Finger motion sensors for fMRI motor studies

The kinematics of motor task performance affect brain activity. However, few functional magnetic resonance imaging (fMRI) motor studies have accounted for on-line kinematics because there are currently few MRI-compatible devices to record motor performance. We built a device based on Micro-Electro-Mechanical System (MEMS) gyroscopes that measures the angular velocity of one segment of each of the 10 fingers while a subject performs a finger motor task during fMRI. Finger position, acceleration, and jerk were computed from the angular velocity measurements. The signal-to-noise ratio (SNR) of the MEMS sensors (range: 27.10-34.36 dB) allowed for clear detection of velocity of finger motion during fMRI motor task performance, and showed good stability over time. We demonstrate that use of the MEMS-based device, while negligibly increasing radiofrequency (RF) noise in the scanning environment, did not cause MR image artifacts nor alter fMRI statistical activation maps. Further, we show that signal from the MEMS sensors was not affected by the high static magnetic field (3 T). Increasing the RF power transmitted during fMRI by using a body coil, as compared to a head coil, decreased the sensor's SNR from 30.7 to 24.2 dB, though this loss in SNR did not interfere with the ability to measure velocity of finger motion. We demonstrate the utility of the MEMS-based device in fMRI motor studies through two experiments that examined the relationship between finger movement kinematics and fMRI activation in the healthy and injured brain. On-line acquisition of motor performance during fMRI, through the use of the MEMS-based device, promises to allow for a more detailed understanding of the relationship between movement kinematics and activation in the healthy and injured brain.