For reasons outlined above, alterations in the shape of the ECTI were expected to correlate to dysplasia, thus SHGM cross sectional images from normal and dysplastic oral mucosa were analyzed. Results from a semi-automated image analysis approach indicate ability to identify and quantify features in ECTI associated with dysplasia while more traditional measures of PI-103 cellular and layer based abnormalities remain possible with these imaging modalities. A Golden Syrian Hamster buccal model of dysplasia and oral cancer involving topical application of 9,10-dimethyl-1,2-benzanthracene was used. Histologically hamster buccal mucosa shows very similar features to sites of the human oral cavity including the floor of the mouth and ventral surface of the tongue, common sites for neoplastic transformation. The DMBA model of oral carcinogenesis is a well-characterized model that shows histological similarities to human oral precancer and cancer, following a similar multistep progression from normal to increasing grades of dysplasia and carcinoma-in-situ to cancer. Additionally, studies have examined the multistep process involving genetic events such as deregulation of oncogenes and tumor suppressor genes in this model and shown parallels in molecular marker alterations to human. The well documented histological and molecular similarities between this model and human oral precancer/cancer make this a useful model to study morphometric features associated with neoplasia. For this study, oral dysplasia was induced by topical application of 0.5% DMBA in mineral oil on the left cheek pouch three times a week for 8 weeks. Additional hamsters were treated only with mineral oil and used as controls. After 8 weeks of DMBA treatment hamsters were anesthetized with a mixture of 150-mg/kg ketamine and 2.5-mg/kg xylazine via intraperitoneal injection. The cheek pouch, a loose pouch of buccal mucosal tissue was stretched outside of the oral cavity and attached to a sample BAY-60-7550 holder having a flat surface. The pouch was rinsed with sterile PBS at room temperature and the hamster placed on a sample holder for microscopy with the buccal pouch accessible to the imaging objective. After MPAM-SHGM imaging the animals were sacrificed and biopsies were obtained from each imaged site and fixed in 10% neutral-buffered formalin. The fixed samples were embedded in paraffin, sectioned and stained by hematoxylin and eosin for histopathological grading. The current study was undertaken to demonstrate a new noninvasive way to visualize and characterize the ECTI during dysplasia when other well characterized early neoplastic changes in the epithelium are known to occur.