X-ray coronary angiography background subtraction by adaptive weighted total variation regularized online RPCA.

Area of Science:

• Medical Imaging
• Cardiovascular Diagnostics
• Image Processing

Background:

• X-ray coronary angiograms (XCA) are crucial for diagnosing and treating cardiovascular diseases.
• Low contrast in XCA images, due to motion and limited contrast agent dose, hinders accurate diagnosis.
• Existing background subtraction methods aim to improve vessel visibility and reduce contrast agent requirements.

Purpose of the Study:

• To propose an adaptive weighted total variation regularized online RPCA (WTV-ORPCA) method for background subtraction in XCA sequences.
• To enhance the visibility and contrast of coronary vessels in low-contrast XCA images.
• To reduce the reliance on high doses of contrast agents.

Main Methods:

• The WTV-ORPCA method utilizes low-rank and sparse subspace decomposition for background subtraction.
• Initial preprocessing involves morphological operators to remove large structures and homogenize images.
• An adaptive weighted TV constraint is applied to the foreground subspace for spatial coherency of extracted vessels.

Main Results:

• Experiments on clinical and synthetic low-contrast XCA datasets demonstrated the method's effectiveness.
• The proposed method achieved superior performance compared to six state-of-the-art techniques.
• Key metrics on clinical data: global CNR (5.976), local CNR (3.173), SSIM (0.987), RE (0.026); on synthetic data: CNR (4.851), local CNR (2.942), SSIM (0.958), RE (0.034).

Conclusions:

• The WTV-ORPCA method significantly improves coronary vessel contrast and visibility in XCA images.
• The technique effectively preserves vessel structure integrity while minimizing reconstruction errors.
• The method offers a computationally efficient solution for enhancing XCA sequences.