VASCULAR STENOSIS DETECTION USING TEMPORAL-SPECTRAL DIFFERENCES IN CORRELATED ACOUSTIC MEASUREMENTS

Area of Science:

• Biomedical Engineering
• Medical Imaging
• Acoustics

Background:

• Central venous stenosis is frequently undiagnosed in hemodialysis patients due to imaging limitations.
• Noninvasive, point-of-care detection methods are needed for early diagnosis.

Purpose of the Study:

• To develop and validate a novel method for detecting central venous stenosis using acoustic signals.
• To assess the efficacy of flexible microphone arrays and advanced signal processing in identifying turbulent blood flow.

Main Methods:

• Flexible microphone arrays were used to capture time-correlated blood flow sounds.
• A signal processing approach involving continuous wavelet transform and spectral feature correlation was developed.
• The system was tested on pulsatile flow phantoms simulating various degrees of stenosis (10-85%).

Main Results:

• An inversion in the timing of systolic spectral content was observed proximal and distal to hemodynamically significant stenoses.
• Specific time shifts (+22 ms for <50% stenosis, -20 to -38 ms for >50% stenosis) were measured.
• Calculated blood velocity increases (142-155 cm/s) in stenotic phantoms align with ultrasound measurements.

Conclusions:

• Flexible microphone arrays offer a promising noninvasive approach for detecting central venous stenosis.
• The developed signal processing technique effectively identifies turbulent flow indicative of stenosis.
• This technology could improve the diagnosis of central venous stenosis in hemodialysis patients.