Physical Sensors: Motion Sensors for Physical Activity Assessment

Pedometers and Accelerometers in Multiple Sclerosis: Current and New Applications

“…These devices have further been applied to assess steps taken in MS [ 31 ], and have presented moderate to high accuracies in recording steps when compared to manually counted steps (manually counted vs. Fitbit—ICC = 0.69; manually counted vs. Fitbit One—relative accuracy = 98.1%) [ 15 , 16 ]. One major problem of using these devices in research concerns the lack of transparency by the manufacturers on how the outputs are derived/calculated [ 8 , 32 ]. This turns the metrics from Fitness/Activity Trackers into a blackbox [ 8 , 32 ].…”

“…One major problem of using these devices in research concerns the lack of transparency by the manufacturers on how the outputs are derived/calculated [ 8 , 32 ]. This turns the metrics from Fitness/Activity Trackers into a blackbox [ 8 , 32 ]. In order for these devices to have better acceptance among researchers, manufacturers would need to consider disclosing the proprietary predictive equations/algorithms in the devices.…”

“…Participant burden also relates to the device placement, with most researchers still adopting the hip as the placement of choice. Nevertheless, there is a tendency for a growing number of researchers to adopt wrist-worn accelerometry protocols, reducing participant burden and allowing for data collection over the 24 h cycle [ 32 ]. In addition to changing the accelerometer placement, the improvement of device size and design by manufacturers are a major aspect for reducing participant burden.…”