Michelle Lis

Year: Senior

Major: Physics and Molecular Biology

Mentor: Anil Sethi

Department: Radiation Oncology

Project Title: Eliminating Motion Distortions in MRI via Post Processing Filtration

Abstract: Due to long imaging time, patient motion in MRI scans can contain significant motion distortions including ghost-like artifacts, blurring, and reduction of intensity in regions of movement. Such motion can also be present in cranial scans used in treatment planning for tumors. This distortion can provide errors in information on location, volume and shapes of tumors and effect dose and beam configurations in treatment planning. Both pre-processing methods, such as patient immobilization, mapping a motion model and reversing motion patterns during the scan, and post-processing, such as filtration via two dimensional Fourier transforms, and computer coding algorithms, are available to correct for these motion distortions. Post-processing is a promising method, in particular, a MatLab program of MRI images to remove rigid rotational motion and remove ghosting effects. By aligning cranial scans, sharpening the images and erasing ghosting, the resulting tumor volume and maximum dose have both been reduced from 26.8 to 25.4 MU. Correction via coding along with other correction methods could provide for more effective patient treatment.