal disorders mostly in the forearm. These injuries are often due to overuse, suggesting the existence of a common forearm region where muscles are often excited during piano playing across subjects. Here we use a grid of electrodes to test this hypothesis, assessing where EMGs with greatest amplitude are more likely to be detected when expert pianists perform different excerpts. METHODS: Tasks were separated into two groups: classical excerpts and octaves, performed by eight, healthy, professional pianists. Monopolar electromyograms (EMGs) were sampled with a grid of 96 electrodes, covering the forearm region where hand and wrist muscles reside. Regions providing consistently high EMG amplitude across subjects were assessed with a non-parametric permutation test, designed for the statistical analysis of neuroimaging experiments. Spatial consistency across trials was assessed with the Binomial test. RESULTS: Spatial consistency of muscle excitation was found across subjects but not across tasks, confining at most 20% of the electrodes in the grid. These local groups of electrodes providing high EMG amplitude were found at the ventral forearm region during classical excerpts and at the dorsal region during octaves, when performed both at preferred and at high, playing speeds. DISCUSSION: Our results revealed that professional pianists consistently load a specific forearm region, depending on whether performing octaves or classical excerpts. This spatial consistency may help furthering our understanding on the incidence of playing-related muscular disorders and provide an anatomical reference for the study of active muscle loading in piano players using surface EMG.

Long short-term memory (LSTM) recurrent neural network for muscle activity detection.

BACKGROUND: The accurate temporal analysis of muscle activation is of great interest in many research areas, spanning from neurorobotic systems to the assessment of altered locomotion patterns in orthopedic and neurological patients and the monitoring of their motor rehabilitation. The performance of the existing muscle activity detectors is strongly affected by both the SNR of the surface electromyography (sEMG) signals and the set of features used to detect the activation intervals. This work aims at introducing and validating a powerful approach to detect muscle activation intervals from sEMG signals, based on long short-term memory (LSTM) recurrent neural networks. METHODS: First, the applicability of the proposed LSTM-based muscle activity detector (LSTM-MAD) is studied through simulated sEMG signals, comparing the LSTM-MAD performance against other two widely used approaches, i.e., the standard approach based on Teager-Kaiser Energy Operator (TKEO) and the traditional approach, used in clinical gait analysis, based on a double-threshold statistical detector (Stat). Second, the effect of the Signal-to-Noise Ratio (SNR) on the performance of the LSTM-MAD is assessed considering simulated signals with nine different SNR values. Finally, the newly introduced approach is validated on real sEMG signals, acquired during both physiological and pathological gait. Electromyography recordings from a total of 20 subjects (8 healthy individuals, 6 orthopedic patients, and 6 neurological patients) were included in the analysis. RESULTS: The proposed algorithm overcomes the main limitations of the other tested approaches and it works directly on sEMG signals, without the need for background-noise and SNR estimation (as in Stat). Results demonstrate that LSTM-MAD outperforms the other approaches, revealing higher values of F1-score (F1-score > 0.91) and Jaccard similarity index (Jaccard > 0.85), and lower values of onset/offset bias (average absolute bias < 6 ms), both on simulated and real sEMG signals. Moreover, the advantages of using the LSTM-MAD algorithm are particularly evident for signals featuring a low to medium SNR. CONCLUSIONS: The presented approach LSTM-MAD revealed excellent performances against TKEO and Stat. The validation carried out both on simulated and real signals, considering normal as well as pathological motor function during locomotion, demonstrated that it can be considered a powerful tool in the accurate and effective recognition/distinction of muscle activity from background noise in sEMG signals.

Dynamic Surface Electromyography Using Stretchable Screen-Printed Textile Electrodes.

