Sparse Reconstruction of Compressive Sensing Magnetic Resonance Imagery using a Cross Domain Stochastic Fully Connected Conditional Random Field Framework

Sparse reconstrction;Biomedical Imaging;Image processing;Computer Vision

Li, Yaguang

University of Waterloo

关键词: Sparse reconstrction;    Biomedical Imaging;    Image processing;    Computer Vision;   

Others  :  https://uwspace.uwaterloo.ca/bitstream/10012/10577/5/Li_Yaguang.pdf

瑞士|英语

来源: UWSPACE Waterloo Institutional Repository

【 摘 要 】

Prostate cancer is a major health care concern in our society. Early detection of prostatecancer is crucial in the successful treatment of the disease. Many current methods used indetecting prostate cancer can either be inconsistent or invasive and discomforting to thepatient. Magnetic resonance imaging (MRI) has demonstrated its ability as a non-invasiveand non-ionizing medical imaging modality with a lengthy acquisition time that can beused for the early diagnosis of cancer. Speeding up the MRI acquisition process can greatlyincrease the number of early detections for prostate cancer diagnosis.Compressive sensing has exhibited the ability to reduce the imaging time for MRI bysampling a sparse yet sufficient set of measurements. Compressive sensing strategies areusually accompanied by strong reconstruction algorithms. This work presents a comprehensiveframework for a cross-domain stochastically fully connected conditional random field (CD-SFCRF) reconstruction approach to facilitate compressive sensing MRI. Thisapproach takes into account original k-space measurements made by the MRI machinewith neighborhood and spatial consistencies of the image in the spatial domain. Thisapproach facilitates the difference in domain between MRI measurements made in thek-space, and the reconstruction results in spatial domain. An adaptive extension of theCD-SFCRF approach that takes into account regions of interest in the image and changesthe CD-SFCRF neighborhood connectivity based on importance is presented and tested aswell. Finally, a compensated CD-SFCRF approach that takes into account MRI machineimaging apparatus properties to correct for degradations and aberrations from the imageacquisition process is presented and tested.Clinical MRI data were collected from twenty patients with ground truth data examinedand con firmed by an expert radiologist with multiple years of prostate cancer diagnosisexperience. Compressive sensing simulations were performed and the reconstructionresults show the CD-SFCRF and extension frameworks having noticeable improvementsover state of the art methods. Tissue structure and image details are well preserved whilesparse sampling artifacts were reduced and eliminated. Future work on this frameworkinclude extending the current work in multiple ways. Extensions including integration intocomputer aided diagnosis applications as well as improving on the compressive sensingstrategy.