Presentation
27 April 2016 Automated segmentation of oral mucosa from wide-field OCT images (Conference Presentation)
Ryan N. Goldan, Anthony M. D. Lee, Lucas Cahill, Kelly Liu, Calum MacAulay, Catherine F. Poh D.D.S., Pierre Lane
Author Affiliations +
Abstract
Optical Coherence Tomography (OCT) can discriminate morphological tissue features important for oral cancer detection such as the presence or absence of basement membrane and epithelial thickness. We previously reported an OCT system employing a rotary-pullback catheter capable of in vivo, rapid, wide-field (up to 90 x 2.5mm2) imaging in the oral cavity. Due to the size and complexity of these OCT data sets, rapid automated image processing software that immediately displays important tissue features is required to facilitate prompt bed-side clinical decisions. We present an automated segmentation algorithm capable of detecting the epithelial surface and basement membrane in 3D OCT images of the oral cavity. The algorithm was trained using volumetric OCT data acquired in vivo from a variety of tissue types and histology-confirmed pathologies spanning normal through cancer (8 sites, 21 patients). The algorithm was validated using a second dataset of similar size and tissue diversity. We demonstrate application of the algorithm to an entire OCT volume to map epithelial thickness, and detection of the basement membrane, over the tissue surface. These maps may be clinically useful for delineating pre-surgical tumor margins, or for biopsy site guidance.
Conference Presentation
© (2016) COPYRIGHT Society of Photo-Optical Instrumentation Engineers (SPIE). Downloading of the abstract is permitted for personal use only.
Ryan N. Goldan, Anthony M. D. Lee, Lucas Cahill, Kelly Liu, Calum MacAulay, Catherine F. Poh D.D.S., and Pierre Lane "Automated segmentation of oral mucosa from wide-field OCT images (Conference Presentation)", Proc. SPIE 9698, Advanced Biomedical and Clinical Diagnostic and Surgical Guidance Systems XIV, 96980R (27 April 2016); https://doi.org/10.1117/12.2211122
Lens.org Logo
CITATIONS
Cited by 1 scholarly publication.
Advertisement
Advertisement
KEYWORDS
Optical coherence tomography

Image segmentation

Tissues

Cancer

Image processing algorithms and systems

Imaging systems

In vivo imaging

Back to Top