@article{oai:kanazawa-u.repo.nii.ac.jp:00013098,
 author = {Takemura, Akihiro and Suzuki, M. and Harauchi, H. and Okumura, Y.},
 issue = {SUPPL.7},
 journal = {Computer-Assisted Radiology and Surgery},
 month = {Jun},
 note = {Accessory cardiac bronchus (ACB) is a rare bronchial anomaly This anomaly often goes undetected due to its low frequency and lack of symptoms in most cases. However, ACB should be diagnosed because it sometimes causes infection and/or hemoptysis. Here, we present a rule-based and 2D-based scheme, which can detect ACB in chest multi-slice computed tomography volume data. We have developed a computer-aided scheme for detection of ACB. This scheme performs extraction of the bronchial tree structure, determination of trachea bifurcation, and detection of ACB on the right bronchi. The sensitivity and specificity of this scheme were evaluated in 9 cases: 3 with ACB. 3 with tracheal bronchus and no ACB, and 3 normal cases. All of the subjects were examined with MSCT scanners (LightSpeed Ultra 16 and LightSpeed Plus; GE Medical Systems, Milwaukee. WI, USA) using a contiguous slice thickness of 2.5 mm. We also measured the time taken for examination using the scheme on a personal computer (CPU: 2.4 GHz Pentium 4, Memory: 1 Gbyte; OS: Red Hat Linux 9). The scheme detected ACB in two of the three cases (Sensitivity: 66.7%), and all of the cases without ACB were identified as having no ACB (Specificity: 100%). The mean processing time was 13.5 seconds (SD: 6.4) per case, and therefore this scheme could be suitable for prechecking before actual diagnosis., 金沢大学大学院医学系研究科},
 pages = {362--364},
 title = {A rule-based scheme for detection of accessory cardiac bronchus in chest computed tomography images},
 volume = {1},
 year = {2006}
}