Paper Submission & Registration
8th Dutch Bio-Medical Engineering Conference
15:20   Muscle
Chair: Joris Jaspers
15 mins
Melt electrowriting for musculoskeletal tissue engineering
Piotr Zielinski, Marleen Kamperman, Malgorzata Wlodarczyk-Biegun
Abstract: Introduction: Musculoskeletal degeneration and complications from injuries have become more prevalent as people live longer and increasingly participate in rigorous athletic and recreational activities. Especially prone to damage are so-called hard-soft tissue interfaces, such as tendon-bone interfaces. These natural interfaces are hierarchical structures with gradually changing physicochemical and biological properties. Until now, most of the clinically available regenerative treatments are not very successful and it is still a major challenge to produce scaffolds that closely imitate the complexity of native tissue. Therefore, in this project I aim at developing polymeric fibrous scaffolds for musculoskeletal tissue engineering, with well controlled mechanical and biological properties. Specifically, I am investigating how the topology influences mechanical and biological performance of the scaffold. Research methods: Fibrous scaffolds with varying pore geometries, inter-fiber distance and fibers thickness were designed and produced from polycaprolactone using Melt Electrowriting (MEW) technique. In this biofabrication method, an electrical field is used to draw well-organized, micron-sized fibers of a molten polymer onto a computer-controlled collector plate, which leads to control of scaffold morphology at both macro and micro level. Mechanical properties of the scaffolds with different designs are investigated using tensile testing. In addition, computer simulation are used to predict the mechanical performance of the scaffolds. Results: Precise scaffolds with different architectures were successfully printed. Preliminary results indicated that the scaffolds stiffness and the ultimate strain are dependent on the design. First computer simulations revealed the possibility to predict the mechanical properties of printed scaffolds. Conclusions and Outlook: Mechanical properties of the scaffolds can be controlled by material topology. In the course of my PhD I will investigate this relation in more detail. In the next steps, I will also analyze cell behavior on the scaffolds with different designs and resulting stiffness ranges. In the future, gradient scaffolds will be produced. Significance: This study shows the applicability of MEW technique to obtain well-controlled scaffolds for use in regeneration of different musculoskeletal tissues. We envision that printed gradient scaffolds will closely mimic complexity of natural tissue interfaces.
15 mins
Towards model-based estimation of muscle history dependent properties and joint stiffness: From experimental data to validation
Christopher P. Cop, Alfred C. Schouten, Bart F.J.M. Koopman, Massimo Sartori
Abstract: Background: Quantifying human joint stiffness in vivo during dynamic movement remains challenging. We propose a data-driven biomechanical model-based approach that does not require joint perturbation, in contrast to well established system identification methods. Biomechanical models often use a standard Hill-type muscle model that does not capture history-dependent properties, e.g. stretch-induced force enhancement (FE). Moreover, muscle architecture features, e.g. pennation angle, are often neglected when estimating stiffness. Our goals are 1) to experimentally capture history-dependent muscle force generation mechanisms; 2) to augment the Hill type muscle model to include history-dependent mechanisms and muscle architecture features; and 3) validate our joint stiffness estimations against a system identification method during dynamic ankle rotations. Methods: A voluntary participant was instructed to keep the tibialis anterior muscle’s (TA) activation level between 15% and 20% of maximum voluntary contraction while the resulting ankle torque was recorded by a dynamometer. The subject performed a) isometric contractions and b) an active stretch followed by an isometric contraction. Additionally, the subject was instructed to follow a sinusoidal plantar-dorsi flexion angle target (amplitude: 0.15 rad, frequency: 0.6 Hz). Electromyography (EMG) and kinematic data were used to estimate joint stiffness using an EMG-driven musculoskeletal model that considers the muscle fiber’s pennation angle. Joint stiffness estimations were compared to an ensemble-based system identification method. Results: Preliminary results show that we were able to elicit FE on the TA (17.8% torque increase compared to the reference contraction). Our model’s joint stiffness estimations were comparable to the system identification reference values (R2= 0.56 and root mean squared error, RMSE,= 3.48 Nm/rad). Discussion: The inclusion of physiological features in musculoskeletal models seems to improve joint stiffness estimation. Next steps include the implementation of a FE model in the Hill-type muscle model to improve its capabilities of force and stiffness estimation during eccentric contractions. Relevance: The ability to decode joint stiffness from EMGs and joint angles at any instant of time, without needing to apply joint perturbations, might help to fill the gap of knowledge between the neural and the muscular systems and enable the subsequent development of tailored neurorehabilitation therapies and biomimetic prostheses and orthoses.
