Output details
11 - Computer Science and Informatics
Kingston University
Shape based computer-aided detection of lung nodules in thoracic CT images
<28> A new Computer-Aided Detection method for lung nodules detection (solid and ground-glass opacity types) is presented. Validated on a clinical dataset of 108 thoracic CT scans (185 solid and 35 GGO nodules) it gave an average detection rate of 90.2%, (8.2 FP/scan). The high detection rate, fast computation and applicability to different imaging conditions/nodule types proved suitable for clinical practice. Patent has been granted (Dehmeshki sole inventor), approvals received by FDA (US) and CE (EU) and it attracted a high citation rate. Method is integrated in commercial applications and used by companies such as Vital Images, Viatronix, Terarecon.