Abstract:

BACKGROUND: Current magnetic resonance (MR) imaging has gray bias field, which affects digital processing of medical images, such as segmentation, registration, and quantization. Correction of this adverse space density changes is a necessary process before imaging analysis.
OBJECTIVE: To propose a model independent correction method of complex image signal bias field based on image data.
METHODS: Data were provided by the Department of Radiology, Fuxing Hospital of Capital Medical University, and collected by MR device (GE, USA). All MR images were provided by the volunteers (imaging visual field 250 mm×250 mm, matrix 256×256, slice thickness 1 mm). Based on entropy method and Mask method, MR image bias field was corrected. 
RESULTS AND CONCLUSION: This method can correct MR image bias field accurately and effectively. These results provided a new method of MR image bias field correction for MR image analysis, which improved the image analysis accuracy and robustness.