One of the first steps in lung cancer detection and staging is the assessment of mediastinal lymph nodes in computed tomography (CT) datasets of the chest. This assessment is usually performed manually, which can be an error-prone and time-consuming process. To overcome these problems, we here present a new method to fully automate the detection of mediastinal lymph node candidates in contrast-enhanced chest CT. Based on the segmentation of mediastinal anatomy such as bronchial tree and aortic arch and making use of anatomical knowledge, we utilize Hessian eigenvalue analysis to detect early lymph node candidates. Their initially high number is then significantly reduced to those candidates, which are blob-like and within a specific intensity interval. We applied our method to 5 cases with suspected lung cancer. Our method did not exceed 5 minutes of computation time, and we achieved an average sensitivity of 82.1% and an average precision of 12.5%.     «

One of the first steps in lung cancer detection and staging is the assessment of mediastinal lymph nodes in computed tomography (CT) datasets of the chest. This assessment is usually performed manually, which can be an error-prone and time-consuming process. To overcome these problems, we here present a new method to fully automate the detection of mediastinal lymph node candidates in contrast-enhanced chest CT. Based on the segmentation of mediastinal anatomy such as bronchial tree and aortic a...     »