Abstract
The treatment of choice for canine alimentary tract neoplasms is surgical excision, but it can sometimes be difficult to achieve wide margins due to neoplasm location, size, or distribution. Optical coherence tomography (OCT) is a rapid, noninvasive imaging modality that uses light to characterise tissue microstructure to allow identification of different tissue types. Spectral domain (SD)-OCT allows for differentiation based on the total light intensity reflected from the tissue. Polarisation sensitive (PS)-OCT detects the polarisation state of light reflected by the tissues. The polarisation properties are phase retardation, degree of polarisation uniformity (DOPU), and optical axis. Our objective was to qualitatively characterise different tissues at the excision sites of alimentary tract neoplasms using OCT. Oral, liver, and other alimentary tumours including stomach, intestine, and pancreas were imaged. Samples were then fixed in formalin, paraffin embedded and stained with haematoxylin and eosin. OCT images and histology slides were compared, and the tissues were qualitatively described by a single investigator. We hypothesized that PS-OCT imaging would provide distinguishing characteristics of tissue appearances that could be used in the future to train observers or artificial intelligence to identify incomplete margins. Our results showed that alimentary tract tumours have disorganised microstructures on SD-OCT and PS-OCT DOPU, and PS-OCT phase retardation and optical axis values that differ from normal tissues. Thus, these characteristics can be used to differentiate neoplastic and normal tissues at surgical margins.