Ultrahigh-Resolution Endoscopic Optical Coherence Tomography for In Vivo Mouse Colonoscopy

Abstract

In vivo monitoring of mouse models of colon cancer promises to reduce the cost of research by improving sacrifice timing and allowing serial studies that observe the progression of disease and drug efficacy in a relatively small set of animals. Optical coherence tomography (OCT) is an optical analog of ultrasound imaging, capable of minimally-invasive mapping of light scatter intensity up to 2 mm deep in tissue. In this work, factors limiting resolution in OCT were examined and devices were created and applied to mouse colon imaging that extended the state-of-the-art in endoscopic ultrahigh-resolution OCT. First, axial chromatic aberration of the objective optics acts as a spectral filter in the sample arm limiting the effective bandwidth of the system. An achromatized endoscope design was demonstrated that achieved axial resolution of 2.3 mum in tissue and 4.4 mum lateral spot diameter with 101 dB sensitivity when interfaced with a time domain OCT system utilizing a 10-femtosecond laser (bandwidth=150 nm FWHM, center wavelength=800 nm). Second, dispersion matching between the sample and reference arms presents the practical resolution limit to endoscopic implementations including a separate, fiber-based reference arm. A second endoscope incorporated the reference arm into the tip of the endoscope using a novel custom beamsplitter prism and achieved 2.4 mum axial resolution in tissue without adjustments for pathlength or dispersion matching when interfaced with a spectrometer-based frequency domain OCT system and a similar laser. Third, non-linear dispersion of the sample media with respect to wavelength causes distortion and broadening of the axial point spread function when data are sampled uniformly in optical frequency. An experiment was performed on high dispersion glass to demonstrate that dispersion artifact free imaging can be achieved without post process corrections if the samples are acquired at equal intervals of media index of refraction divided by vacuum wavelength. Finally, other microscopic modalities that depend on tissue scatter intensity are used to find the origins of scatter in the mouse colonic mucosa. These observations are used to explain unexpected features found in ultrahigh-resolution tomograms collected with the two endoscopes presented.