3-dimensional throat region segmentation from MRI data based on Fourier interpolation and 3-dimensional level set...

Area of Science:

• Medical Imaging
• Image Processing
• Computational Anatomy

Background:

• Accurate segmentation of the throat region in Magnetic Resonance Imaging (MRI) is crucial for various clinical applications.
• Existing segmentation methods may struggle with image artifacts and contrast variations inherent in MRI data.

Purpose of the Study:

• To introduce a new algorithm for robust and accurate 3D segmentation of the throat region from MRI data.
• To improve the delineation of the throat region by addressing preprocessing challenges and utilizing advanced segmentation techniques.

Main Methods:

• The algorithm incorporates preprocessing steps to enhance contrast and reduce artifacts in MRI data.
• Isotropic 3D volume reconstruction is achieved via Fourier interpolation.
• A level set method is employed to evolve a bounding cube, defining a smooth 3D boundary of the throat region.

Main Results:

• The proposed algorithm demonstrates effective segmentation on both real and synthetic MRI datasets.
• Validation confirms the algorithm's robustness and accuracy in delineating the 3D throat region.

Conclusions:

• The developed algorithm offers a reliable solution for 3D throat region segmentation in MRI.
• This technique has the potential to enhance downstream analysis and applications relying on precise anatomical segmentation.