MRI-Compatible Soft Robotic Sensing Pad for Head Motion Detection

Mishek Musa1, Saikat Sengupta2, Yue Chen3

  • 1University of Arkansas
  • 2Vanderbilt University Medical Center
  • 3Georgia Institute of Technology

Details

16:20 - 16:25 | Thu 26 May | Room 122A | ThB15.09

Session: Soft Sensors and Actuators

Abstract

Head motion remains a prevalent challenge to overcome when performing MRI scans, as it results in severe degradation of the image quality. This paper presents the design, fabrication, and evaluation of a pneumatic based MRI-compatible soft sensing pad to be used for head motion detection. A modal based calibration method is first used to estimate the weight of the head on the sensing pad. A calibration error of 0.12±0.06 kg was observed. Gaussian process regression is then employed to establish a mapping from the weight of the head and the internal pressures of the sensing pad to the head motion based on experimental data. The testing result of the regression indicates that the sensing pad is capable of predicting the angle of the head with an RMSE of 1.35±0.99 deg over a range of ±40 deg. Over the range observed in clinical practice, ±10 deg, the sensing pad is able to predict the head angle with an RMSE of 0.88±0.54 deg. The sensing pad has been deemed MRI safe and no significant SNR variation was observed during MRI compatibility experiments.