Automatic Detection and Volumetric Assessment of Lymph Nodes (WIP)

One of the most important measures of therapy efficacy in cancer patients is the assessment of lymph nodes over time. Lymph node segmentation, in particular, plays a key role in tumor staging (first time CT scans) and tumor progression (re-staging and comparison of time series in order to determine the permanence of the tumor’s response to radiotherapy or chemotherapy).

Lymph nodes are not only often difficult to distinguish from surrounding tissue, they are especially difficult to measure volumetrically. This is partly due to their unpredictable shape caused by the irregular expansion of malignant structures. It is therefore not sufficient to identify and segment the lymph nodes, the surrounding organs must be checked as well.

Based on Definiens Cognition Network Technology®, abnormal lymph nodes in CT images can be automatically detected, segmented and volumetrically measured. This can assist the researching radiologist in detecting all the relevant lymph nodes and may support immediate feedback on the efficacy of the cancer therapy. Organs are treated as context objects and can be automatically segmented - a vital step towards the detection of lymph nodes.