Automatic detection of motion artifacts in infant functional optical topography studies.

ABSTRACT

A key advantage of optical topography (OT) for cognitive neurodevelopmental studies over other imaging techniques such as magnetic resonance (MRI) or positron emission tomography (PET) is that it is less intrusive in the experimental set up. This is, in part, because in OT there is no need to impose strict movement constraints on the participants. However, large head movements can cause signal disruption, thus there is a need to (i) detect artifacts caused by movement and (ii) implement strategies to remove the noise component from the optical data. We have developed a motion sensor compatible with our in-house OT system and suitable for infant studies. With the data collected we have adjusted the thresholds that were used in the past to automatically discard data affected by movement artifacts. We have also compared the performance of two different head probe designs by measuring the amount of signal disruption present in recordings with each design. Finally, we have studied the feasibility of using the sensor data as external input of an adaptive filter to reduce the movement component of the optical data.