Researchers and clinicians have called for early pregnancy biomarkers of preterm delivery, especially in the uterine cervix, a key component in the timing of birth. Speckle statistics analysis is a quantitative ultrasound tool for characterizing sub-resolution acoustic scattering which may offer a unique lens into the dynamic microstructure of the pregnant cervix. Our goal is to develop speckle statistics biomarkers to characterize the microstructural remodeling of the cervix to improve our understanding of the physiological processes which precede birth. However, the cervix presents unique challenges to quantitative analysis. For example, cervical stroma is spatially heterogenous and anisotropic. There is a lack of speckle statistics models for elongated structures like collagen—a prominent acoustic scatterer in the cervix. Additionally, speckle statistics metrics are relative to the imaging system used, presenting a barrier to commercial and clinical translation. These challenges are presented as opportunities for innovation in speckle statistics analysis—including methods that account for power loss in acoustic signals, the unique spatially varying ultrasonic resolution of each imaging system, and the presence of fiber-like acoustic scatterers. Application in an in vivo longitudinal study of the pregnant human cervix demonstrates the impact of these methods on the detection of microstructural changes during gestation.
Alexandra Christensen (2025) studied this question.