Combining a deformable model and a probabilistic framework for an automatic 3D segmentation of prostate on MRI

Area of Science:

• Medical Imaging
• Computational Anatomy
• Radiology

Background:

• Accurate prostate segmentation in MRI is vital for cancer diagnosis and therapy.
• Current methods often require manual expert correction, limiting efficiency.
• There is a need for automated segmentation techniques.

Purpose of the Study:

• To introduce a novel automatic 3D segmentation method for the prostate gland using MRI.
• To evaluate the accuracy and efficiency of this automated method.
• To reduce reliance on manual segmentation in clinical practice.

Main Methods:

• Utilized a statistical shape model for prior knowledge.
• Modeled gray level distribution using Gaussian mixture fitting.
• Employed Markov fields for contextual information and iterative conditional mode for optimization.

Main Results:

• The automated method achieved a mean Hausdorff distance of 9.94 mm and an overlap ratio of 0.83.
• Performance was validated against expert radiologist delineations.
• The method demonstrated accuracy even at the prostate's base and apex.

Conclusions:

• The developed automatic prostate MRI segmentation method yields satisfactory results.
• The approach is computationally feasible and efficient.
• This automated technique shows promise for improving prostate cancer diagnosis and treatment planning.