scholarly journals Bringing Proportional Recovery into Proportion: Bayesian Hierarchical Modelling of Post-Stroke Motor Performance

2019 ◽  
Author(s):  
Anna K. Bonkhoff ◽  
Thomas Hope ◽  
Danilo Bzdok ◽  
Adrian G. Guggisberg ◽  
Rachel L. Hawe ◽  
...  
Keyword(s):  
Motor Performance ◽  
Model Comparison ◽  
Synthetic Data ◽  
Motor Recovery ◽  
Stroke Recovery ◽  
Future Research ◽  
Single Subject ◽  
Bayesian Hierarchical ◽  
Post Stroke ◽  
Function Recovery

AbstractAccurate predictions of motor performance after stroke are of cardinal importance for the patient, clinician, and health care system. More than ten years ago, the proportional recovery rule was introduced by promising just that: high-fidelity predictions of recovery following stroke based only on the initially lost motor performance, at least for a specific fraction of patients. However, emerging evidence suggests that this recovery rule is subject to various confounds and may apply less universally than assumed by many.We systematically revisited stroke outcome predictions by casting the data in a less confounded form and employing more integrative and flexible hierarchical Bayesian models. We jointly analyzed n=385 post-stroke trajectories from six separate studies – the currently largest overall dataset of upper limb motor recovery. We addressed confounding ceiling effects by introducing a subset approach and ensured correct model estimation through synthetic data simulations. Finally, we used model comparisons to assess the underlying nature of recovery within our empirical recovery data.The first model comparison, relying on the conventional fraction of patients called fitters, pointed to a combination of constant and proportional to lost function recovery. Proportional to lost here describes the original notion of proportionality, indicating greater recovery in case of a more pronounced initial deficit. This combination explained only 32% of the variance in recovery, which is in stark contrast to previous reports of >80%. When instead analyzing the complete spectrum of subjects, model comparison selected a composite of constant and proportional to spared function recovery, implying a more significant improvement in case of more preserved function. Explained variance was at 53%.Therefore, our data suggest that motor recovery post-stroke may exhibit some characteristics of proportionality. However, the levels of explanatory value were substantially reduced compared to what has previously been reported. This finding motivates future research moving beyond solely behavior scores to explain stroke recovery and establish robust single-subject predictions.

scholarly journals Bringing proportional recovery into proportion: Bayesian modelling of post-stroke motor impairment

Brain ◽  
2020 ◽  
Vol 143 (7) ◽  
pp. 2189-2206 ◽  
Author(s):  
Anna K Bonkhoff ◽  
Thomas Hope ◽  
Danilo Bzdok ◽  
Adrian G Guggisberg ◽  
Rachel L Hawe ◽  
...  
Keyword(s):  
Model Comparison ◽  
Motor Impairment ◽  
Synthetic Data ◽  
Motor Recovery ◽  
Stroke Recovery ◽  
Future Research ◽  
Single Subject ◽  
Post Stroke ◽  
Specific Fraction ◽  
Explained Variance

Abstract Accurate predictions of motor impairment after stroke are of cardinal importance for the patient, clinician, and healthcare system. More than 10 years ago, the proportional recovery rule was introduced by promising that high-fidelity predictions of recovery following stroke were based only on the initially lost motor function, at least for a specific fraction of patients. However, emerging evidence suggests that this recovery rule is subject to various confounds and may apply less universally than previously assumed. Here, we systematically revisited stroke outcome predictions by applying strategies to avoid confounds and fitting hierarchical Bayesian models. We jointly analysed 385 post-stroke trajectories from six separate studies—one of the largest overall datasets of upper limb motor recovery. We addressed confounding ceiling effects by introducing a subset approach and ensured correct model estimation through synthetic data simulations. Subsequently, we used model comparisons to assess the underlying nature of recovery within our empirical recovery data. The first model comparison, relying on the conventional fraction of patients called ‘fitters’, pointed to a combination of proportional to lost function and constant recovery. ‘Proportional to lost’ here describes the original notion of proportionality, indicating greater recovery in case of a more severe initial impairment. This combination explained only 32% of the variance in recovery, which is in stark contrast to previous reports of >80%. When instead analysing the complete spectrum of subjects, ‘fitters’ and ‘non-fitters’, a combination of proportional to spared function and constant recovery was favoured, implying a more significant improvement in case of more preserved function. Explained variance was at 53%. Therefore, our quantitative findings suggest that motor recovery post-stroke may exhibit some characteristics of proportionality. However, the variance explained was substantially reduced compared to what has previously been reported. This finding motivates future research moving beyond solely behaviour scores to explain stroke recovery and establish robust and discriminating single-subject predictions.

