Integrated Detection Network (IDN) for Pose and Boundary Estimation in Medical Images

Abstract

The expanding role of complex object detection algorithms introduces a need for flexible architectures that simplify interfacing with machine learning techniques and offer easy-to-use training and detection procedures. To address this need, the Integrated Detection Network (IDN) proposes a conceptual design for rapid prototyping of object and boundary detection systems. The IDN uses a strong spatial prior present in the medical imaging domain and a large annotated database of images to train robust detectors. The best detection hypotheses are propagated throughout the detection network using sequential sampling techniques. The effectiveness of the IDN is demonstrated on two learning-based algorithms: (1) automatic detection of fetal brain structures in ultrasound volumes, and (2) liver boundary detection in MRI volumes. Modifying the detection pipeline is simple and allows for immediate adaptation to the variations of the desired algorithms. Both systems achieved low detection error (3.09 and 4.20 mm for two brain structures and 2.53 mm for boundary).

Results

Figure 1: Two IDN configurations for localizing cerebellum in ultrasound volumes of the fetal head. The pipelines A and B have different subnetworks at 4 mm and 2 mm resolutions and same modules at 1 mm resolution..

Figure 2: The MRI liver segmentation network uses the rigid detector to locate the liver. Then several layers boundary detection is performed on different image and mesh resolutions.

Publications and Further Reading