OBJECTIVE: Wearable devices have created new opportunities in healthcare and sport sciences by unobtrusively monitoring physiological signals. Textile polymer-based electrodes proved to be effective in detecting electrophysiological potentials but suffer mechanical fragility and low stretch resistance. The goal of this research is to develop and validate in dynamic conditions cost-effective and easily manufacturable electrodes characterized by adequate robustness and signal quality. METHODS: We here propose an optimized screen printing technique for the fabrication of PEDOT:PSS-based textile electrodes directly into finished stretchable garments for surface electromyography (sEMG) applications. A sensorised stretchable leg sleeve was developed, targeting five muscles of interest in rehabilitation and sport science. An experimental validation was performed to assess the accuracy of signal detection during dynamic exercises, including sit-to-stand, leg extension, calf raise, walking, and cycling. RESULTS: The electrodes can resist up to 500 stretch cycles. Tests on five subjects revealed excellent contact impedance, and cross-correlation between sEMG envelopes simultaneously detected from the leg muscles by the textile and Ag/AgCl electrodes was generally greater than 0.9, which proves that it is possible to obtain good quality signals with performance comparable with disposable electrodes. CONCLUSIONS: An effective technique to embed polymer-based electrodes in stretchable smart garments was presented, revealing good performance for dynamic sEMG detections. SIGNIFICANCE: The achieved results pave the way to the integration of unobtrusive electrodes, obtained by screen printing of conductive polymers, into technical fabrics for rehabilitation and sport monitoring, and in general where the detection of sEMG in dynamic conditions is necessary.

Design and Characterization of a Textile Electrode System for the Detection of High-Density sEMG.

Muscle activity monitoring in dynamic conditions is a crucial need in different scenarios, ranging from sport to rehabilitation science and applied physiology. The acquisition of surface electromyographic (sEMG) signals by means of grids of electrodes (High-Density sEMG, HD-sEMG) allows obtaining relevant information on muscle function and recruitment strategies. During dynamic conditions, this possibility demands both a wearable and miniaturized acquisition system and a system of electrodes easy to wear, assuring a stable electrode-skin interface. While recent advancements have been made on the former issue, detection systems specifically designed for dynamic conditions are at best incipient. The aim of this work is to design, characterize, and test a wearable, HD-sEMG detection system based on textile technology. A 32-electrodes, 15 mm inter-electrode distance textile grid was designed and prototyped. The electrical properties of the material constituting the detection system and of the electrode-skin interface were characterized. The quality of sEMG signals was assessed in both static and dynamic contractions. The performance of the textile detection system was comparable to that of conventional systems in terms of stability of the traces, properties of the electrode-skin interface and quality of the collected sEMG signals during quasi-isometric and highly dynamic tasks.

Investigating the effect of a passive trunk exoskeleton on local discomfort, perceived effort and spatial distribution of back muscles activity.

This study aimed at determining the effect of a passive exoskeleton on local perceived discomfort, perceived effort and low back muscles' activity. Thirteen volunteers performed two simulated working tasks with and without the exoskeleton. In the static task, the exoskeleton decreased the lumbar perceived discomfort, the perceived effort and the level of low back muscles' activity (∼10%) while increasing discomfort in the chest and feet. The percent decrease in EMG amplitude was correlated with the percent increase in perceived effort with exoskeleton. For the dynamic task, the exoskeleton increased the discomfort in the chest and decreased the level of back muscle activity (∼5%). Current findings suggest exoskeleton is effective in reducing the back load while increasing the perceived discomfort at non-targeted body regions in both working tasks. The concurrent increase of discomfort in non-targeted areas probably led to a higher perceived effort despite the reduction of low back muscle activity. Practitioner summary: This study provided insights into exoskeleton effects on local discomfort, perceived effort and muscle activity. Overall, the potential benefits of passive exoskeleton should be considered alongside its adverse effects on the non-targeted body regions that can lead to an increase of perceived effort despite the reduction of back muscle activity.

Understanding regional activation of thoraco-lumbar muscles in chronic low back pain and its relationship to clinically relevant domains.

