A voxel-based neural approach (VBNA) to identify lung nodules in the ANODE09 study

摘要: The computer-aided detection (CAD) system we applied on the ANODE09 dataset is devoted to identify pulmonary nodules in low-dose and thin-slice computed tomography (CT) images: developed two different systems for internal (CADI) juxtapleural (CADJP) framework of italian MAGIC-5 collaboration. basic modules CADI subsystem are: a 3D dot-enhancement algorithm nodule candidate identification an original approach, referred as Voxel-Based Neural Approach (VBNA), reduce amount false-positive findings based neural classifier working at voxel level. To detect CADJP procedure enhancing regions where many pleura surface normals intersect, followed by VBNA classification. We present both FROC curves obtained 5 annotated example dataset, all 50 test cases.