Inpatient Stroke Recovery Using Sensors

The 6MWT measures the distance a subject can walk indoors on a flat, hard surface in 6 minutes, using assistive devices as necessary. The test is a reliable and valid evaluation of functional exercise capacity and is used as a sub-maximal test of aerobic capacity and endurance. The minimal detectable change for people with sub-acute stroke is 60.98 meters. Wearable sensors are used to monitor the movement of the subject during the test.

The 10MWT measures the amount of time it takes to walk 10 meters. Time will be recorded using a stopwatch and recorded to the one hundredth of a second. The effects of acceleration and deceleration are minimized by adding 1 meter at the beginning and end of the course to isolate the subject's steady state speed. The test will be recorded 3 times each at a normal self-selected pace and at a faster pace, with adequate rest in between. Results will be averaged from 3 trials. Any assistive devices or orthotics should be kept consistent throughout and documented. Wearable sensors are used to monitor the movement of the subject during the test.

The BBS is a 14-item test, scored on a 5-level ordinal scale and validated against length of stay and discharge destination for stroke patients. The total score is expressed as a number between 0 and 56, where lower score mean increased balance impairment. It measures functional balance in a clinical setting during static and dynamic tasks (sitting, standing, transitioning from sit to stand, etc.). Wearable sensors are used to monitor the movement of the subject during the test.

The TUG assesses mobility by measuring the time a person take to rise from a chair, walk 3 meters, turn around, walk back to the chair, and sit down. It can detect longitudinal changes in mobility in stroke patients. The subject wears their routine footwear and orthotics and can use their mobility aids. Wearable sensors are used to monitor the movement of the subject during the test.

The FGA is comprised of 10-item that contains 7 of 8 items (except walking around obstacles) from the Dynamic Gait Index and 3 additional tasks, including walking with a narrow base of support, walking with the eyes closed, and walking backward. Subjects' performance of each test item was rated on a 4-point scale (0-3), with the total score ranging between 0 and 30. Wearable sensors are used to monitor the movement of the subject during the test.

Gait analysis provides a quantitative means of assessing walking function based on spatiotemporal parameters of gait. Subjects walk at both a comfortable and a fast pace over the GaitRite system, an electronic walkway with integrated sensors. Data from GaitRite is reliable and valid for evaluating walking characteristics and provides a gold standard for validating gait parameters from the sensors. Wearable sensors are used to monitor the movement of the subject during the test.

Error between clinical scores estimated from machine learning algorithms trained on sensor data from the Admission time-point, and true scores assessed at three months after the stroke event.

Error between clinical scores estimated from machine learning algorithms trained on sensor data from the Admission time-point, and true scores assessed at six months after the stroke event.

Error between clinical scores estimated from machine learning algorithms trained on sensor data from the Admission time-point, and true scores assessed at twelve months after the stroke event.