Search

Background: Brain hypoperfusion is linked with worse physical, cognitive and MRI outcomes in multiple sclerosis (MS). Understanding the proteomic signatures related to hypoperfusion could provide insights into the pathophysiological mechanism., Methods: 140 people with MS (pwMS; 86 clinically isolated syndrome (CIS)/relapsing-remitting (RRMS) and 54 progressive (PMS)) were included. Cerebral arterial blood flow (CABF) was determined using ultrasound Doppler measurement as the sum of blood flow in the bilateral common carotid arteries and vertebral arteries. Proteomic analysis was performed using the Multiple Sclerosis Disease Activity (MSDA) test assay panel performed on the Olink™ platform. The MSDA test measures the concentrations of 18 proteins that are age and sex-adjusted. It utilizes a stacked classifier logistic regression model to determine 4 disease pathway scores (immunomodulation, neuroinflammation, myelin biology, and neuroaxonal integrity) as well as an overall disease activity score (1 to 10). MRI measures of T2 lesion volume (LV) and whole brain volume (WBV) were derived., Results: The pwMS were on average 54 years old and had an average CABF of 951 mL/min. There were no differences in CABF between CIS/RRMS vs. PMS groups. Lower CABF levels were correlated with the overall disease activity score (r = -0.26, p = 0.003) and with the neuroinflammation (r = -0.29, p = 0.001), immunomodulation (r = -0.26, p = 0.003) and neuroaxonal integrity (r = -0.23, p = 0.007) pathway scores. After age and body mass index (BMI)-adjustment, lower CABF remained associated with the neuroinflammatory (r = -0.23, p = 0.011) and immunomodulation (r = -0.20, p = 0.024) pathway scores. The relationship between CABF and the neuroinflammation pathway score remained significant after adjusting for T2-LV and WBV (p = 0.038). Individual analyses identified neurofilament light chain, CCL-20 and TNFSF13B as contributors. When compared to the highest quartile (>1133.5 mL/min), the pwMS in the lowest CABF quartile (<764 mL/min) had greater overall disease activity score (p = 0.003), neuroinflammation (p = 0.001), immunomodulation (p = 0.004) and neuroaxonal integrity pathway scores (p = 0.007)., Conclusion: Lower cerebral arterial perfusion in MS is associated with changes in neuroinflammatory/immunomodulation pathways and their respective proteomic biomarkers. These findings may suggest a relationship between the hypoperfusion and pro-inflammatory MS changes rather than being merely an epiphenomenon subsequent to lower energy demands., Competing Interests: Declaration of competing interest Dejan Jakimovski, Karen Marr and Niels Bergsland have nothing to disclose. Ferhan Qureshi, Anisha Keshavan, Ati Ghoreyshi, Kian Jalaleddini and Kelly Leyden are employees of Octave Bioscience. Murali Ramanathan received research funding from the Department of Defense and National Institute of Neurological Diseases and Stroke. Michael G. Dwyer received compensation from Keystone Heart for consultant fees. He received financial support for research activities from Bristol Myers Squibb, Mapi Pharma, Keystone Heart, Protembis and V-WAVE Medical. Bianca Weinstock-Guttman received honoraria for serving in advisory boards and educational programs from Biogen Idec, Novartis, Genentech, Genzyme and Sanofi, Janssen, Abbvie and Bayer. She also received support for research activities from the National Institutes of Health, National Multiple Sclerosis Society, Department of Defense, and Biogen Idec, Novartis, Genentech, Genzyme and Sanofi. Robert Zivadinov has received personal compensation from Bristol Myers Squibb, EMD Serono, Sanofi, Keystone Heart, Protembis and Novartis for speaking and consultant fees. He received financial support for research activities from Sanofi, Novartis, Bristol Myers Squibb, Octave, Mapi Pharma, Keystone Heart, Protembis and V-WAVE Medical., (Copyright © 2024 Elsevier B.V. All rights reserved.)

Background: Atrophied lesion volume (aLV), a proposed biomarker of disability progression in multiple sclerosis (MS) and transition into progressive MS (PMS), depicts chronic periventricular white matter (WM) pathology. Meningeal infiltrates, imaged as leptomeningeal contrast enhancement (LMCE), are linked with greater cortical pathology., Objectives: To determine the relationship between serum-derived proteomic data with the development of aLV and LMCE in a heterogeneous group of people with MS (pwMS)., Methods: Proteomic and MRI data for 202 pwMS (148 clinically isolated syndrome /relapsing-remitting MS and 54 progressive MS (PMS)) were acquired at baseline and at 5.4-year follow-up. The concentrations of 21 proteins related to multiple MS pathophysiology pathways were derived using a custom-developed Proximity Extension Assay on the Olink™ platform. The accrual of aLV was determined as the volume of baseline T2-weighted lesions that were replaced by cerebrospinal fluid over the follow-up. Regression models and age-adjusted analysis of covariance (ANCOVA) were used., Results: Older age (standardized beta = 0.176, p = 0.022), higher glial fibrillary acidic protein (standardized beta = 0.312, p = 0.001), and lower myelin oligodendrocyte glycoprotein levels (standardized beta = -0.271, p = 0.002) were associated with accrual of aLV over follow-up. This relationship was driven by the pwPMS population. The presence of LMCE at the follow-up visit was not predicted by any baseline proteomic biomarker nor cross-sectionally associated with any protein concentration., Conclusion: Proteomic markers of glial activation are associated with chronic lesional WM pathology (measured as aLV) and may be specific to the progressive MS phenotype. LMCE presence in MS does not appear to relate to proteomic measures., (Copyright © 2024 Elsevier B.V. All rights reserved.)

