Active Learning for Pulmonary Nodule Detection

Expensive and time-consuming medical imaging annotation is one of the big challenges for the deep-learning-based computer-aided diagnose (CAD) nodule detection system on the low-dose computed tomography (CT). To address this, we propose a novel active learning approach to improve the training efficiency for a deep-network-based lung nodule detection framework as well as reducing the annotation cost. The informative CT scans, such as the samples that inconspicuous or likely to produce higher false positives rates, are selected and further annotated by the radiologists for the nodule detector network training. A simple yet effective schema suggests the sample by ranking the uncertainty loss predicted from global scan-level and local proposal-level perspective through multi-layer feature maps from the backbone network and the region of interests (ROIs) from the detector network respectively.

• Machine Perception