Multidimensional Clutter Filtering of Aperture Domain Data for Improved Blood Flow Sensitivity

Area of Science:

• Medical Imaging
• Signal Processing
• Ultrasound Technology

Background:

• Singular Value Decomposition (SVD) is a standard technique for clutter rejection in power Doppler imaging.
• Conventional SVD methods applied to Casorati matrices offer filtering based on spatial or temporal data characteristics.
• Limitations exist in conventional SVD, particularly its robustness with short ensemble lengths.

Purpose of the Study:

• To introduce a novel clutter filtering method utilizing Higher Order Singular Value Decomposition (HOSVD).
• To apply HOSVD to a tensor of aperture data for enhanced clutter rejection in power Doppler imaging.
• To demonstrate the improved sensitivity and robustness of the HOSVD method compared to conventional SVD.

Main Methods:

• Implementation of Higher Order Singular Value Decomposition (HOSVD) on a tensor of aperture data (e.g., delayed channel data).
• Leveraging temporal, spatial, and aperture domain features for multidimensional filtering.
• Validation using Field II simulations and in vivo ultrasound data.

Main Results:

• The HOSVD method demonstrates improved sensitivity towards detecting blood flow.
• HOSVD shows greater robustness compared to conventional SVD when dealing with short ensemble lengths.
• Multidimensional filtering approach enhances clutter rejection capabilities.

Conclusions:

• HOSVD applied to aperture data offers a superior clutter filtering technique for power Doppler imaging.
• The proposed method enhances blood flow detection sensitivity and maintains robustness with limited data.
• This advanced SVD technique represents a significant advancement in ultrasound signal processing for improved diagnostic accuracy.