Background: A quantitative measurement of serum proteome biomarkers that would associate with disease progression endpoints can provide risk stratification for persons with multiple sclerosis (PwMS) and supplement the clinical decision-making process., Materials and Methods: In total, 202 PwMS were enrolled in a longitudinal study with measurements at two time points with an average follow-up time of 5.4 years. Clinical measures included the Expanded Disability Status Scale, Timed 25-foot Walk, 9-Hole Peg, and Symbol Digit Modalities Tests. Subjects underwent magnetic resonance imaging to determine the volumetric measures of the whole brain, gray matter, deep gray matter, and lateral ventricles. Serum samples were analyzed using a custom immunoassay panel on the Olink™ platform, and concentrations of 18 protein biomarkers were measured. Linear mixed-effects models and adjustment for multiple comparisons were performed., Results: Subjects had a significant 55.6% increase in chemokine ligand 20 (9.7 pg/mL vs. 15.1 pg/mL, p < 0.001) and neurofilament light polypeptide (10.5 pg/mL vs. 11.5 pg/mL, p = 0.003) at the follow-up time point. Additional changes in CUB domain-containing protein 1, Contactin 2, Glial fibrillary acidic protein, Myelin oligodendrocyte glycoprotein, and Osteopontin were noted but did not survive multiple comparison correction. Worse clinical performance in the 9-HPT was associated with neurofilament light polypeptide (p = 0.001). Increases in several biomarker candidates were correlated with greater neurodegenerative changes as measured by different brain volumes., Conclusion: Multiple proteins, selected from a disease activity test that represent diverse biological pathways, are associated with physical, cognitive, and radiographic outcomes. Future studies should determine the utility of multiple protein assays in routine clinical care., (© 2024 The Authors. Annals of Clinical and Translational Neurology published by Wiley Periodicals LLC on behalf of American Neurological Association.)

Objectives: We evaluate the potential clinical and cost impacts of discontinuing disease-modifying therapy (DMT) in people with multiple sclerosis (PwMS) when age-related immunosenescence can reduce DMT efficacy while increasing associated risks., Methods: A Markov model simulated clinical and cost impacts to the patient and payers when a proportion of eligible patients with relapsing remitting multiple sclerosis (RRMS) discontinue DMT. Eligibility was defined as age >55 years, an RRMS diagnosis of >5 years, and no history of relapses for 5 years. Increasing the proportion of eligible patients willing to discontinue therapy was also modeled. Clinical and cost inputs were from published literature., Results: Difference in EDSS progression between eligible patients who did and did not attempt discontinuation was not significant. After 1 year of eligibility, per-patient costs were $96k lower in the cohort that attempted discontinuation; however a higher proportion of relapses were seen in this group. When the proportion of patients willing to discontinue DMT increased, clinical findings remained consistent while the average cost per patient decreased., Conclusion: While there are increased clinical and cost benefits as more eligible patients attempt discontinuation, the risk of relapses can increase. Timely disease monitoring is required to manage safe DMT discontinuation., Competing Interests: Declaration of conflicting interestsThe author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

