Complex threshold method for identifying pixels that contain predominantly noise in magnetic resonance images

Abstract

Purpose: To create a robust means to remove noise pixels using complex data. Materials and Methods:A receiver operating characteristic (ROC) curve was used to determine the appropriate choice of magnitude and phase thresholds as well as connectivity values to determine what pixels represent noise in the image. To fine-tune the results, a spike removal and hole replacement operator is applied to reduce Type I error and remove small islands of noise. Results:The use of phase information improves the magnitude-only thresholding approach by further recognizing pixels that contain only noise. The performance of the method is enhanced using local connectivity of magnitude and phase data. An ROC analysis on simulated data shows that the Type I and Type II errors are less than 10 Ϫ4 and 10 Ϫ3, respectively, without connectivity and 0 and 10 Ϫ3 , respectively, with connectivity for a signal-to-noise ratio (SNR) of 3:1 or higher. Conclusion:The joint use of both magnitude and phase images helps to improve the removal of noise points in magnetic resonance images. This can prove useful in automating the visualization of phase images without the highly distractive phase noise in noise regions. Also, it is useful in susceptibility weighted imaging when taking the minimum intensity projections of variably sized regions.

Imaging the Effects of Oxygen Saturation Changes in Voluntary Apnea and Hyperventilation on Susceptibility-Weighted Imaging

“…This algorithm involves first applying a threshold to remove background noise and structures (ie, skull). 23 A shape-filtering noise removal algorithm was then used to remove falsely identified venous structures from the threshold-generated map. Further filtering of false-positives was performed by removal of clusters of connected voxels below a certain size.…”

Decreased oxygen saturation in asymmetrically prominent cortical veins in patients with cerebral ischemic stroke

“…Thirdly, phase k-space was interpolated by zero filling the phase images to a larger matrix size for the purpose of reducing the aliasing artifacts in the final SM data. Fourthly, skull stripping and complex thresholding were performed to remove unwanted regions of low signal [18]. Fifthly, a regularization threshold of 0.1 was used for the inverse process.…”

“…The processing included the following steps: Firstly, a high resolution SWI gradient echo sequence with velocity compensation in 3 directions was acquired with both the magnitude and phase images being obtained. Secondly, phase images were high-pass filtered using a central 96 × 96 homodyne filter [18]. Thirdly, phase k-space was interpolated by zero filling the phase images to a larger matrix size for the purpose of reducing the aliasing artifacts in the final SM data.…”

Background-Suppressed MR Venography of the Brain Using Magnitude Data: A High-Pass Filtering Approach

“…These techniques include phase unwrapping [14, 15], a regional suppression of phase artifact using local field gradient mapping [17], and the use of a susceptibility model [18]. Signal loss in the peripheral region of the brain was reduced by the use of tissue-air volume segmentation algorithms [17, 19–21]. Despite the effectiveness of these techniques, residual image artifacts often remain in SWI.…”