Automated classification of four types of developmental odontogenic cysts.

Odontogenic cysts originate from remnants of the tooth forming epithelium in the jaws and gingiva. There are various kinds of such cysts with different biological behaviours that carry different patient risks and require different treatment plans. Types of odontogenic cysts can be distinguished by the properties of their epithelial layers in H&E stained samples. Herein we detail a set of image features for automatically distinguishing between four types of odontogenic cyst in digital micrographs and evaluate their effectiveness using two statistical classifiers - a support vector machine (SVM) and bagging with logistic regression as the base learner (BLR). Cyst type was correctly predicted from among four classes of odontogenic cysts between 83.8% and 92.3% of the time with an SVM and between 90 ± 0.92% and 95.4 ± 1.94% with a BLR. One particular cyst type was associated with the majority of misclassifications. Omission of this cyst type from the data set improved the classification rate for the remaining three cyst types to 96.2% for both SVM and BLR.

Segmentation of epithelium in H&E stained odontogenic cysts.

An algorithm for the automated segmentation of epithelial tissue in digital images of histologic tissue sections of odontogenic cysts (cysts originating from residual odontogenic epithelium) is presented. The algorithm features an image standardization process that greatly reduces variation in luminance and chrominance between images due to variations in sample preparation. Segmentation of the epithelial regions of images uses an algorithm based on binary graph cuts where graph weights depend on probabilities obtained from colour histogram models of epithelium and stroma image regions. Algorithm training used a data set of 38 images of four types of odontogenic cyst and was tested using a separate data set of 35 images of the same four cyst types. The best parameters for the segmentation algorithm were determined using a response-surface optimizer. The best parameter set resulted in an overall mean (± std. dev.) sensitivity of 91.5 ± 17% and overall mean specificity of 85.1 ± 18.6% on the training set. Particularly good results were obtained for dentigerous and odontogenic keratocysts for which the mean sensitivities/specificities were 91.9 ± 6.15%/97.4 ± 2.15% and 96.1 ± 1.98%/98.7 ± 3.16%, respectively. Our method is potentially applicable to many pathological conditions in similar tissues, such as skin and mucous membranes where there is a clear microscopic distinction between epithelium and connective tissues.