An 18-protein multiple sclerosis (MS) disease activity (DA) test was validated based on associations between algorithm scores and clinical/radiographic assessments (N = 614 serum samples; Train [n = 426; algorithm development] and Test [n = 188; evaluation] subsets). The multi-protein model was trained based on presence/absence of gadolinium-positive (Gd+) lesions and was also strongly associated with new/enlarging T2 lesions, and active versus stable disease (composite of radiographic and clinical evidence of DA) with improved performance (p < 0.05) compared to the neurofilament light single protein model. The odds of having ≥1 Gd+ lesions with a moderate/high DA score were 4.49 times that of a low DA score, and the odds of having ≥2 Gd+ lesions with a high DA score were 20.99 times that of a low/moderate DA score. The MSDA Test was clinically validated with improved performance compared to the top-performing single-protein model and can serve as a quantitative tool to enhance the care of MS patients., Competing Interests: Declaration of Competing Interest Tanuja Chitnis has received compensation for consulting from Biogen, Novartis Pharmaceuticals, Roche Genentech, and Sanofi Genzyme, and has received research support from the National Institutes of Health, National MS Society, US Department of Defense, EMD Serono, I-Mab Biopharma, Mallinckrodt ARD, Novartis Pharmaceuticals, Octave Bioscience, Inc., Roche Genentech, and Tiziana Life Sciences. This research was conducted in part with the support of the Department of Defense through the Multiple Sclerosis Research Program under Award No. W81XWH-18-1-0648 (to T. Chitnis). John Foley has received research support from Biogen, Novartis, Adamas, Octave Bioscience, Inc., Genentech, and Mallinckrodt, has received speakers' honoraria and acted as a consultant for EMD Serono, Genzyme, Novartis, Biogen, and Genentech, has equity interest in Octave Bioscience, Inc., and is the founder of InterPro Bioscience. Carolina Ionete has received research support from Biogen, Serono, Genentech, NMSS, and the Department of Defense, and received compensation for advisory board activity from Sanofi-Genzyme. Nabil K. El Ayoubi has received support to attend scientific educational courses from Novartis, Merck Serono, Sanofi, Biologix, and has received speaker honoraria for scientific presentations on Multiple Sclerosis from Biologix, Sanofi, Merck Serono, and Novartis. Shrishti Saxena, Patricia Gaitan-Walsh, Anu Paul, and Fermisk Saleh have no disclosures. Hrishikesh Lokhande has received research support from the US Department of Defense and Octave Bioscience, Inc. Howard Weiner has received research support from the Department of Defense, Genentech, Inc., National Institutes of Health, National Multiple Sclerosis Society, Novartis, and Sanofi Genzyme. He has received compensation for consulting from Genentech, Inc., IM Therapeutics, IMAB Biopharma, MedDay Pharmaceuticals, Tiziana Life Sciences, and vTv Therapeutics. Ferhan Qureshi, Anisha Keshavan, Kian Jalaleddini, and Ati Ghoreyshi are employees of Octave Bioscience, Inc. Michael J. Becich, Fatima Rubio da Costa, Victor M. Gehman, and Fujun Zhang were employees of Octave Bioscience, Inc., at the time the study was completed. Samia J. Khoury has received compensation for scientific advisory board activity from Merck and Roche and for serving on IDMC for Biogen., (Copyright © 2023 The Authors. Published by Elsevier Inc. All rights reserved.)

Acute ischemic stroke is a source of long-term disability in the United States, of which a large portion of cases are a result of large vessel occlusion (LVO). LVO strokes have high rates of morbidity and mortality due to difficulty of treatments in achieving recanalization. Recently, however, results of randomized clinical trials have shown that treatment options are expanding in both availability and efficacy. As these methods of intervention become more optimal, so must the preceding methods of assessment. Transcranial Doppler (TCD) ultrasound is a non-invasive method of evaluating cerebral hemodynamics, and has a long history in stroke assessment. Despite the importance of information provided by a TCD exam, its utilization in the acute stroke workflow has remained low because of its dependence on expert analysis. Here, we review the evolution of morphological analysis of TCD waveforms for the indication, localization, and monitoring of acute LVO., Competing Interests: Declaration of Competing Interest The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper., (Copyright © 2020 The Authors. Published by Elsevier Ltd.. All rights reserved.)

Measurement of intracranial pressure (ICP) is crucial in the management of many neurological conditions. However, due to the invasiveness, high cost, and required expertise of available ICP monitoring techniques, many patients who could benefit from ICP monitoring do not receive it. As a result, there has been a substantial effort to explore and develop novel noninvasive ICP monitoring techniques to improve the overall clinical care of patients who may be suffering from ICP disorders. This review attempts to summarize the general pathophysiology of ICP, discuss the importance and current state of ICP monitoring, and describe the many methods that have been proposed for noninvasive ICP monitoring. These noninvasive methods can be broken down into four major categories: fluid dynamic, otic, ophthalmic, and electrophysiologic. Each category is discussed in detail along with its associated techniques and their advantages, disadvantages, and reported accuracy. A particular emphasis in this review will be dedicated to methods based on the use of transcranial Doppler ultrasound. At present, it appears that the available noninvasive methods are either not sufficiently accurate, reliable, or robust enough for widespread clinical adoption or require additional independent validation. However, several methods appear promising and through additional study and clinical validation, could eventually make their way into clinical practice.

Objective: Transcranial Doppler (TCD) ultrasonography measures pulsatile cerebral blood flow velocity in the arteries and veins of the head and neck. Similar to other real-time measurement modalities, especially in healthcare, the identification of high-quality signals is essential for clinical interpretation. Our goal is to identify poor quality beats and remove them prior to further analysis of the TCD signal., Methods: We selected objective features for this purpose including Euclidean distance between individual and average beat waveforms, cross-correlation between individual and average beat waveforms, ratio of the high-frequency power to the total beat power, beat length, and variance of the diastolic portion of the beat waveform. We developed an iterative outlier detection algorithm to identify and remove the beats that are different from others in a recording. Finally, we tested the algorithm on a dataset consisting of more than 15 h of TCD data recorded from 48 stroke and 34 in-hospital control subjects., Results: We assessed the performance of the algorithm in the improvement of estimation of clinically important TCD parameters by comparing them to that of manual beat annotation. The results show that there is a strong correlation between the two, that demonstrates the algorithm has successfully recovered the clinically important features. We obtained significant improvement in estimating the TCD parameters using the algorithm accepted beats compared to using all beats., Significance: Our algorithm provides a valuable tool to clinicians for automated detection of the reliable portion of the data. Moreover, it can be used as a pre-processing tool to improve the data quality for automated diagnosis of pathologic beat waveforms using machine learning.

