2022
Open-source dataset reveals relationship between walking bout duration and fall risk classification performance in persons with multiple sclerosis
Meyer B, Tulipani L, Gurchiek R, Allen D, Solomon A, Cheney N, McGinnis R. Open-source dataset reveals relationship between walking bout duration and fall risk classification performance in persons with multiple sclerosis. PLOS Digital Health 2022, 1: e0000120. PMID: 36812538, PMCID: PMC9931255, DOI: 10.1371/journal.pdig.0000120.Peer-Reviewed Original ResearchWalking boutsFall riskGait parametersOpen-source datasetsNon-fallersFeature-based modelWalking bout durationClassification performanceWearable sensorsFall risk classificationDaily activity performancePatient-reported surveysBiannual clinical visitsFall risk estimationBout durationDeep learning modelsFall historyShort boutsClinic visitsInvestigate fall riskWalking dataRisk estimatesAssociated with morbiditySensor dataDeep learning
2021
Wearables and Deep Learning Classify Fall Risk From Gait in Multiple Sclerosis
Meyer B, Tulipani L, Gurchiek R, Allen D, Adamowicz L, Larie D, Solomon A, Cheney N, McGinnis R. Wearables and Deep Learning Classify Fall Risk From Gait in Multiple Sclerosis. IEEE Journal Of Biomedical And Health Informatics 2021, 25: 1824-1831. PMID: 32946403, PMCID: PMC8221405, DOI: 10.1109/jbhi.2020.3025049.Peer-Reviewed Original ResearchConceptsWearable sensorsPreventive interventionsBidirectional long short-termInexpensive wearable sensorsDeep neural networksWearable sensor dataFall prevention interventionsSpatiotemporal gait parametersFall risk assessmentLong-Short-TermMachine learning modelsGait biomechanicsGait parametersSensor dataFall riskNeural networkHealthcare providersStatistical featuresLearning modelsPatient reportsAccelerometer dataIdentified measuresMultiple sclerosisWearableGood performance