15 mins
Does pain reduction through local anaesthesia influence motor strategies in overhead athletes with ongoing pain? - A study protocol using OpenSim for kinematic analysis
Ton Leenen, Norman D'hondt, Michel van den Bekerom, Marco Hoozemans, Dirkjan Veeger
Abstract: Overhead athletes suffering from acute nociceptive pain due to a shoulder injury immediately execute motor strategies that protect the injured site from further harm or threat. Although these motor strategies yield short-term benefits, they may have detrimental long-term consequences when they undesirably persist. Ongoing nociceptive input is believed to play an eminent role in the failure to explore other motor strategies to regain pre-pain motor control. However, cause-effect mechanisms are still unclear. To determine the influence of nociceptive input on motor strategies in overhead athletes with persistent nociceptive shoulder pain, a multicenter trial involving four clinics will be conducted. Using a within-subject, repeated measures study-design, electromyography (EMG) and kinematics of the symptomatic upper limb and the intensity of perceived pain will be measured during repeated target-oriented active arm elevations with and without local anaesthesia. EMG of force coupled thoracohumeral, scapulohumeral, and scapulothoracic muscles will be recorded with wireless biosignalsplux sensors, to analyze the differences in muscle onsets and offsets between the pre- and post-anaesthesia condition. Kinematical data of the thorax, scapula and upper limb will be collected with wireless Xsens DOT sensors (inertial measurement units; IMU), adhering to ISB-guidelines. Unique to this study is the use of open-source modelling software OpenSim, enabling reconstruction of the performed active arm elevations through a biomechanical scapulothoracic model, driven by IMU sensor data. This model will be used to analyze between-condition differences in thoracohumeral, scapulohumeral and scapulothoracic motion. Scapulothoracic motion will be described by translation and rotation of the scapula on the thorax, modelled as an ellipsoid surface, providing the following four position coordinates: abduction-adduction, elevation-depression, upward-downward rotation, and internal-external rotation. Thoracohumeral motion will be described by the elevation angle, internal rotation and the displacement in elevation. When available, preliminary results will be presented at this conference.
15 mins
Developing an automated tool to detect fasciculations from ultrasound recordings
Diederik Stikvoort Garcia, Stephan Goedee, Boudewijn Sleutjes, Leonard van den Berg
Abstract: Background: The presence of fasciculations is one of the diagnostic criteria for amyotrophic lateral sclerosis (ALS). Fasciculations are conventionally scored with needle EMG. However, observer studies reported a higher sensitivity when using ultrasound (US) to grade fasciculations by their spatial and temporal nature. Quantifying fasciculation patterns from US requires automated protocols. Still, such protocols predominantly focus on producing temporal measures of fasciculations. Quantification of the spatial nature of fasciculations may further improve diagnostic sensitivity and monitoring in ALS. Objectives: To expand established automatic fasciculation detection protocols for US sequences with spatial metrics, in congruence with clinical practice. Methods: US recordings of 30 sec were obtained from 8 patients with suspected ALS recruited through our outpatient clinic from the abductor pollicis brevis (n = 8) and biceps brachii (n = 7). Each sequence received a score from an experienced US rater as a combination of a) focal, multifocal, diffuse (distribution), b) sporadic, intermittent, continuous (frequency). The algorithm uses adaptive Gaussian mixture model background subtraction and co-occurrence and spatial distribution driven spectral clustering. It yields the number of regions with motion, the number of movements per 30 sec and area as percentage of total image size. Metric values were discretized to thirds of the maximum metric value and compared to the clinician’s scores. Results: Fasciculations were detected in 13 sequences by the algorithm and 14 by the clinician. The mean time to run the algorithm was 42.9 sec (range 21.2-111.3 sec). Scores by the clinician were focal-sporadic (n = 6); multifocal-intermittent (n = 2), and diffuse-continuous (n = 6). The mean number of regions was 5.7 (range 0-13), the mean number of movements was 22 (range 0-91) and mean area was 16.9 (range 0-42 %). The number of regions corresponded most often with the clinical scores (9 out 14 classified correctly). Reduced algorithm performance compared to operator scoring mostly occurred in high contrast regions with probe motion or voluntary movement. Discussion: Using our initially implemented algorithm, we successfully quantified the distribution of fasciculations and their frequency of movement. Further development is required to more reliably detect fasciculation specific movement patterns. The use of the algorithm may eventually support the diagnostic phase by objective fasciculation scoring.

end %-->