Cerebral Blood Flow Velocity waveforms acquired via Transcranial Doppler (TCD) can provide evidence for cerebrovascular occlusion and stenosis. Thrombolysis in Brain Ischemia (TIBI) flow grades are widely used for this purpose, but require subjective assessment by expert evaluators to be reliable. In this work we seek to determine whether TCD morphology can be objectively assessed using an unsupervised machine learning approach to waveform categorization. TCD beat waveforms were recorded at multiple depths from the Middle Cerebral Arteries of 106 subjects; 33 with Large Vessel Occlusion (LVO). From each waveform, three morphological features were extracted, quantifying onset of maximal velocity, systolic canopy length, and the number/prominence of peaks/troughs. Spectral clustering identified groups implicit in the resultant three-dimensional feature space, with gap statistic criteria establishing the optimal cluster number. We found that gap statistic disparity was maximized at four clusters, referred to as flow types I, II, III, and IV. Types I and II were primarily composed of control subject waveforms, whereas types III and IV derived mainly from LVO patients. Cluster morphologies for types I and IV aligned clearly with Normal and Blunted TIBI flows, respectively. Types II and III represented commonly observed flow-types not delineated by TIBI, which nonetheless deviate from normal and blunted flows. We conclude that important morphological variability exists beyond that currently quantified by TIBI in populations experiencing or at-risk for acute ischemic stroke, and posit that the observed flow-types provide the foundation for objective methods of real-time automated flow type classification., Competing Interests: At the time this research was conducted, authors ST, CT, KJ, AD, NC, SW, and RH were salaried employees of Neural Analytics, Inc., and TD was a paid consultant. FS is also a former paid consultant. All authors either hold stock or stock options in the company. The funding institution, Neural Analytics, Inc., holds numerous patents related to transcranial doppler technology from which the authors do not directly stand to benefit. Our commercial affiliation does not alter our adherence to PLOS ONE policies on sharing data and materials. This research was also supported in part by NINDS-1R43NS105340.

Transcranial Doppler (TCD) ultrasound has been demonstrated to be a valuable tool for assessing cerebral hemodynamics via measurement of cerebral blood flow velocity (CBFV), with a number of established clinical indications. However, CBFV waveform analysis depends on reliable pulse onset detection, an inherently difficult task for CBFV signals acquired via TCD. We study the application of a new algorithm for CBFV pulse segmentation, which locates pulse onsets in a sequential manner using a moving difference filter and adaptive thresholding. The test data set used in this study consists of 92,012 annotated CBFV pulses, whose quality is representative of real world data. On this test set, the algorithm achieves a true positive rate of 99.998% (2 false negatives), positive predictive value of 99.998% (2 false positives), and mean temporal offset error of 6.10 ± 4.75 ms. We do note that in this context, the way in which true positives, false positives, and false negatives are defined caries some nuance, so care should be taken when drawing comparisons to other algorithms. Additionally, we find that 97.8% and 99.5% of onsets are detected within 10 and 30 ms, respectively, of the true onsets. The algorithm's performance in spite of the large degree of variation in signal quality and waveform morphology present in the test data suggests that it may serve as a valuable tool for the accurate and reliable identification of CBFV pulse onsets in neurocritical care settings., (Copyright © 2019 Canac, Ranjbaran, O'Brien, Asgari, Scalzo, Thorpe, Jalaleddini, Thibeault, Wilk and Hamilton.)

In addition to a vestibular system, birds uniquely have a balance-sensing organ within the pelvis, called the lumbosacral organ (LSO). The LSO is well developed in terrestrial birds, possibly to facilitate balance control in perching and terrestrial locomotion. No previous studies have quantified the functional benefits of the LSO for balance. We suggest two main benefits of hip-localized balance sense: reduced sensorimotor delay and improved estimation of foot-ground acceleration. We used system identification to test the hypothesis that hip-localized balance sense improves estimates of foot acceleration compared to a head-localized sense, due to closer proximity to the feet. We built a physical model of a standing guinea fowl perched on a platform, and used 3D accelerometers at the hip and head to replicate balance sense by the LSO and vestibular systems. The horizontal platform was attached to the end effector of a 6 DOF robotic arm, allowing us to apply perturbations to the platform analogous to motions of a compliant branch. We also compared state estimation between models with low and high neck stiffness. Cross-correlations revealed that foot-to-hip sensing delays were shorter than foot-to-head, as expected. We used multi-variable output error state-space (MOESP) system identification to estimate foot-ground acceleration as a function of hip- and head-localized sensing, individually and combined. Hip-localized sensors alone provided the best state estimates, which were not improved when fused with head-localized sensors. However, estimates from head-localized sensors improved with higher neck stiffness. Our findings support the hypothesis that hip-localized balance sense improves the speed and accuracy of foot state estimation compared to head-localized sense. The findings also suggest a role of neck muscles for active sensing for balance control: increased neck stiffness through muscle co-contraction can improve the utility of vestibular signals. Our engineering approach provides, to our knowledge, the first quantitative evidence for functional benefits of the LSO balance sense in birds. The findings support notions of control modularity in birds, with preferential vestibular sense for head stability and gaze, and LSO for body balance control,respectively. The findings also suggest advantages for distributed and active sensing for agile locomotion in compliant bipedal robots., (Copyright © 2018 Urbina-Meléndez, Jalaleddini, Daley and Valero-Cuevas.)

