Fast Automated Segmentation and Reproducible Volumetry of Pulmonary Metastases in CT-Scans for Therapy Monitoring

摘要: The assessment of metastatic growth under chemotherapy belongs to the daily radiological routine and is currently performed by manual measurements largest nodule diameters. As in lung cancer screening where 3d volumetry methods have already been developed other groups, computer assistance would be beneficial improve speed reliability assessment. We propose a new morphology model based approach for fast reproducible pulmonary nodules that was explicitly applicable metastases which are frequently large, not necessarily spherical, often complexly attached vasculature chest wall. A database over 700 from more than 50 patient CT scans various scanners used test algorithm during development. An vivo reproducibility study conducted concerning volumetric analysis 105 8 patients were subjected low dose scan twice within several minutes. Low median volume deviations inter-observer (0.1%) inter-scan (4.7%) tests negligible average computation time 0.3 seconds measured. experiments revealed clinically significant change can detected reliably method.