Motion-Based Segmentation of Medical MRI Images: A Review and Analysis
Author
Adhikari, Anil
Abstract
Image data produced in medical field assists doctors, surgeons and researchers to study muscle and tissue structure to identify and diagnose patient specific diseases and deficiency. Due to the complexity of muscle organization and structure, the study and even the treatment becomes cumbersome. So, various image processing and segmentation mechanisms have been evolving in literature to provide effective interpretation of the image data. Most of the image segmentation algorithms and methods use pixel intensity, texture and shape feature to process image. But in some situation where two independently moving muscles attach together making the boundary not clearly visible which means the intensity of pixels around the boundary is almost same, then those segmentation methods do not work efficiently. So, in this thesis research, we have devised an approach to segment the muscles based on their movement using optical flow estimation technique. We have also evaluated the efficiency of a popular optical flow estimation technique with low-quality and high-quality image datasets. Segmentation and boundary detection results have been provided with the accuracy and performance evaluation.
Subject
Date
2018-07-20
Citation:
APA:
Adhikari, Anil.
(July 2018).
Motion-Based Segmentation of Medical MRI Images: A Review and Analysis
(Master's Thesis, East Carolina University). Retrieved from the Scholarship.
(http://hdl.handle.net/10342/6951.)
MLA:
Adhikari, Anil.
Motion-Based Segmentation of Medical MRI Images: A Review and Analysis.
Master's Thesis. East Carolina University,
July 2018. The Scholarship.
http://hdl.handle.net/10342/6951.
September 30, 2023.
Chicago:
Adhikari, Anil,
“Motion-Based Segmentation of Medical MRI Images: A Review and Analysis”
(Master's Thesis., East Carolina University,
July 2018).
AMA:
Adhikari, Anil.
Motion-Based Segmentation of Medical MRI Images: A Review and Analysis
[Master's Thesis]. Greenville, NC: East Carolina University;
July 2018.
Collections
Publisher
East Carolina University