Compliant finger sensor for sensorimotor studies in MEG and MR environment

Abstract

Magnetoencephalography (MEG) and functional magnetic resonance imaging (fMRI) are widely used for functional brain imaging. The correlations between the sensorimotor functions of the hand and brain activities have been investigated in MEG/fMRI studies. Currently, limited information can be drawn from these studies due to the limitations of existing motion sensors that are used to detect hand movements. One major challenge in designing these motion sensors is to limit the signal interference between the motion sensors and the MEG/fMRI. In this work, a novel finger motion sensor, which contains low-ferromagnetic and non-conductive materials, is introduced. The finger sensor consists of four air-filled chambers. When compressed by finger(s), the pressure change in the chambers can be detected by the electronics of the finger sensor. Our study has validated that the interference between the finger sensor and an MEG is negligible. Also, by applying a support vector machine algorithm to the data obtained from the finger sensor, at least 11 finger patterns can be discriminated. Comparing to the use of traditional electromyography (EMG) in detecting finger motion, our proposed finger motion sensor is not only MEG/fMRI compatible, it is also easy to use. As the signals acquired from the sensor have a higher SNR than that of the EMG, no complex algorithms are required to detect different finger movement patterns. Future studies can utilize this motion sensor to investigate brain activations during different finger motions and correlate the activations with the sensory and motor functions respectively.