An automated method for the normalization of scintigraphic images

Area of Science:

• Medical Imaging
• Computer Vision
• Nuclear Medicine

Background:

• Accurate comparison of organ images requires precise normalization.
• Existing normalization methods can be sensitive to structural variations in medical scans.

Purpose of the Study:

• To introduce a novel, automated image normalization technique.
• To create a method insensitive to structural modifications in organ images.

Main Methods:

• An iterative, automated normalization approach was developed.
• Normalization factor determined by maximizing sign changes in scanned subtraction images.
• A unidimensional optimization method was employed for calculation.

Main Results:

• The proposed method demonstrates high insensitivity to structural image modifications.
• The algorithm is computationally efficient and cost-effective.
• Successful implementation is feasible on standard nuclear medicine data-processing systems.

Conclusions:

• The developed automated normalization method offers a robust solution for comparative organ imaging.
• Its efficiency and low computational cost make it suitable for widespread adoption in nuclear medicine.