How Accurate Are Smartphone Accelerometers to Identify Intermittent Claudication?

Claudication is a cramping pain that is worsened by walking and relieved with rest. It is caused by inadequate blood flow to the leg muscles because of atherosclerosis. Recently, smartphones and their sensors have been proposed in the context of mobile health to monitor gait. However, their use remains disputed: objections concern the quality of the collected data. Therefore, the work presented in this paper proposes to study three main sources of noise observed in smartphone accelerometers and to objectively assess their impact on claudication detection. To do so, we first observe three noise sources in four different smartphones to get an idea of their ranges; we second compare the smartphones’ signals to a ground truth from a vision-based system and third propose to detect claudication by estimating duty cycle from the vertical accelerometer signal and to evaluate the impact of the three noise sources on this basis.