An accurate model for ElectroMyoGram (EMG)-torque dynamics has many uses. One of its applications which has gained high attention among researchers is its use, in estimating the muscle contraction level for the efficient control of prosthesis. In this paper, the dynamic relationship between the surface EMG and torque during isometric contractions at the human ankle was studied using system identification techniques. Subjects voluntarily modulated their ankle torque in dorsiflexion direction, by activating their tibialis anterior muscle, while tracking a pseudo-random binary sequence in a torque matching task. The effects of contraction bandwidth, described by torque spectrum, on EMG-torque dynamics were evaluated by varying the visual command switching time. Nonparametric impulse response functions (IRF) were estimated between the processed surface EMG and torque. It was demonstrated that: 1) at low contraction bandwidths, the identified IRFs had unphysiological anticipatory (i.e., non-causal) components, whose amplitude decreased as the contraction bandwidth increased. We hypothesized that this non-causal behavior arose, because the EMG input contained a component due to feedback from the output torque, i.e., it was recorded from within a closed-loop. Vision was not the feedback source since the non-causal behavior persisted when visual feedback was removed. Repeating the identification using a nonparametric closed-loop identification algorithm yielded causal IRFs at all bandwidths, supporting this hypothesis. 2) EMG-torque dynamics became faster and the bandwidth of system increased as contraction modulation rate increased. Thus, accurate prediction of torque from EMG signals must take into account the contraction bandwidth sensitivity of this system.

Key Points: In tonic, isometric, plantarflexion contractions, physiological tremor increases as the ankle joint becomes plantarflexed. Modulation of physiological tremor as a function of muscle stretch differs from that of the stretch reflex amplitude. Amplitude of physiological tremor may be altered as a function of reflex pathway gains. Healthy humans likely increase their γ-static fusimotor drive when muscles shorten. Quantification of physiological tremor by manipulation of joint angle may be a useful experimental probe of afferent gains and/or the integrity of automatic fusimotor control., Abstract: The involuntary force fluctuations associated with physiological (as distinct from pathological) tremor are an unavoidable component of human motor control. While the origins of physiological tremor are known to depend on muscle afferentation, it is possible that the mechanical properties of muscle-tendon systems also affect its generation, amplification and maintenance. In this paper, we investigated the dependence of physiological tremor on muscle length in healthy individuals. We measured physiological tremor during tonic, isometric plantarflexion torque at 30% of maximum at three ankle angles. The amplitude of physiological tremor increased as calf muscles shortened in contrast to the stretch reflex whose amplitude decreases as muscle shortens. We used a published closed-loop simulation model of afferented muscle to explore the mechanisms responsible for this behaviour. We demonstrate that changing muscle lengths does not suffice to explain our experimental findings. Rather, the model consistently required the modulation of  γ-static fusimotor drive to produce increases in physiological tremor with muscle shortening - while successfully replicating the concomitant reduction in stretch reflex amplitude. This need to control γ-static fusimotor drive explicitly as a function of muscle length has important implications. First, it permits the amplitudes of physiological tremor and stretch reflex to be decoupled. Second, it postulates neuromechanical interactions that require length-dependent γ drive modulation to be independent from α drive to the parent muscle. Lastly, it suggests that physiological tremor can be used as a simple, non-invasive measure of the afferent mechanisms underlying healthy motor function, and their disruption in neurological conditions., (© 2017 The Authors. The Journal of Physiology © 2017 The Physiological Society.)

This paper describes a new small signal parametric model of ankle joint intrinsic mechanics in normal subjects. We found that intrinsic ankle mechanics is a third-order system and the second-order mass-spring-damper model, referred to as IBK, used by many researchers in the literature cannot adequately represent ankle dynamics at all frequencies in a number of important tasks. This was demonstrated using experimental data from five healthy subjects with no voluntary muscle contraction and at seven ankle positions covering the range of motion. We showed that the difference between the new third-order model and the conventional IBK model increased from dorsi to plantarflexed position. The new model was obtained using a multi-step identification procedure applied to experimental input/output data of the ankle joint. The procedure first identifies a non-parametric model of intrinsic joint stiffness where ankle position is the input and torque is the output. Then, in several steps, the model is converted into a continuous-time transfer function of ankle compliance, which is the inverse of stiffness. Finally, we showed that the third-order model is indeed structurally consistent with agonist-antagonist musculoskeletal structure of human ankle, which is not the case for the IBK model.