Abstract 14: Effects of Lesion Laterality on Post-Stroke Motor Performance: An ENIGMA Stroke Recovery Analysis

Stroke ◽  
2017 ◽  
Vol 48 (suppl_1) ◽  
Author(s):  
Sook-Lei Liew ◽  
Neda Jahanshad ◽  
Lisa Aziz-Zadeh ◽  
Niels Birbaumer ◽  
Michael Borich ◽  
...  
Keyword(s):  
Motor Performance ◽  
Statistical Significance ◽  
Motor Recovery ◽  
Small Sample ◽  
Stroke Recovery ◽  
Intracranial Volume ◽  
Brain Volumes ◽  
Post Stroke ◽  
Small Sample Sizes ◽  
Cortical Regions

The laterality of the lesioned hemisphere is often overlooked in stroke recovery research due to small sample sizes. Here, we used a well-powered dataset from ENIGMA Stroke Recovery (a consortium that harmonizes post-stroke MRIs and behavioral data worldwide; http://enigma.usc.edu) to analyze the effects of left (LHL) versus right (RHL) hemisphere lesions on motor performance. Given the different functional roles of each hemisphere, we hypothesized that the LHL group should show better motor performance, and, consequently, different brain-behavior relationships, compared to the RHL group. Data from over 2000 stroke patients across 20 sites worldwide has been committed. To date, structural T1-weighted MRIs from n=343 (10 sites) have been analyzed (LHL n=174; RHL n=169). ENIGMA protocols extracted volumes of subcortical regions of interest and provided quality control. Regression analyses examined brain volumes as predictors of motor outcomes. Motor scores were combined across scales/sites, with each score expressed as a percentage of the maximum score. Covariates (e.g., age, sex, intracranial volume) and manually marked lesion effects were also modeled. Statistical significance was assessed nonparametrically by permutation. As anticipated, the LHL group had better motor performance compared to the RHL group (t(1,341)=3.07, p=0.0023). In addition, while the combined LHL+RHL analyses showed significant associations between motor scores and volumes of the basal ganglia/lateral ventricles, separate group analyses showed strong associations for the LHL group, but only one association for the RHL group (Table 1). This may suggest that motor recovery following RH lesions is more heterogeneous or relies more on cortical regions/networks that were not assessed here. While further research is needed, these results suggest that laterality of the lesioned hemisphere affects neural patterns related to motor recovery and should be carefully examined.

Abstract TP210: Examining the Role of Context: Neighborhood Socioeconomic Status Shaping Post-Stroke Outcomes

Stroke ◽  
2020 ◽  
Vol 51 (Suppl_1) ◽  
Author(s):  
Eric Stulberg ◽  
Erica Twardzik ◽  
Chia-Wei Hsu ◽  
Sehee Kim ◽  
Philippa Clarke ◽  
...  
Keyword(s):  
Socioeconomic Status ◽  
Activities Of Daily Living ◽  
Daily Living ◽  
Neighborhood Context ◽  
Stroke Recovery ◽  
Future Research ◽  
Stroke Severity ◽  
Stroke Outcomes ◽  
Neighborhood Socioeconomic Status ◽  
Post Stroke

Introduction: Neighborhoods may influence post-stroke recovery. We examined the association between neighborhood socioeconomic status (nSES) and 90-day post-stroke function, depression, cognition, and quality of life (QoL). Methods: Stroke survivors (N=782) were identified from the population-based Brain Attack Surveillance in Corpus Christi (BASIC) Project. An nSES index – composite of 2010 census-tract level income, wealth, education, employment – was the exposure; higher values indicate higher nSES (median -4.56; IQR: -7.48 to -0.46). Function was measured with 22 self-reported activities of daily living/instrumental activities of daily living, depression with Patient Health Questionnaire-8, QoL with the Stroke Specific QoL Scale, and cognition with the Modified Mini Mental State Examination. Confounder-adjusted generalized estimating equations were used to estimate associations between nSES (comparing 75 th to 25 th percentile) and 90-day outcomes. We tested for effect modification by initial stroke severity (NIH Stroke Scale (NIHSS) ≤ 5 or >5) by including interaction terms in adjusted models. Results: Higher nSES was associated with significantly better function, better QoL, and less depression after adjusting for person-level confounders in those with NIHSS >5. Higher nSES was associated with better cognition, but this result was not significant. In those with NIHSS ≤5, higher nSES had a statistically significant (though attenuated) association with function and cognition. Conclusions: Future research should identify features of higher nSES neighborhoods that contribute to more favorable stroke outcomes. Our findings highlight the need for examining the individual and joint influence of neighborhood context and stroke severity on post-stroke recovery.

