Inpatient Stroke Recovery Using Sensors

OBJECTIVES:

Aim 1: Assess the feasibility of continuous long-term monitoring of inpatients with stroke using wearable sensors.

The investigators will obtain quantitative health data from research-grade, wireless, wearable sensors on individuals with subacute and chronic stroke in the clinical setting, as well as healthy controls.

The investigators will specifically check for variability in device data, as well as consistency and periodicity of sensor readings across the clinical study period. The investigators will analyze test-retest reliability and inter-rater reliability of using the wearable sensor technology for clinical and monitoring applications. Furthermore, The investigators will determine whether the sensors can distinguish biometric and activity characteristics between healthy controls and individuals with stroke.

Aim 2: Quantify upper and lower extremity movement impairments, mobility-related activities, speech and swallowing activities, and clinical parameters during stroke recovery.

The investigators will obtain continuous biometric and movement-based sensor data for clinical symptoms (e.g., muscle activation, heart rate variability, talk time, and gait quality) during the performance of validated clinical tests and during general inpatient activities (e.g., therapy, eating, and sleeping).

The investigators will compare device data with clinically validated measures of movement and language function, such as the Modified Ashworth Scale or Western Aphasia Battery. The investigators will describe variation of device data in subgroups of subjects defined by clinician assessed clinically validated measures (10-Meter Walk Test, Mini-Mental Status Exam, etc.). The investigators will also assess the ability of the sensors to capture response to treatment, such as movement therapy, speech therapy, medication, and Botox by comparing sensor data before and after treatment. The investigators will provide evidence about the degree to which the measured variables are intercorrelated. Lastly, The investigators will evaluate and compare the state of recovery between patients at time of discharge using sensor-based outcomes. Due to heterogeneity of clinical symptoms after stroke, as differing etiologies and degrees of recovery result in different types and levels of gait impairments, preliminary analyses may be performed on specific sub-datasets to determine which predictors, data types, and other data compositions affect algorithm performance before their application and evaluation on the dataset collected at the end of this study. Sub-datasets may be sampled based on (but not limited to) the population characteristics, the sensors used, the data available at the time analysis.

• Change in 6-Minute Walk Test (6MWT) [Time frame: From Admission time-point to 1-2 weeks, 1 month, 3 months, 6 months, 12 months post-stroke]
• Change in 10-Meter Walk Test (10MWT) [Time frame: From Admission time-point to 1-2 weeks, 1 month, 3 months, 6 months, 12 months post-stroke]
• Change in Berg Balance Scale (BBS) [Time frame: From Admission time-point to 1-2 weeks, 1 month, 3 months, 6 months, 12 months post-stroke]
• Change in Timed Up and Go (TUG) [Time frame: From Admission time-point to 1-2 weeks, 1 month, 3 months, 6 months, 12 months post-stroke]
• Change in Functional Gait Assessment (FGA) [Time frame: From Admission time-point to 1-2 weeks, 1 month, 3 months, 6 months, 12 months post-stroke]
• Change in Gait Analysis [Time frame: From Admission time-point to 1-2 weeks, 1 month, 3 months, 6 months, 12 months post-stroke]
• Three months clinical score estimation [Time frame: 3 months after the stroke event]
• Six months clinical score estimation [Time frame: 6 months after the stroke event]
• Twelve months clinical score estimation [Time frame: 12 months after the stroke event]