This paper presents our new method, Short Segment-Structural Decomposition SubSpace (SS-SDSS), for the estimation of dynamic joint stiffness from short data segments. The main application is for data sets that are only piecewise stationary. Our approach is to: 1) derive a data-driven, mathematical model for dynamic stiffness for short data segments; 2) bin the non-stationary data into a number of short, stationary data segments; and 3) estimate the model parameters from subsets of segments with the same properties. This method extends our previous state-spacework by recognizing that initial conditions have important effects for short data segments; consequently, initial conditions are incorporated into the stiffness model and estimated for each segment. A simulation study that faithfully replicated experimental conditions delineated the range of experimental conditions for which the method can successfully identify stiffness. An experimental study on the ankle of a healthy subject during a torque matching tasks demonstrated the successful estimation of dynamic stiffness in a slow, time-varying experiment. Together, the simulation and experimental studies demonstrate that the SS-SDSS method is a valuable tool to measure stiffness in functionally important tasks.

Objective: The purpose of this paper is to present a structural decomposition subspace (SDSS) method for decomposition of the joint torque to intrinsic, reflexive, and voluntary torques and identification of joint dynamic stiffness., Methods: First, it formulates a novel state-space representation for the joint dynamic stiffness modeled by a parallel-cascade structure with a concise parameter set that provides a direct link between the state-space representation matrices and the parallel-cascade parameters. Second, it presents a subspace method for the identification of the new state-space model that involves two steps: 1) the decomposition of the intrinsic and reflex pathways and 2) the identification of an impulse response model of the intrinsic pathway and a Hammerstein model of the reflex pathway., Results: Extensive simulation studies demonstrate that SDSS has significant performance advantages over some other methods. Thus, SDSS was more robust under high noise conditions, converging where others failed; it was more accurate, giving estimates with lower bias and random errors. The method also worked well in practice and yielded high-quality estimates of intrinsic and reflex stiffnesses when applied to experimental data at three muscle activation levels., Conclusion: The simulation and experimental results demonstrate that SDSS accurately decomposes the intrinsic and reflex torques and provides accurate estimates of physiologically meaningful parameters., Significance: SDSS will be a valuable tool for studying joint stiffness under functionally important conditions. It has important clinical implications for the diagnosis, assessment, objective quantification, and monitoring of neuromuscular diseases that change the muscle tone.

Children with cerebral palsy (CP) often develop reduced passive range of motion with age. The determining factor underlying this process is believed to be progressive development of contracture in skeletal muscle that likely changes the biomechanics of the joints. Consequently, to identify the underlying mechanisms, we modeled the mechanical characteristics of the forearm flexors acting across the wrist joint. We investigated skeletal muscle strength (Grippit®) and passive stiffness and viscosity of the forearm flexors in 15 typically developing (TD) children (10 boys/5 girls, mean age 12 years, range 8-18 yrs) and nine children with CP Nine children (6 boys/3 girls, mean age 11 ± 3 years (yrs), range 7-15 yrs) using the NeuroFlexor® apparatus. The muscle stiffness we estimate and report is the instantaneous mechanical response of the tissue that is independent of reflex activity. Furthermore, we assessed cross-sectional area of the flexor carpi radialis (FCR) muscle using ultrasound. Age and body weight did not differ significantly between the two groups. Children with CP had a significantly weaker (-65%, p < 0.01) grip and had smaller cross-sectional area (-43%, p < 0.01) of the FCR muscle. Passive stiffness of the forearm muscles in children with CP was increased 2-fold ( p < 0.05) whereas viscosity did not differ significantly between CP and TD children. FCR cross-sectional area correlated to age ( R 2 = 0.58, p < 0.01), body weight ( R 2 = 0.92, p < 0.0001) and grip strength ( R 2 = 0.82, p < 0.0001) in TD children but only to grip strength ( R 2 = 0.60, p < 0.05) in children with CP. We conclude that children with CP have weaker, thinner, and stiffer forearm flexors as compared to typically developing children.

Objective: One goal of neuromorphic engineering is to create 'realistic' robotic systems that interact with the physical world by adopting neuromechanical principles from biology. Critical to this is the methodology to implement the spinal circuitry responsible for the behavior of afferented muscles. At its core, muscle afferentation is the closed-loop behavior arising from the interactions among populations of muscle spindle afferents, alpha and gamma motoneurons, and muscle fibers to enable useful behaviors., Approach: We used programmable very- large-scale-circuit (VLSI) hardware to implement simple models of spiking neurons, skeletal muscles, muscle spindle proprioceptors, alpha-motoneuron recruitment, gamma motoneuron control of spindle sensitivity, and the monosynaptic circuitry connecting them. This multi-scale system of populations of spiking neurons emulated the physiological properties of a pair of antagonistic afferented mammalian muscles (each simulated by 1024 alpha- and gamma-motoneurones) acting on a joint via long tendons., Main Results: This integrated system was able to maintain a joint angle, and reproduced stretch reflex responses even when driving the nonlinear biomechanics of an actual cadaveric finger. Moreover, this system allowed us to explore numerous values and combinations of gamma-static and gamma-dynamic gains when driving a robotic finger, some of which replicated some human pathological conditions. Lastly, we explored the behavioral consequences of adopting three alternative models of isometric muscle force production. We found that the dynamic responses to rate-coded spike trains produce force ramps that can be very sensitive to tendon elasticity, especially at high force output., Significance: Our methodology produced, to our knowledge, the first example of an autonomous, multi-scale, neuromorphic, neuromechanical system capable of creating realistic reflex behavior in cadaveric fingers. This research platform allows us to explore the mechanisms behind healthy and pathological sensorimotor function in the physical world by building them from first principles, and it is a precursor to neuromorphic robotic systems.

