Applications and challenges for isolation of ultrasound signals from swept source optical coherent tomography

Abstract

Ultrasound is a powerful tool for performing cross sectional imaging. Non-contact detection of ultrasound via light is important when physical coupling to the sample is precluded by sample sensitivity and has applications in metrology and medical imaging. Optical coherence tomography provides similar structural information to ultrasound, typically with higher resolution and lower penetration depths. The phase sensitivity of optical coherence tomography lends itself well to detection of acoustic waves. Furthermore, swept source optical coherence tomography has demonstrated sensitivity to megahertz frequency acoustic waves by monitoring changes in optical pathlength as a function of wavelength during the sweep. Recent advances in swept source laser technology have improved the ability to detect and isolate ultrasound acoustic signals. Swept source optical coherence tomography has ultimate sensitivity to acoustic waves defined by the shot-noise limit of the imaging system and can achieve theoretical sensitivities as low as ~10 pm from ultrasonic waves from an ideal reflector. Measurement of ultrasound via optical coherence tomography is complicated by tissue dispersion, signal crosstalk between imaging depths, non-ideal tissue reflectors, and laser stability. Here, we present and compare methods for extracting ultrasound signals from OCT datasets based on techniques for demodulation of frequency modulated signals. The overall sensitivity, signal to noise ratio, and signal isolation are compared between methods of quadrature demodulation, Hilbert transform based signal extraction, and phase locked loops, and future directions of this technology are discussed.