Hope Center member publications

Comparing stress prediction models using smartwatch physiological signals and participant self-reports
(2021) Computer Methods and Programs in Biomedicine, 208, art. no. 106207, . 

Dai, R.^a , Lu, C.^a , Yun, L.^b , Lenze, E.^c , Avidan, M.^b , Kannampallil, T.^b d

^a Department of Computer Science, McKelvey School of Engineering, United States
^b Department of Anesthesiology, United States
^c Department of Psychiatry, United States
^d Institute for Informatics, School of Medicine, Washington University in St. Louis, St Louis, MO, United States

Abstract
Recent advances in wearable technology have facilitated the non-obtrusive monitoring of physiological signals, creating opportunities to monitor and predict stress. Researchers have utilized machine learning methods using these physiological signals to develop stress prediction models. Many of these prediction models have utilized objective stressor tasks (e.g., a public speaking task or solving math problems). Alternatively, the subjective user responses with self-reports have also been used for measuring stress. In this paper, we describe a methodological approach (a) to compare the prediction performance of models developed using objective markers of stress using participant-reported subjective markers of stress from self-reports; and (b) to develop personalized stress models by accounting for inter-individual differences. Towards this end, we conducted a laboratory-based study with 32 healthy volunteers. Participants completed a series of stressor tasks—social, cognitive and physical—wearing an instrumented commercial smartwatch that collected physiological signals and participant responses using timed self-reports. After extensive data preprocessing using a combination of signal processing techniques, we developed two types of models: objective stress models using the stressor tasks as labels; and subjective stress models using participant responses to each task as the label for that stress task. We trained and tested several machine learning algorithms—support vector machine (SVM), random forest (RF), gradient boosted trees (GBT), AdaBoost, and Logistic Regression (LR)—and evaluated their performance. SVM had the best performance for the models using the objective stressor (i.e., stressor tasks) with an AUROC of 0.790 and an F-1 score of 0.623. SVM also had the highest performance for the models using the subjective stress (i.e., participant self-reports) with an AUROC of 0.719 and an F-1 score of 0.520. Model performance improved with a personalized threshold model to an AUROC of 0.751 and an F-1 score of 0.599. The performance of the stress models using an instrumented commercial smartwatch was comparable to similar models from other state-of-the-art laboratory-based studies. However, the subjective stress models had a lower performance, indicating the need for further research on the use of self-reports for stress-related studies. The improvement in performance with the personalized threshold-based models provide new directions for building stress prediction models. © 2021 Elsevier B.V.

Author Keywords
Objective stress;  Personalized threshold;  Smartwatch;  Subjective stress

Funding details
Washington University School of Medicine in St. Louis

Document Type: Article
Publication Stage: Final
Source: Scopus

Black-White racial health disparities in inflammation and physical health: Cumulative stress, social isolation, and health behaviors
(2021) Psychoneuroendocrinology, 131, art. no. 105251, . 

McClendon, J.^a b , Chang, K.^c , J. Boudreaux, M.^d , Oltmanns, T.F.^e , Bogdan, R.^e

^a National Center for PTSD, VA Boston Healthcare System, Boston, MA, United States
^b Boston University School of Medicine, Boston, MA, United States
^c University of Rochester
^d Hogan Assessment Systems, United States
^e Washington University in St. Louis, St. Louis, MO, United States

Abstract
Black Americans have vastly increased odds and earlier onsets of stress- and age-related disease compared to White Americans. However, what contributes to these racial health disparities remains poorly understood. Using a sample of 1577 older adults (32.7% Black; ages 55–65 at baseline), we examined whether stress, health behaviors, social isolation, and inflammation are associated with racial disparities in self-reported physical health. A latent cumulative stress factor and unique stress-domain specific factors were modeled by applying bifactor confirmatory analysis to assessments across the lifespan (i.e., childhood maltreatment, trauma exposure, discrimination, stressful life events, and indices of socioeconomic status). Physical health, health behavior, and social isolation were assessed using self-report. Interleukin-6 (IL-6) and C-reactive protein (CRP) were assayed from morning fasting serum samples; a z-scored inflammation index was formed across these 2 cytokines. A parallel serial mediational model tested whether race (i.e., Black/White) is indirectly associated with health through the following 3 independent pathways: (1) cumulative stress to preventative health behaviors (e.g., healthy eating) to inflammation, (2) cumulative stress to risky health behaviors (e.g., substance use) to inflammation; and (3) cumulative stress to social isolation to inflammation. There were significant indirect effects between race and self-reported physical health through cumulative stress, preventative health behaviors, and inflammation (B = −0.02, 95% CI: −0.05, −0.01). Specifically, Black Americans were exposed to greater cumulative stress, which was associated with reduced engagement in preventative health behaviors, which was, in turn, associated with greater inflammation and reduced physical health. A unique SES factor also indirectly linked race to physical health through preventative health behaviors. Cumulative stress exposure and unique aspects of socioeconomic status are indirectly associated with Black-White racial health disparities through behavioral (i.e., preventative health behavior) and biological (i.e., inflammation) factors. Culturally responsive evidence-based interventions that enhance engagement in preventative health behaviors are needed to directly confront health disparities. Ultimately, large scale anti-racist public policies that reduce cumulative stress burden (e.g., a living wage, universal healthcare) may best attenuate racial health disparities. © 2021 Elsevier Ltd