Objective: We studied the fundamentals of muscle afferentation by building a Neuro-mechano-morphic system actuating a cadaveric finger. This system is a faithful implementation of the stretch reflex circuitry. It allowed the systematic exploration of the effects of different fusimotor drives to the muscle spindle on the closed-loop stretch reflex response., Approach: As in Part I of this work, sensory neurons conveyed proprioceptive information from muscle spindles (with static and dynamic fusimotor drive) to populations of α-motor neurons (with recruitment and rate coding properties). The motor commands were transformed into tendon forces by a Hill-type muscle model (with activation-contraction dynamics) via brushless DC motors. Two independent afferented muscles emulated the forces of flexor digitorum profundus and the extensor indicis proprius muscles, forming an antagonist pair at the metacarpophalangeal joint of a cadaveric index finger. We measured the physical response to repetitions of bi-directional ramp-and-hold rotational perturbations for 81 combinations of static and dynamic fusimotor drives, across four ramp velocities, and three levels of constant cortical drive to the α-motor neuron pool., Main Results: We found that this system produced responses compatible with the physiological literature. Fusimotor and cortical drives had nonlinear effects on the reflex forces. In particular, only cortical drive affected the sensitivity of reflex forces to static fusimotor drive. In contrast, both static fusimotor and cortical drives reduced the sensitivity to dynamic fusimotor drive. Interestingly, realistic signal-dependent motor noise emerged naturally in our system without having been explicitly modeled., Significance: We demonstrate that these fundamental features of spinal afferentation sufficed to produce muscle function. As such, our Neuro-mechano-morphic system is a viable platform to study the spinal mechanisms for healthy muscle function-and its pathologies such as dystonia and spasticity. In addition, it is a working prototype of a robust biomorphic controller for compliant robotic limbs and exoskeletons.

The dynamic relationship between the joint position and reflex EMG in ankle muscles of healthy human subjects was studied for time-varying (TV) contractions. A linear parameter varying (LPV) identification algorithm was used to estimate the Hammerstein system relating ankle position to the reflex EMG response. The estimated Hammerstein system comprised a time-invariant (TI) linear element and a TV static nonlinearity that resembled a half-wave rectifier with a threshold and linear gain. The results demonstrated a systematic change in the reflex nonlinearity with the activation level. The gain of TV nonlinearity increased with activation level reaching its peak at 20-30% maximum voluntary contraction and then decreased. The threshold of the nonlinearity decreased with increasing activation level reaching it minimum at the same point where the gain was maximal. Using the LPV-Hammerstein method in this work, the underlying TV dynamics were extracted from small number of trials. Thus, this method can be used to study stretch reflexes in subjects with neuromuscular disorders.

This paper describes a novel method for the identification of time-varying ankle joint dynamic stiffness during large passive movements. The method estimates a linear parameter varying parallel-cascade (LPV-PC) model of joint stiffness consisting of two pathways: (a) an LPV impulse response function (IRF) for intrinsic mechanics and (b) an LPV Hammerstein cascade with time-varying static nonlinearity and a time-invariant linear dynamics for the reflex pathway. A subspace identification technique is used to estimate a statespace representation of the reflex stiffness dynamics. Then, an orthogonal projection decouples intrinsic from reflex response and subsequently identifies an LPV-IRF model of intrinsic stiffness. Finally, an LPV model of the reflex static nonlinearity is estimated using an iterative, separable least squares method. The LPV method was validated using experimental data from two healthy subjects where the ankle was moved passively by an actuator through its range of motion first without and then with perturbations. The identification results demonstrated that (a) the dynamic response of the intrinsic pathway changes systematically with joint position; and (b) the static nonlinearity of the reflex pathway resembles a half-wave rectifier whose threshold decreases and gain increases as ankle is moved to dorsiflexed position.

This paper presents a state-space (subspace) method for identification of parallel-cascade joint stiffness from short segments of data. It provides unbiased estimates of stiffness by accounting for the contributions of initial conditions of each segment. The method is important in situations where it is not possible to acquire a long stationary data due to switching or time-varying behavior. The power of the method was demonstrated by using it to efficiently characterize ankle joint stiffness through the joint's range of motion.

