Automatic mediastinal lymph node detection in chest CT

Abstract

Computed tomography (CT) of the chest is a very common staging investigation for the assessment of mediastinal, hilar, and intrapulmonary lymph nodes in the context of lung cancer. In the current clinical workflow, the detection and assessment of lymph nodes is usually performed manually, which can be error-prone and timeconsuming. We therefore propose a method for the automatic detection of mediastinal, hilar, and intrapulmonary lymph node candidates in contrast-enhanced chest CT. Based on the segmentation of important mediastinal anatomy (bronchial tree, aortic arch) and making use of anatomical knowledge, we utilize Hessian eigenvalues to detect lymph node candidates. As lymph nodes can be characterized as blob-like structures of varying size and shape within a specific intensity interval, we can utilize these characteristics to reduce the number of false positive candidates significantly. We applied our method to 5 cases suspected to have lung cancer. The processing time of our algorithm did not exceed 6 minutes, and we achieved an average sensitivity of 82.1% and an average precision of 13.3%.

Automatic Detection and Segmentation of Axillary Lymph Nodes

“…There is a limited amount of work directed to automatic lymph node detection [4,5,7]. These works target mediastinal [5], abdominal [7] and neck [4] lymph nodes while our work targets axillary lymph nodes.…”

“…These works target mediastinal [5], abdominal [7] and neck [4] lymph nodes while our work targets axillary lymph nodes. The axillary lymph nodes are far from airways or major vessels, so a segmentation of the vessels or airways is not necessary.…”

“…Previous work [5] relied on Hessian-based blobness measures and other filters for removing lymph node false positives. Our work comes from a learning perspective, where the appropriate features that discriminate the lymph nodes from the negatives are learned from training examples.…”

Segmentation Based Features for Lymph Node Detection from 3-D Chest CT

“…The comparability is however limited because of different data, different criterions for a detection, different body regions and different minimum lymph node sizes used for evaluation. Both [9] and [5] report a very high number of false alarms and also consider a lymph node already as detected if there is just overlap with the automatic segmentation. In [3], very good results are reported, but the method was evaluated on a single dataset.…”

“…In [9,5], lymph nodes are detected by a cascade of filters (Hessian based, morphological operations and a so-called 3-D Min-DD filter). In [3], lymph nodes are detected and segmented by fitting a mass-spring model at different positions.…”