Abstract TMP31: Peripheral Nerve Stimulation Paired with Constraint-Induced Therapy to Enhance Post-Stroke Upper Extremity Motor Performance

Stroke ◽  
2013 ◽  
Vol 44 (suppl_1) ◽  
Author(s):  
Cheryl Carrico ◽  
KC Chelette ◽  
Laurie Nichols ◽  
Lumy Sawaki
Keyword(s):  
Peripheral Nerve ◽  
Upper Extremity ◽  
Motor Function ◽  
Nerve Stimulation ◽  
Motor Performance ◽  
Motor Recovery ◽  
Peripheral Nerve Stimulation ◽  
Motor Deficit ◽  
Motor Training ◽  
Post Stroke

Research has shown that peripheral nerve stimulation (PNS) can enhance motor learning following cortical lesions. Studies have also shown that intensive upper extremity motor training can significantly enhance post-stroke motor performance. Constraint-induced therapy (CIT) is a form of intensive training that restricts use of the non-paretic upper extremity during repetitive, task-oriented motor training of the paretic extremity. Extensive evidence has validated the effectiveness of CIT for enhancing post-stroke upper extremity motor recovery. No studies have evaluated how PNS may modulate the effects of CIT. Therefore, we conducted a pilot study of PNS paired with CIT and hypothesized that in subjects with stroke, pairing CIT with active PNS would lead to significantly more improved motor function in the paretic upper extremity than CIT paired with sham PNS. Outcome measures included the Fugl-Meyer Assessment Scale (FMA; primary outcome measure), the Wolf Motor Function Test (WMFT), and the Action Research Arm Test (ARAT). Nineteen chronic stroke subjects with mild to moderate upper extremity motor deficit received 2 hours of either active (n=10) or sham (n=9) PNS preceding 4 hours of CIT for 10 consecutive weekdays. Changes in FMA, WMFT, and ARAT were analyzed using factorial ANOVA. Results showed significant (p<0.05) change in all measures at completion evaluation compared with baseline (FMA (p=0.005); WMFT (p=0.030); ARAT (p=0.020)) as well as 1-month follow-up compared with baseline (FMA (p=0.048); WMFT (p=0.045); ARAT (p=0.047)). These results highlight the enormous potential for PNS paired with CIT to enhance post-stroke upper extremity motor recovery more effectively than CIT alone.

scholarly journals Targeting GABAC Receptors Improves Post-Stroke Motor Recovery

2021 ◽  
Vol 11 (3) ◽  
pp. 315
Author(s):  
Petra S. van Nieuwenhuijzen ◽  
Kim Parker ◽  
Vivian Liao ◽  
Josh Houlton ◽  
Hye-Lim Kim ◽  
...  
Keyword(s):  
Infarct Size ◽  
Motor Function ◽  
Treatment Options ◽  
Motor Recovery ◽  
Stroke Recovery ◽  
Xenopus Laevis Oocytes ◽  
Reactive Astrogliosis ◽  
Delayed Treatment ◽  
Post Stroke ◽  
Gabac Receptors

Ischemic stroke remains a leading cause of disability worldwide, with limited treatment options available. This study investigates GABAC receptors as novel pharmacological targets for stroke recovery. The expression of ρ1 and ρ2 mRNA in mice were determined in peri-infarct tissue following photothrombotic motor cortex stroke. (R)-4-amino-cyclopent-1-enyl butylphosphinic acid (R)-4-ACPBPA and (S)-4-ACPBPA were assessed using 2-elecotrode voltage electrophysiology in Xenopus laevis oocytes. Stroke mice were treated for 4 weeks with either vehicle, the α5-selective negative allosteric modulator, L655,708, or the ρ1/2 antagonists, (R)-4-ACPBPA and (S)-4-ACPBPA respectively from 3 days post-stroke. Infarct size and expression levels of GAT3 and reactive astrogliosis were determined using histochemistry and immunohistochemistry respectively, and motor function was assessed using both the grid-walking and cylinder tasks. After stroke, significant increases in ρ1 and ρ2 mRNAs were observed on day 3, with ρ2 showing a further increase on day 7. (R)- and (S)-4-ACPBPA are both potent antagonists at ρ2 and only weak inhibitors of α5β2γ2 receptors. Treatment with either L655,708, (S)-4-ACPBPA (ρ1/2 antagonist; 5 mM only), or (R)-4-ACPBPA (ρ2 antagonist; 2.5 and 5 mM) from 3 days after stroke resulted in a significant improvement in motor recovery on the grid-walking task, with L655,708 and (R)-4-ACPBPA also showing an improvement in the cylinder task. Infarct size was unaffected, and only (R)-4-ACPBPA significantly increased peri-infarct GAT3 expression and decreased the level of reactive astrogliosis. Importantly, inhibiting GABAC receptors affords significant improvement in motor function after stroke. Targeting the ρ-subunit could provide a novel delayed treatment option for stroke recovery.