BACKGROUND: Altered regional activation of the lumbar extensors has been previously observed in individuals with low back pain (LBP) performing high-effort and fatiguing tasks. It is currently unknown whether similar alterations can be observed during low-effort functional tasks. Similarly, previous studies did not investigate whether side differences in regional activation are present in individuals with LBP. Finally, there is limited evidence of whether the extent of the alteration of regional activation is associated with clinical factors. Therefore, the aim of this study was to investigate whether individuals with LBP exhibit asymmetric regional activation of the thoraco-lumbar extensor muscles during functional tasks, and if the extent of neuromuscular control alteration is associated with clinical and psychosocial outcome domains. METHODS: 21 participants with and 21 without LBP performed five functional tasks (gait, sit-to-stand, forward trunk flexion, shoulder flexion and anterior pelvic tilt). The spatial distribution of activation of the thoraco-lumbar extensor muscles was assessed bilaterally using high-density electromyography. For each side, the distribution of electromyographic (EMG) amplitude was characterized in terms of intensity, location and size. Indices of asymmetry were calculated from these features and comparisons between groups and tasks were performed using ANOVA. The features that significantly differed between groups were correlated with self-reported measures of pain intensity and other outcome domains. RESULTS: Indices of asymmetry did not differ between participants with and without LBP (p > 0.11). The cranio-caudal location of the activation differed between tasks (p < 0.05), but not between groups (p = 0.64). Participants with LBP showed reduced EMG amplitude during anterior pelvic tilt and loading response phase during gait (both p < 0.05). Pearson correlation revealed that greater pain intensity was associated with lower EMG amplitude for both tasks (R<-0.5, p < 0.05). CONCLUSIONS: Despite clear differences between tasks, individuals with and without LBP exhibited similar distributions of EMG amplitude during low-effort functional activities, both within and between sides. However, individuals with LBP demonstrated lower activation of the thoraco-lumbar muscles during gait and anterior pelvic tilt, especially those reporting higher pain intensity. These results have implications in the development or refinement of assessment and intervention strategies focusing on motor control in patients with chronic LBP.

Characterization of the stimulation output of four devices for focal muscle vibration.

Different devices for mechano-acoustic muscle vibration became available on the market in the last ten years. Although the use of these vibrators is increasing in research and clinical settings, the features of their stimulation output were never described in literature. In this study we aimed to quantify and compare the stimulation output of the four most widespread pneumatic devices for focal muscle vibration available on the market. A piezoelectric pressure sensor was used to measure the pressure profile generated by the four selected devices in the following experimental conditions: i) measurement of the output changes associated with variations of the stimulation amplitude for three stimulation frequencies (100 Hz, 200 Hz, and 300 Hz); ii) measurement of the output changes during a 20-min long stimulation at constant frequency (300 Hz) and amplitude; iii) measurement of the output changes associated with the progressive activation of all stimulation channels at constant frequency (200 Hz) for different amplitudes. The maximum peak-to-peak amplitudes of the pressure waves were in the range 102 mbar - 369 mbar (below the maximum values declared by the different manufacturers). The shape of the pressure waves generated by the four devices was quasi-sinusoidal and asymmetric with respect to the atmospheric pressure. All output features had a remarkable intra- and inter-device variability. Further studies are required to support the technological improvement of the currently available devices and to focus the issues of vibration effectiveness, limitations, proper protocols, modalities of its application and assessment in neuromuscular training and rehabilitation.

Timing and Modulation of Activity in the Lower Limb Muscles During Indoor Rowing: What Are the Key Muscles to Target in FES-Rowing Protocols?

The transcutaneous stimulation of lower limb muscles during indoor rowing (FES Rowing) has led to a new sport and recreation and significantly increased health benefits in paraplegia. Stimulation is often delivered to quadriceps and hamstrings; this muscle selection seems based on intuition and not biomechanics and is likely suboptimal. Here, we sample surface EMGs from 20 elite rowers to assess which, when, and how muscles are activated during indoor rowing. From EMG amplitude we specifically quantified the onset of activation and silencing, the duration of activity and how similarly soleus, gastrocnemius medialis, tibialis anterior, rectus femoris, vastus lateralis and medialis, semitendinosus, and biceps femoris muscles were activated between limbs. Current results revealed that the eight muscles tested were recruited during rowing, at different instants and for different durations. Rectus and biceps femoris were respectively active for the longest and briefest periods. Tibialis anterior was the only muscle recruited within the recovery phase. No side differences in the timing of muscle activity were observed. Regression analysis further revealed similar, bilateral modulation of activity. The relevance of these results in determining which muscles to target during FES Rowing is discussed. Here, we suggest a new strategy based on the stimulation of vasti and soleus during drive and of tibialis anterior during recovery.

A Modular, Smart, and Wearable System for High Density sEMG Detection.

