AUTOMATIC DETECTION OF GROUND GLASS OPACITIES ON LUNG HRCT USING MULTIPLE NEURAL NETWORKS

The purpose of this study was to implement neural networks and expert rules for the automatic detection of ground glass opacities (GG) on hi gh-resolution computed tomography (HRCT). Different approaches using s elf-organizing neural nets as well as classifications of lung HRCT wit h and without the use of explicit textural parameters have been applie d in preliminary studies. In the present study a hybrid network of thr ee single nets and an expert rule was applied for the detection of GG on 120 HRCT scans from 20 patients suffering from different lung disea ses. Single nets alone were not capable to reliably detect or exclude GG since the false-positive rate was greater than 100 % with regard to the area truly involved, more than 50 pixels throughout, and the true -positive rate was greater than 95 %. The hybrid network correctly cla ssified 91 of 120 scans. Mild GG was false positive in 15 cases with l ess than 50 pixels, which was judged not clinically relevant. The : pi tfalls were: partial volume effects of bronchovascular bundles and the chest wall. Motion artefacts and diaphragm were responsible for 11 mi sclassifications. Hybrid networks represent a promising tool for an au tomatic pathology-detecting system. They are ready to use as a diagnos tic assistant for detection, quantification and follow-up of ground gl ass opacities, and further applications are underway.