Comparing CST Lesion Metrics as Biomarkers for Recovery of Motor and Proprioceptive Impairments After Stroke

2019 ◽  
Vol 33 (10) ◽  
pp. 848-861 ◽  
Author(s):  
Sonja E. Findlater ◽  
Rachel L. Hawe ◽  
Erin L. Mazerolle ◽  
Abdulaziz S. Al Sultan ◽  
Jessica M. Cassidy ◽  
...  
Keyword(s):  
Magnetic Resonance Imaging ◽  
Diffusion Magnetic Resonance Imaging ◽  
Motor Performance ◽  
Corticospinal Tract ◽  
Region Of Interest ◽  
Internal Capsule ◽  
Stroke Recovery ◽  
Future Research ◽  
Resonance Imaging ◽  
Posterior Limb

Background. Corticospinal tract (CST) damage is considered a biomarker for stroke recovery. Several methods have been used to define CST damage and examine its relationship to motor performance, but which method is most useful remains unclear. Proprioceptive impairment also affects stroke recovery and may be related to CST damage. Methods. Robotic assessment quantified upper-limb motor and proprioceptive performance at 2 weeks and 6 months poststroke (n = 149). Three previously-established CST lesion metrics were calculated using clinical neuroimaging. Diffusion magnetic resonance imaging quantified CST microstructure in a subset of participants (n = 21). Statistical region of interest (sROI) analysis identified lesion locations associated with motor and proprioceptive deficits. Results. CST lesion metrics were moderately correlated with motor scores at 2 weeks and 6 months poststroke. CST fractional anisotropy (FA) was correlated with motor scores at 1 month poststroke, but not at 6 months. The FA ratio of the posterior limb of the internal capsule was not correlated with motor performance. CST lesion metrics were moderately correlated with proprioceptive scores at 2 weeks and 6 months poststroke. sROI analysis confirmed that CST damage was associated with motor and proprioceptive deficits and additionally found that putamen, internal capsule, and corticopontocerebellar tract lesions were associated with poor motor performance. Conclusions. Across all methods used to quantify CST damage, correlations with motor or proprioceptive performance were moderate at best. Future research is needed to identify complementary or alternative biomarkers to address the complexity and heterogeneity of stroke recovery.

scholarly journals Dynamic connectivity predicts acute motor impairment and recovery post-stroke

2020 ◽  
Author(s):  
Anna K Bonkhoff ◽  
Anne Rehme ◽  
Lukas Hensel ◽  
Caroline Tscherpel ◽  
Lukas Volz ◽  
...  
Keyword(s):  
Acute Phase ◽  
Resting State ◽  
Prediction Models ◽  
Motor Impairment ◽  
Motor Recovery ◽  
Stroke Recovery ◽  
Integrated Network ◽  
Post Stroke ◽  
Model Based ◽  
Dynamic Connectivity

Objective Thorough assessment of cerebral dysfunction after acute brain lesions is paramount to optimize predicting short- and long-term clinical outcomes. The potential of dynamic resting-state connectivity for prognosticating motor recovery has not been explored so far. Methods We built random forest classifier-based prediction models of acute upper limb motor impairment and recovery after stroke. Predictions were based on structural and resting-state fMRI data from 54 ischemic stroke patients scanned within the first days of symptom onset. Functional connectivity was estimated using both a static and dynamic approach. Individual motor performance was phenotyped in the acute phase and six months later. Results A model based on the time spent in specific dynamic connectivity configurations achieved the best discrimination between patients with and without motor impairments (out-of-sample area under the curve and 95%-confidence interval (AUC±95%-CI): 0.67±0.01). In contrast, patients with moderate-to-severe impairments could be differentiated from patients with mild deficits using a model based on the variability of dynamic connectivity (AUC±95%-CI: 0.83±0.01). Here, the variability of the connectivity between ipsilesional sensorimotor cortex and putamen discriminated the most between patients. Finally, motor recovery was best predicted by the time spent in specific connectivity configurations (AUC±95%-CI: 0.89±0.01) in combination with the initial motor impairment. Here, better recovery was linked to a shorter time spent in a functionally integrated network configuration in the acute phase post-stroke. Interpretation Dynamic connectivity-derived parameters constitute potent predictors of acute motor impairment and post-stroke recovery, which in the future might inform personalized therapy regimens to promote recovery from acute stroke.

