Adaptive Spectral Envelope Estimation for Doppler Ultrasound.

Area of Science:

• Medical Imaging
• Biomedical Engineering
• Fluid Dynamics

Background:

• Accurate estimation of maximum velocities and spectral envelopes in ultrasound Doppler blood flow spectrograms is crucial for clinical diagnosis.
• Spectral broadening and power spectrum variance present challenges in obtaining precise maximum velocity measurements.

Purpose of the Study:

• To develop and validate a novel method for maximum velocity point detection in ultrasound Doppler spectrograms.
• To improve the accuracy and reliability of spectral envelope estimation for enhanced blood flow analysis.

Main Methods:

• A hybrid method combining signal noise slope intersection and geometric method principles was developed for maximum velocity detection.
• Adaptive noise estimation was employed to ensure a smooth spectral envelope post-detection.
• The method was validated using simulated Doppler signals, phantom recordings (steady and pulsatile), and in vivo data from various arteries.

Main Results:

• Simulation experiments showed a bias of less than 2.5% in maximum velocity estimations across various conditions (peak velocities, Doppler angles, SNR).
• Low standard deviation (<2%) in spectral envelope estimation was achieved for simulated, phantom, and in vivo data, even with varying flow conditions and SNR.
• The method demonstrated potential for automated blood flow measurements, exemplified by pulsatility index estimation in uterine and umbilical arteries.

Conclusions:

• The proposed method effectively estimates maximum velocities and spectral envelopes in ultrasound Doppler spectrograms.
• The achieved accuracy and low variability support the use of this method for automated clinical blood flow measurements.
• This advancement has significant implications for improving diagnostic capabilities in vascular ultrasound.