Towards the prediction of spontaneous preterm birth by cervical collagen imaging using polarization-sensitive OCT

Introduction: Preterm birth (PTB) remains one of the leading risk factors of neonatal mortality despite the ongoing research to reduce its incidence. Spontaneous PTB (sPTB) can arise as a result of premature remodeling of the collagenrich extracellular matrix (ECM) of the cervix. Non-invasive biomarkers of cervical collagen content, such as birefringence, may thus have diagnostic value in predicting sPTB. In our previous work, we showed that polarizationsensitive OCT (PS-OCT) can characterize in vitro cervical collagen alignment through measurements of the backscattered light’s polarization state, differentiating collagen from the surrounding ECM of the cervical epithelium. To advance this towards in vivo measurements, we have developed a PS-OCT system, based on Jones Matrix calculus (JMOCT), which is an extension of PS-OCT optimized for fiber-optic beam delivery. We have coupled this to an in-house designed colposcopic probe, to characterize changes in cervical collagen organization during term and preterm pregnancies. This study aims to assess the performance of this system prior to clinical studies.

Methods: JM-OCT images of several material and biological samples with expected birefringent properties and depthbased landmarks were obtained with the colposcopic probe, including in vitro ovine cervix. OCT B-scans containing 700 A-scans were acquired with filtered phase retardance to assess the system’s resolution, imaging depth and area, and ability to present changes in sample birefringence.

Results: The system successfully produced high intensity OCT B-scans with high axial resolution (14 μm) and imaging depth (up to 2.1 mm), displaying depth-resolved landmarks including blood vessels within the in vivo skin’s papillary dermis. Changes in material and ECM birefringence were clearly displayed in phase retardance images.

Conclusion: The performance of our in-house designed colposcopic probe coupled with the in-house built JM-OCT engine shows great potential to image in vivo cervical collagen orientation and abundance in future studies. With its high axial resolution and phase retardance imaging, we aim to investigate the application of our imaging system clinically, to assess its value in understanding, predicting, and preventing PTB.