Motion correction and compensation strategies for fetal Magnetic Resonance Imaging

Abstract

Magnetic Resonance Imaging (MRI) is one of the imaging modalities that are capable of providing a rich range of soft tissue contrast. MRI is also non-invasive and non-ionizing in nature. Longer scan/acquisition time in the MRI makes it more prone to motion artifacts. This enables motion detection, correction, and compensation as one of the well-established research domain in the field of MRI. The vast range of pulse sequences, reconstruction techniques, and image processing algorithms provide solutions to minimize the artifacts appeared in the images. The motion artifacts that appear in the images are as a result of a complex combination of patient/subject motion, anatomy, pulse sequences, reconstruction technique utilized. As many numerous factors govern the appearance of motion artifacts in the image, it makes it harder for generalizing the solution. Thus the general strategy employed for addressing motion is by tailoring techniques to a specific requirement. Motion compensation/correction can be projected as a complex problem having many partial solutions. Fetal MRI is employed as an alternative imaging modality for ultrasound imaging for validating abnormalities. Typical motion is caused by the subject/patient and/or fetus. The approaches practiced for reducing motion are sedation, fast imaging, reconstruction or post-processing. The work carried out follows the image processing based approach. The work compares the generic algorithms that might be more suitable for fetal MR image post-processing.The image processing based approach was integrated with the optical-based sensor for tracking of the patient movement. The anatomy considered for study is MSK in case of humans. The displacement measured by the sensor was utilized as the aprior information for motion correction. Measured displacement was utilized for motion correction in the frequency domain.