Author Keywords
Disparities;  Health;  Inflammation;  Race;  Stress

Funding details
R01-AG045231, R01-AG052564, R01-AG061162, R01-DA046224, R01-HD083614, R21-AA027827, R34-DA050272, R56-AG059265
Klingenstein Third Generation FoundationKTGF

Document Type: Article
Publication Stage: Final
Source: Scopus

Acute Trem2 reduction triggers increased microglial phagocytosis, slowing amyloid deposition in mice
(2021) Proceedings of the National Academy of Sciences of the United States of America, 118 (27), art. no. e2100356118, . 

Schoch, K.M.^a , Ezerskiy, L.A.^a , Morhaus, M.M.^a , Bannon, R.N.^a , Sauerbeck, A.D.^a , Shabsovich, M.^a , Jafar-Nejad, P.^b , Rigo, F.^b , Miller, T.M.^a

^a Department of Neurology, Hope Center for Neurological Disorders, Washington University in St. Louis, St. Louis, MO 63110, United States
^b Ionis Pharmaceuticals, Carlsbad, CA 92010, United States

Abstract
Heterozygous genetic variants within the TREM2 gene show a strong association with increased Alzheimer’s disease (AD) risk. Amyloid beta-depositing mouse models haploinsufficient or null for Trem2 have identified important relationships among TREM2, microglia, and AD pathology; however, results are challenging to interpret in the context of varying microglial phenotypes and disease progression. We hypothesized that acute Trem2 reduction may alter amyloid pathology and microglial responses independent of genetic Trem2 deletion in mouse models. We developed antisense oligonucleotides (ASOs) that potently but transiently lower Trem2 messenger RNA throughout the brain and administered them to APP/PS1 mice at varying stages of plaque pathology. Late-stage ASO-mediated Trem2 knockdown significantly reduced plaque deposition and attenuated microglial association around plaque deposits when evaluated 1 mo after ASO injection. Changes in microglial gene signatures 1 wk after ASO administration and phagocytosis measured in ASO-treated cells together indicate that microglia may be activated with short-term Trem2 reduction. These results suggest a time- and/or dose-dependent role for TREM2 in mediating plaque deposition and microglial responses in which loss of TREM2 function may be beneficial for microglial activation and plaque removal in an acute context. © 2021 National Academy of Sciences. All rights reserved.

Author Keywords
Alzheimer’s disease;  Amyloid;  Antisense oligonucleotide;  Microglia;  Trem2

Funding details
CDI-CORE-2015-505, CDI-CORE-2019-813
National Institutes of HealthNIH
National Institute of Neurological Disorders and StrokeNINDSR01NS078398
Washington University in St. LouisWUSTL
Foundation for Barnes-Jewish Hospital3770, 4642
University of WashingtonUWS10 RR027552
Hope Center for Neurological Disorders

Document Type: Article
Publication Stage: Final
Source: Scopus

The Stroke Neuro-Imaging Phenotype Repository: An Open Data Science Platform for Stroke Research
(2021) Frontiers in Neuroinformatics, 15, art. no. 597708, . 

