Content-Based Estimation of Brain MRI Tilt in Three Orthogonal Directions

Area of Science:

• Medical Imaging
• Neuroimaging
• Image Processing

Background:

• Brain magnetic resonance imaging (MRI) scans can suffer from tilt, leading to distortions and misalignment.
• Manual tilt correction is labor-intensive, costly, and requires specialized anatomical knowledge.
• Accurate image registration is crucial for various medical applications, making tilt correction essential.

Purpose of the Study:

• To develop an automated method for correcting tilt in brain MRI images across three orthogonal directions (X, Y, Z).
• To measure pitch, yaw, and roll angles for precise tilt estimation and correction.
• To enhance the accuracy and robustness of brain MRI image registration.

Main Methods:

• Utilized image processing techniques, principal component analysis, and similarity measures for Z-axis tilt correction.
• Employed morphological operations for X-axis tilt correction.
• Applied orthogonal regression for Y-axis tilt correction, processing both T1- and T2-weighted MR images.

Main Results:

• The automated method achieved a simulation error of 0.40 ± 0.09°.
• Measured tilt angles were 6.2 ± 3.94° (X), 2.35 ± 2.61° (Z), and 5 ± 4.36° (Y).
• The proposed algorithm demonstrated superior accuracy and robustness compared to existing methods.

Conclusions:

• The developed automated approach effectively corrects tilt in brain MRI images.
• This method offers a more accurate and robust solution than manual correction or previous automated techniques.
• Improved image alignment facilitates more reliable medical image registration and analysis.