Anatomical adjustments in brain positron emission tomography using CT images

Area of Science:

• Medical Imaging
• Radiology
• Nuclear Medicine

Background:

• Accurate image registration between different modalities like X-ray CT and PET is crucial for comprehensive patient diagnosis.
• Current methods may lack precision or ease of application across various imaging devices.
• Overlaying functional PET data with anatomical CT data requires precise spatial alignment.

Purpose of the Study:

• To develop and validate a novel method for automatic alignment of regions of interest between X-ray CT and PET images.
• To enhance the accuracy of anatomical localization by precisely adjusting PET and XCT image slices.
• To provide a universally applicable technique for PET/XCT image registration.

Main Methods:

• Utilized a face mask with defined landmarks during both PET and XCT scans for slice position matching.
• Employed a video system in conjunction with facial landmarks for accurate anatomical adjustment between different image slices.
• Developed an automated process for aligning regions of interest based on landmark data.

Main Results:

• Achieved accurate slice position matching between PET and XCT images, with average misadjustments below 2 mm axially and 1.0 degree angularly.
• Demonstrated the method's effectiveness in aligning anatomical locations across different imaging slices.
• Validated the technique through clinical application, confirming its reliability.

Conclusions:

• The developed method offers precise and automatic alignment of XCT and PET images using facial landmarks.
• This technique minimizes positional misadjustments, enhancing the accuracy of combined imaging analysis.
• The method is easily applicable to any PET or XCT imaging device, offering broad clinical utility.