Mohammadian Foroushani, H.^a , Dhar, R.^b , Chen, Y.^c , Gurney, J.^d , Hamzehloo, A.^b , Lee, J.-M.^c , Marcus, D.S.^d

^a Department of Electrical and System Engineering, School of Engineering, Washington University in St. Louis, St. Louis, MO, United States
^b Division of Neurocritical Care, Department of Neurology, Washington University School of Medicine, St. Louis, MO, United States
^c Division of Cerebrovascular Disease, Department of Neurology, Washington University School of Medicine, St. Louis, MO, United States
^d Department of Radiology, Washington University School of Medicine, St. Louis, MO, United States

Abstract
Stroke is one of the leading causes of death and disability worldwide. Reducing this disease burden through drug discovery and evaluation of stroke patient outcomes requires broader characterization of stroke pathophysiology, yet the underlying biologic and genetic factors contributing to outcomes are largely unknown. Remedying this critical knowledge gap requires deeper phenotyping, including large-scale integration of demographic, clinical, genomic, and imaging features. Such big data approaches will be facilitated by developing and running processing pipelines to extract stroke-related phenotypes at large scale. Millions of stroke patients undergo routine brain imaging each year, capturing a rich set of data on stroke-related injury and outcomes. The Stroke Neuroimaging Phenotype Repository (SNIPR) was developed as a multi-center centralized imaging repository of clinical computed tomography (CT) and magnetic resonance imaging (MRI) scans from stroke patients worldwide, based on the open source XNAT imaging informatics platform. The aims of this repository are to: (i) store, manage, process, and facilitate sharing of high-value stroke imaging data sets, (ii) implement containerized automated computational methods to extract image characteristics and disease-specific features from contributed images, (iii) facilitate integration of imaging, genomic, and clinical data to perform large-scale analysis of complications after stroke; and (iv) develop SNIPR as a collaborative platform aimed at both data scientists and clinical investigators. Currently, SNIPR hosts research projects encompassing ischemic and hemorrhagic stroke, with data from 2,246 subjects, and 6,149 imaging sessions from Washington University’s clinical image archive as well as contributions from collaborators in different countries, including Finland, Poland, and Spain. Moreover, we have extended the XNAT data model to include relevant clinical features, including subject demographics, stroke severity (NIH Stroke Scale), stroke subtype (using TOAST classification), and outcome [modified Rankin Scale (mRS)]. Image processing pipelines are deployed on SNIPR using containerized modules, which facilitate replicability at a large scale. The first such pipeline identifies axial brain CT scans from DICOM header data and image data using a meta deep learning scan classifier, registers serial scans to an atlas, segments tissue compartments, and calculates CSF volume. The resulting volume can be used to quantify the progression of cerebral edema after ischemic stroke. SNIPR thus enables the development and validation of pipelines to automatically extract imaging phenotypes and couple them with clinical data with the overarching aim of enabling a broad understanding of stroke progression and outcomes. © Copyright © 2021 Mohammadian Foroushani, Dhar, Chen, Gurney, Hamzehloo, Lee and Marcus.

Author Keywords
big data;  containerized pipeline;  deep learning;  informatics;  phenotype repository;  stroke neuroimaging;  XNAT

Funding details
National Institute of Neurological Disorders and StrokeNINDSK23 NS099440, P30 NS098577, R01 EB009352, R01 NS085419

Document Type: Article
Publication Stage: Final
Source: Scopus

Does raising the arms modify head tremor severity in cervical dystonia?
(2021) Tremor and Other Hyperkinetic Movements, 11 (1), art. no. 21, . 

Cisneros, E.^a , Vu, J.P.^a , Lee, H.Y.^a , Chen, Q.^a , Benadof, C.N.^a , Zhang, Z.^a , Pettitt, E.A.^a , Joshi, S.K.^a , Barbano, R.L.^b , Jankovic, J.^c , Jinnah, H.A.^d , Perlmutter, J.S.^e f , Berman, B.D.^g , Mahajan, A.^h , Goetz, C.G.^h , Stebbins, G.T.^h , Comella, C.L.^h , Peterson, D.A.^a i

^a Institute for Neural Computation, University of California, San Diego, La Jolla, CA, United States
^b Department of Neurology, University of Rochester, Rochester, NY, United States
^c Parkinson’s Disease Center and Movement Disorders Clinic, Department of Neurology, Baylor College of Medicine, Houston, TX, United States
^d Departments of Neurology and Human Genetics, Emory University, Atlanta, GA, United States
^e Department of Neurology, Washington University School of Medicine, St. Louis, MO, United States
^f Departments of Radiology, Neuroscience, Physical Therapy, and Occupational Therapy, Washington University School of Medicine, St. Louis, MO, United States
^g Department of Neurology, Virginia Commonwealth University, Richmond, VA, United States
^h Department of Neurological Sciences, Rush University Medical Center, Chicago, IL, United States
ⁱ Computational Neurobiology Laboratory, Salk Institute for Biological Studies, La Jolla, CA, United States

