Endoscopic evaluation of the colorectum is limited to the mucosal surface and provides no functional or structural information regarding subsurface changes. Targeted diagnostics and individualized treatment, however, requires this information. In this ex vivo study of human colorectal tissue, we use swept-source optical coherence tomography to create quantified subsurface scattering coefficient maps of normal and cancerous tissue. Specifically, we use a novel wavelet-based-curve-fitting method to generate subsurface scattering coefficient maps. The angular spectra of scattering coefficient maps of normal tissues exhibit a spatial feature distinct from those of abnormal tissues. The en face scattering coefficient maps of the normal colon contain a large area of homogenous scattering coefficients with periodic dot patterns, while the scattering coefficient map of cancer region shows a large area of heterogeneous scattering coefficients. An angular spectrum index to quantify the differences between the normal, abnormal, and treated tissues is derived, and its strength in revealing subsurface cancer in ex vivo samples is statistically analyzed. Using this index, we differentiate malignant from normal colonic tissue. Additionally, we also found that rectal cancers completely destroyed by preoperative treatment appear much more similarly to normal tissue than the original malignancy. The study demonstrates that the angular spectrum of the scattering coefficient map can effectively reveal subsurface colorectal cancer and help clinicians identify patients who don’t require surgical intervention.
|