A non-rigid registration method for serial lower extremity hybrid SPECT/CT imaging.

Area of Science:

• Biomedical Imaging
• Medical Physics
• Animal Models

Background:

• Serial microCT imaging aids in evaluating hindlimb ischemia in mouse models.
• Non-rigid image registration is crucial for quantifying changes in perfusion and molecular markers.
• Existing registration methods struggle with large displacements in whole-body imaging.

Purpose of the Study:

• To develop an improved non-rigid registration method for serial microCT imaging in mice.
• To enhance the quantification of disease progression and therapeutic interventions in hindlimb ischemia models.

Main Methods:

• A novel method based on the extended demons algorithm was developed.
• The method utilizes a level-set representation of mouse body contour and skeletal structure.
• The registration accounts for bone movement as a framework and skin movement as a constraint.

Main Results:

• The proposed method demonstrated improved performance in registering serial microCT mouse images.
• Successfully applied to hybrid microCT-SPECT image quantification.
• Outperformed existing registration techniques in handling anatomical variations.

Conclusions:

• The new serial registration method effectively handles large anatomical deformations in mouse imaging.
• This advancement improves the quantitative analysis of hindlimb ischemia and therapeutic responses.
• The method offers a robust solution for serial whole-body imaging registration challenges.