Automated Detection of Lymphadenopathy in Neck Resulting from Oral SCC on 3D CT Imaging

Abstract

Background/Objectives: Detection of lymphadenopathy is challenging issue in medical field. Squamous Cell Carcinoma (SCC) can lead to a certain type of lymphadenopathy which is different from all other pathological factors. Nowadays detection of these cases is done by expert radiologists, Computing Tomography (CT), and sonography and etc., challenges in detection of this type of lymphadenopathy in neck is mainly due to neck anatomy which similar objects are close to each other. Methods/Statistical Analysis: In this paper, was presented a method to detect this type of the lymphadenopathy in neck by 3 dimensional (3D) image processing techniques and Computer-Aided Diagnosis (CAD) systems. This method consists of four steps. The first is preprocessing, the second is thresholding and morphological operation, the third is feature extraction and the last is classification. By using this method, detection of lymphadenopathy will be done more accurate and less time consuming. Findings: This method is done in 18 neck CT data sets, consisting of lymphadenopathy by SCC cells. The sensitivity of using this method is 94%. Applications/Improvements: The goal is to introduce a sensitive, accurate and generalizable method with the least false positive.