Detection of Human Cervical Cancer by Probe-Based Quantitative Optical Coherence Tomography

Abstract

Most cervical cancers develop from squamous cells in the exocervix followed by stromal invasion, which alters the organization and morphology of collagen fibers. Therefore, morpho-structural remodeling of collagen fibers is closely associated with cancer progression. Collagen-based cancer detection requires not only techniques capable of qualified large-depth imaging but also computational sensitivity to extract subtle changes. Here, optical coherence tomography (OCT) is applied to collagen fibers in the exocervix. High-quality imaging into deep stroma is guaranteed by an all-fiber probe designed to have an extended depth of focus through the formation of the quasi-Bessel focusing beam. Collagen fibers provide dominant scattering signals in OCT images, and volume information is utilized to establish an optical biomarker reflecting variation gradient in fiber alignment and crimping. Detection of cervical cancer with a multi-parametric method is then evaluated by ex vivo imaging of human specimens and in vivo imaging of a murine model harboring human cervical cancer. Finally, the tumor potential index (TPI) is proposed by merging multiple metrics. The TPI map provides an intuitive illustration of cancer risk, which may guide clinicians more accurately to the correct location for biopsy.