scholarly journals Community detection in networks without observing edges

2020 ◽  
Vol 6 (4) ◽  
pp. eaav1478 ◽  
Author(s):  
Till Hoffmann ◽  
Leto Peel ◽  
Renaud Lambiotte ◽  
Nick S. Jones
Keyword(s):  
Community Detection ◽  
Model Comparison ◽  
Synthetic Data ◽  
Detection Algorithm ◽  
Climate Data ◽  
Bayesian Hierarchical ◽  
Point Estimates ◽  
Community Detection Algorithm ◽  
Daily Returns ◽  
Selection Of

We develop a Bayesian hierarchical model to identify communities of time series. Fitting the model provides an end-to-end community detection algorithm that does not extract information as a sequence of point estimates but propagates uncertainties from the raw data to the community labels. Our approach naturally supports multiscale community detection and the selection of an optimal scale using model comparison. We study the properties of the algorithm using synthetic data and apply it to daily returns of constituents of the S&P100 index and climate data from U.S. cities.

scholarly journals 3393 Biomarkers of Stroke Recovery Study

2019 ◽  
Vol 3 (s1) ◽  
pp. 35-36
Author(s):  
Matthew A. Edwardson ◽  
Amrita Cheema ◽  
Ming Tan ◽  
Alexander Dromerick
Keyword(s):  
Motor Impairment ◽  
Skilled Nursing Facilities ◽  
Stroke Recovery ◽  
Future Research ◽  
Nursing Facilities ◽  
Early Recovery ◽  
Skilled Nursing ◽  
Neural Repair ◽  
Blood Samples ◽  
Post Stroke

OBJECTIVES/SPECIFIC AIMS: There are currently no established blood-based biomarkers of recovery and neural repair following stroke in humans. Such biomarkers would be extremely valuable for aiding in stroke prognosis, timing rehabilitation therapies, and designing drugs to augment natural repair mechanisms. Metabolites, including lipids and amino acids, are engaged in many cellular processes and cross the blood-brain barrier more easily than proteins. Recent advances in liquid chromatography / mass spectrometry (LCMS) allow researchers to obtain a biochemical fingerprint of the metabolites in various biofluids. Thus, metabolite biomarkers of neural repair after brain injury are a promising avenue for future research. Objective: Design and conduct a study to identify metabolite changes in the blood associated with good and poor motor recovery following stroke. METHODS/STUDY POPULATION: We launched the Biomarkers of Stroke Recovery (BIOREC) study, which seeks to enroll 70 participants suffering arm motor impairment following stroke and 35 matched controls. BIOREC is a longitudinal observational study. Fasting blood samples are collected at 5, 15, and 30 days post-stroke, processed, and stored in the Georgetown Lombardi biorepository. Outcome measures, including measures of motor impairment, cognition and language, are assessed at 5, 15, 30, and 90 days post-stroke. The primary outcome measure is the upper extremity Fugl-Meyer score. Control participants are matched for age +/− 1 yr, race, gender, cardiovascular comorbities, and statin use through a computer algorithm that screens the entire MedStar electronic health record (EHR). Control participants provide 2 fasting blood samples one month apart. Once all samples are collected and sent for LCMS analysis, logistic regression analysis will identify potential metabolite biomarkers by comparing participants with good recovery to those with poor recovery as well as stroke participants to controls. RESULTS/ANTICIPATED RESULTS: To date, forty stroke participants have enrolled from 4 acute care hospitals in the Washington, DC metro region and completed all study procedures. Twenty stroke participants either dropped out or were withdrawn due to other medical concerns. Stroke patients ended up at a variety of venues following their acute hospitalization including the acute rehabilitation hospital, skilled nursing facilities, and home. We learned to overcome these logistical challenges by traveling to wherever the patients were sent and notifying medical providers of their study participation. In rare cases we have paid to transport patients from skilled nursing facilities to the clinic, which has reduced dropouts. In addition to the stroke participants, we have enrolled 7 healthy control participants using the EHR screening algorithm. DISCUSSION/SIGNIFICANCE OF IMPACT: Performing a longitudinal study in the early recovery phase following stroke is logistically challenging, but feasible. Difficulty in identifying participants with isolated motor impairment requires added effort to eliminate dropouts. Screening the EHR is an effective method to identify matched controls. Future metabolomics analysis of stored blood samples holds promise to identify biomarkers of stroke recovery and neural repair.

Sign in / Sign up

Export Citation Format

Share Document