Abstract
Background: A defining characteristic of dystonia is its position-dependence. In cervical dystonia (CD), sensory tricks ameliorate head tremor (HT). But it remains unknown whether raising the arms alone has the same impact. Methods: We analyzed data collected from patients enrolled by the Dystonia Coalition. For 120 patients with HT, we assessed how raising their arms without touching their head changed their HT severity. Results: Forty-eight out of 120 patients exhibited changes in HT severity when raising their arms. These patients were more likely to exhibit decreases in HT severity (N = 35) than increases (N = 13, χ2 (1, N = 48) = 10.1, p = 0.002). Demographic factors and sensory trick efficacy were not significant predictors of whether HT severity changed when raising their arms. Discussion: Raising the arms without touching the head is a posture that can reduce HT severity in some CD patients. Our results extend the concept of position-dependent motor symptoms in CD to include the position of the arms. Highlights Head tremor (HT) is a prevalent symptom of cervical dystonia (CD) that can often be disabling. This study demonstrates that raising the arms without touching the head is a posture that can reduce HT severity in some CD patients. Our findings also identify a novel form of position-dependence in CD. © 2021 The Author(s).

Author Keywords
Cervical dystonia;  Dystonic tremor;  Head tremor;  Posture

Funding details
U54 TR001456
U54 NS065701, U54 NS116025
W81XWH-17-1-0393
U.S. Department of DefenseDOD
Dystonia Coalition
Rare Diseases Clinical Research NetworkRDCRN

Document Type: Article
Publication Stage: Final
Source: Scopus

Long Acellular Nerve Allografts Cap Transected Nerve to Arrest Axon Regeneration and Alter Upstream Gene Expression in a Rat Neuroma Model
(2021) Plastic and Reconstructive Surgery, pp. 32E-41E. 

Pan, D., Bichanich, M., Wood, I.S., Hunter, D.A., Tintle, S.M., Davis, T.A., Wood, M.D., Moore, A.M.

From the Division of Plastic and Reconstructive Surgery Washington University School of Medicine; the Department of Plastic and Reconstructive Surgery Medical College of Wisconsin Affiliated Hospitals; Orthopaedics and the Department of Surgery Uniformed Services University of the Health Sciences-Walter Reed National Military Medical Center; and the Department of Plastic and Reconstructive Surgery The Ohio State University Wexner Medical Center.

Abstract
Background: Treatments to manage painful neuroma are needed. An operative strategy that isolates and controls chaotic axonal growth could prevent neuroma. Using long acellular nerve allograft to “cap” damaged nerve could control axonal regeneration and, in turn, regulate upstream gene expression patterns. Methods: Rat sciatic nerve was transected, and the distal nerve end was reversed and ligated to generate a model end-neuroma. Three groups were used to assess their effects immediately following this nerve injury: no treatment (control), traction neurectomy, or 5-cm acellular nerve allograft cap attached to the proximal nerve. Regeneration of axons from the injured nerve was assessed over 5 months and paired with concurrent measurements of gene expression from upstream affected dorsal root ganglia. Results: Both control and traction neurectomy groups demonstrated uncontrolled axon regeneration revealed using Thy1-GFP rat axon imaging and histomorphometric measures of regenerated axons within the most terminal region of regenerated tissue. The acellular nerve allograft group arrested axons within the acellular nerve allograft, where no axons reached the most terminal region even after 5 months. At 5 months, gene expression associated with regeneration and pain sensitization, including Bdnf, cfos, and Gal, was decreased within dorsal root ganglia obtained from the acellular nerve allograft group compared to control or traction neurectomy group dorsal root ganglia. Conclusions: Long acellular nerve allografts to cap a severed nerve arrested axon regeneration within the acellular nerve allograft. This growth arrest corresponded with changes in regenerative and pain-related genes upstream. Acellular nerve allografts may be useful for surgical intervention of neuroma. © 2021 Lippincott Williams and Wilkins. All rights reserved.

Document Type: Article
Publication Stage: Article in Press
Source: Scopus