Home > Research > Research Areas  
Active Noise Control for MRI
Recently, MRI(Magnetic Resonance Imaging) scanner is continually used for medical diagnosis and many biomedical researches. When it operates, however, intense noise is generated. The SPL (Sound Pressure Level) of the noise approaches 130 dB especially in 3T MRI. It is expected that the SPL would increase in the future MRI because the strength of magnetic field will increase for higher-resolution image. Ití»s a safety risk because a patient is exposed to intense noise for over 20 minutes and it could cause temporary hearing loss to the patient in extreme case. Ear plugs or ear protector have been used to reduce the noise. But, ití»s effective only for reducing high-frequency noises. For this reason, active noise control systems have been researched for reducing low-frequency noises in which MRI noise has a high SPL. The previous systems mainly have used feedback FXLMS (Filtered-X Least Mean Square) algorithm which is able to reduce the noise at target point inside bore. However, it is able to control only narrowband noises and possible to diverge in severe case.

We determined the property of MRI noise signal measured in MRI. The signal has quasi-periodicity. And the main frequency peaks in frequency domain are multiple of base frequency. Using this property, we are studying about adaptive active noise control algorithm for reducing the noise. This algorithm is focused on the main frequency peaks of MRI noise, and able to reduce the noise effectively.