This paper describes a new subspace-based algorithm for the identification of Hammerstein systems. It extends a previous approach which described the Hammerstein cascade by a state-space model and identified it with subspace methods that are fast and require little a priori knowledge. The resulting state-space models predict the system response well but have many redundant parameters and provide limited insight into the system since they depend on both the nonlinear and linear elements. This paper addresses these issues by reformulating the problem so that there are many fewer parameters and each parameter is related directly to either the linear dynamics or the static nonlinearity. Consequently, it is straightforward to construct the continuous-time Hammerstein models corresponding to the estimated state-space model. Simulation studies demonstrated that the new method performs better than other well-known methods in the nonideal conditions that prevail during practical experiments. Moreover, it accurately distinguished changes in the linear component from those in the static nonlinearity. The practical application of the new algorithm was demonstrated by applying it to experimental data from a study of the stretch reflex at the human ankle. Hammerstein models were estimated between the velocity of ankle perturbations and the EMG activity of triceps surae for voluntary contractions in the plantarflexing and dorsiflexion directions. The resulting models described the behavior well, displayed the expected unidirectional rate sensitivity, and revealed that both the gain of the linear element and the threshold of the nonlinear changed with contraction direction.

This paper describes a novel model structure and identification method for the time-varying, intrinsic stiffness of human ankle joint during imposed walking (IW) movements. The model structure is based on the superposition of a large signal, linear, time-invariant (LTI) model and a small signal linear-parameter varying (LPV) model. The methodology is based on a two-step algorithm; the LTI model is first estimated using data from an unperturbed IW trial. Then, the LPV model is identified using data from a perturbed IW trial with the output predictions of the LTI model removed from the measured torque. Experimental results demonstrate that the method accurately tracks the continuous-time variation of normal ankle intrinsic stiffness when the joint position changes during the IW movement. Intrinsic stiffness gain decreases from full plantarflexion to near the mid-point of plantarflexion and then increases substantially as the ankle is dosriflexed.

This paper describes a state-space representation of the parallel-cascade model of ankle joint stiffness whose parameters are directly related to the underlying dynamics of the system. It then proposes a two step subspace method to identify this model. In the first step, the intrinsic stiffness is estimated using proper orthogonal projections. In the second step, the reflexive pathway is estimated by iterating between estimating its nonlinear and linear components. The identified models can be easily converted to continuous-time for physiological interpretation. Monte-Carlo studies using simulated data which replicate closely the experimental conditions, were used to compare the performance of the new method with the previous parallel-cascade, and subspace methods. The new method is more robust to noise and is guaranteed to converge.

Dynamic ankle joint stiffness defines the relationship between the position of the ankle and the torque acting about it and can be separated into intrinsic and reflex components. Under stationary conditions, intrinsic stiffness can described by a linear second order system while reflex stiffness is described by Hammerstein system whose input is delayed velocity. Given that reflex and intrinsic torque cannot be measured separately, there has been much interest in the development of system identification techniques to separate them analytically. To date, most methods have been nonparametric and as a result there is no direct link between the estimated parameters and those of the stiffness model. This paper presents a novel algorithm for identification of a discrete-time model of ankle stiffness. Through simulations we show that the algorithm gives unbiased results even in the presence of large, non-white noise. Application of the method to experimental data demonstrates that it produces results consistent with previous findings.

This paper describes a novel method for the identification of Hammerstein systems with time-varying (TV) static nonlinearities and time invariant (TI) linear elements. This paper develops a linear parameter varying (LPV) state-space representation for such systems and presents a subspace identification technique that gives individual estimates of the Hammerstein components. The identification method is validated using simulated data of a TV model of ankle joint reflex stiffness where the threshold and gain of the model change as nonlinear functions of an exogenous signal. Pilot experiment of TV reflex EMG response identification in normal ankle joint during an imposed walking task demonstrate systematic changes in the reflex nonlinearity with the trajectory of joint position.

Noise characteristics play an important role in evaluating tools developed to study biomedical systems. Despite usual assumptions, noise in biomedical systems is often nonwhite or non-Gaussian. In this paper, we present a method to analyze the noise component of a biomedical system. We demonstrate the effectiveness of the method in the analysis of noise in voluntary ankle torque measured by a torque transducer and eye movements measured by electro-oculography (EOG).

This paper presents an algorithm for the identification of Hammerstein cascades with hard nonlinearities. The nonlinearity of the cascade is described using a B-spline basis with fixed knot locations; the linear dynamics are described using a state-space model. The algorithm automatically estimates both the order of the linear system and the number and locations of the knots used to characterize the nonlinearity. Therefore, it significantly reduces the a priori knowledge about the underlying system required for identification. A simulation study on a model of reflex stiffness shows that the new method estimates the nonlinearity accurately in the presence of output noise.

Reflex stiffness is often modeled as a Hammerstein system comprising a cascade of a static nonlinear element and a linear dynamic element. The nonlinearity is frequently modeled as a half wave rectifier so that changes in the reflex response can only be modeled by changes in the parameters of the linear element. This is an oversimplification since there are physiological mechanisms that could change both the threshold of the nonlinearity and the linear dynamics. This study explores the ability of a new subspace identification algorithm to distinguish changes in parameters of the nonlinear element from those of the linear element. Simulation studies demonstrate that the method does so very effectively even in the presence of substantial output noise. Pilot experiments in which the method was applied to stretch reflex EMG data revealed that both the threshold of the nonlinearity and the gain of the linear element change with muscle activation.