OBJECTIVE: The use of linear or bi-dimensional electrode arrays for surface EMG detection (HD-sEMG) is gaining attention as it increases the amount and reliability of information extracted from the surface EMG. However, the complexity of the setup and the encumbrance of HD-sEMG hardware currently limits its use in dynamic conditions. The aim of this paper was to develop a miniaturized, wireless, and modular HD-sEMG acquisition system for applications requiring high portability and robustness to movement artifacts. METHODS: A system with modular architecture was designed. Its core is a miniaturized 32-channel amplifier (Sensor Unit - SU) sampling at 2048 sps/ch with 16 bit resolution and wirelessly transmitting data to a PC or a mobile device. Each SU is a node of a Body Sensor Network for the synchronous signal acquisition from different muscles. RESULTS: A prototype with two SUs was developed and tested. Each SU is small (3.4 cm × 3 cm × 1.5 cm), light (16.7 g), and can be connected directly to the electrodes; thus, avoiding the need for customary, wired setup. It allows to detect HD-sEMG signals with an average noise of 1.8 µVRMS and high performance in terms of rejection of power-line interference and motion artefacts. Tests performed on two SUs showed no data loss in a 22 m range and a ±500 µs maximum synchronization delay. CONCLUSIONS: Data collected in a wide spectrum of experimental conditions confirmed the functionality of the designed architecture and the quality of the acquired signals. SIGNIFICANCE: By simplifying the experimental setup, reducing the hardware encumbrance, and improving signal quality during dynamic contractions, the developed system opens new perspectives in the use of HD-sEMG in applied and clinical settings.

Design of a Wireless, Modular and Programmable Neuromuscular Electrical Stimulator.

The use of electrical stimulation to elicit single twitches and tetanic contractions of skeletal muscles has increased markedly in the last years, with applications ranging from basic physiology to clinical settings. Addressing all possible needs required by different applications with an electrical stimulator is challenging as it requires the device to be highly flexible in terms of stimulation configurations (number of channels and electrode location), and possibility to control the stimulation patterns (timing and stimulation profiles). This paper describes a new wireless, modular, and programmable electrical stimulator integrating the possibility to acquire and use biomechanical signals to trigger the stimulation output. A closed-loop FES Cycling setup has been presented to show a possible application of the system.

Towards the Design of an Impedance-Controlled HD-sEMG Amplifier: A Feasibility Study.

The use of multiple surface EMG electrodes (High-Density surface EMG - HD-sEMG) allows the extraction of anatomical and physiological information either at the muscle or at the motor unit level with applications in several fields ranging from clinical neurophysiology to the control of prosthetic devices. These applications need to acquire monopolar sEMG signals free from power line interference arising from the capacitive coupling between the subject, the acquisition system and the power line. The aim of this work is to provide a common mode analysis of the detection system used to collect monopolar sEMG signals, characterizing different configuration of the reference electrodes leading to different behaviors in terms of immunity to the power line interference. Based on the experimental results, a new impedance-controlled HD-sEMG signal amplifier is proposed and discussed.

Design and Test of a Biomechanical Model for the Estimation of Knee Joint Angle During Indoor Rowing: Implications for FES-Rowing Protocols in Paraplegia.

Functional electrical stimulation of lower limb muscles during rowing provides a means for the cardiovascular conditioning in paraplegia. The possibility of shaping stimulation profiles according to changes in knee angle, so far conceived as changes in seat position, may help circumventing open issues associated with muscle fatigue and movement coordination. Here, we present a subject-specific biomechanical model for the estimation of knee joint angle during indoor rowing. Anthropometric measurements and foot and seat positions are inputs to the model. We tested our model on two samples of elite rowers; 15 able-bodied, and 11 participants in the Rio 2016 Paralympic games. Paralympic rowers presented minor physical disabilities (LTA-PD classification), enabling them to perform the full rowing cycle (with legs, trunks, and arms). Knee angle was estimated from the rowing machine seat position, measured with a linear encoder, and transmitted wirelessly to a computer. Key results indicate the root mean square error (RMSE) between estimated and measured angles did not depend on group and stroke rate ( ). Significantly greater RMSE values were observed, however, within the rowing cycle ( ), reaching on average 8 deg in the mid-recovery phase. Differences between estimated and measured knee angle values resulted in slightly earlier (5%) detection of knee flexion, regardless of the group and stroke rate considered. Offset of knee extension, knee angle at catch and range of knee motion were identified equally well with our model and with inertial sensors. These results suggest our model describes accurately the movement of knee joint during indoor rowing.