Identification of oral cancer in OCT images based on an optical attenuation model.
Lasers Med Sci
; 35(9): 1999-2007, 2020 Dec.
Article
em En
| MEDLINE
| ID: mdl-32335743
ABSTRACT
Surgery is still the first choice to treat oral cancer, where it is important to detect surgical margins in order to reduce cancer recurrence and maintain oral-maxillofacial function simultaneously. As a non-invasive and in situ imaging technique, optical coherence tomography (OCT) can obtain images close to the resolution of histopathology, which makes it have great potential in intraoperative diagnosis. However, it is not enough to find surgical margins accurately just observing OCT images directly and qualitatively. The purpose of this study is to identify oral cancer in OCT images by investigating the quantitative difference of cancer and non-cancer tissue. Based on an available optical attenuation model and the axial confocal PSF of a home-made swept source OCT system, by using fresh ex vivo human oral tissues from 14 patients of oral squamous cell carcinoma (OSCC) as the samples, diagnosis with sensitivity (97.88%) and specificity (83.77%) was achieved at the attenuation threshold of 4.7 mm-1, and the accuracy of identification reached 91.15% in our study. Our preliminary results demonstrated that the oral cancer resection will be guided accurately and the clinical application of OCT will be further promoted by deeply mining the information hidden in OCT images.
Palavras-chave
Texto completo:
1
Coleções:
01-internacional
Base de dados:
MEDLINE
Assunto principal:
Neoplasias Bucais
/
Tomografia de Coerência Óptica
Tipo de estudo:
Diagnostic_studies
/
Prognostic_studies
Limite:
Adult
/
Aged
/
Female
/
Humans
/
Male
/
Middle aged
Idioma:
En
Revista:
Lasers Med Sci
Assunto da revista:
BIOTECNOLOGIA
/
RADIOLOGIA
Ano de publicação:
2020
Tipo de documento:
Article
País de afiliação:
China