作者: Burçin Buket Oğul , Polat Koşucu , Ahmet Özçam , Sümeyra Demir Kanik
DOI: 10.1007/978-3-319-11128-5_38
关键词:
摘要: The high fatality rate of lung cancer brings a lot attention to Computer Aided Detection (CAD) systems for nodule detection. A CAD can help radiologist reduce the time, and effort in analyzing images increase accuracy. In this paper fully automated system is presented detect nodules from X-ray images. Proposed segments area, identifies candidates, extract features classifies candidates as or not. output highlighted candidate areas with size information. Publicly available JSRT (Japanese Society Radiological Technology) are used validate system. We achieved %80 sensitivity an average 6.4 FPs. tested on different dataset 417 %76 6.7 shows potential automate detection satisfying