Ultrasonic tissue characterization for the differentiation of parotid gland tumors

The first ultrasonic tissue characterization system for the computerized differentiation of tumors of the parotid gland is presented. The system is based on a multifeature tissue characterization approach involving spectrum and texture parameters and using fuzzy inference systems as higher order classifiers. Baseband ultrasound echo data were acquired during conventional ultrasound imaging examinations of the salivary glands. Several tissue-describing parameters were calculated within numerous small regions of interest in order to evaluate local spectral and textural tissue properties. The parameters were processed by an adaptive network-based fuzzy inference system using the results of conventional histology after parotidectomy as the gold standard. Cases of parotid gland tumors and alterations include basal cell adenomas, monomorphic adenomas, pleomorphic adenomas, adenoid cysts, cysts and canaticulous adenomas. The results of the classification procedure are presented as a numerical score indicating the probability of a certain tumor or alteration for each parotid gland. In a pilot study, the system was evaluated on 23 cases of benign and malignant parotid gland tumors of patients undergoing parotidectomy. The ROC curve area given as the cross-validation mean and cr oss-vatidation standard deviation is A(ROC)=0.95 +/- 0.07 when using four-fold cross-validation over cases and differentiating between various malignant and benign parotid gland tumors as the positive target group and monomorphic adenomas as the negative target group. An exceptional equal error rate of E-EER=0.92 +/- 0.08 is achieved for the same setup. Some alterations which are of benign nature were counted to the positive group, as they occur too seldom to achieve a high probability for being considered safe if left untreated.