Go back    Home Solutions nnDetection: A Self-configuring Method for Medical Object Detection

nnDetection: A Self-configuring Method for Medical Object Detection (nnDetection)

Authors: Michael Baumgartner, Paul F. Jäger, Fabian Isensee, Klaus H. Maier-Hein

Keywords: detection, 3d-object-detection, pytorch-implementation, retina-unet, out-of-the-box, biomedical image analysis

Simultaneous localisation and categorisation of objects, also referred to as object detection, is of high relevance for many applications which rely on rating of objects rather than e.g. pixels or entire images. For this task, the cumbersome and iterative process of method configuration constitutes a major research and application bottleneck. Recently, nnU-Net has tackled this challenge for the task of image segmentation with great success. Following nnU-Net’s agenda, we systematised and automated the configuration process for 3D object detection. The resulting self-configuring method, nnDetection, adapts itself without any manual intervention to arbitrary 3D detection problems while achieving results en par with or superior to the state-of-the-art. While it was initially developed for medical imaging, it can be used for any volumetric imaging modality.


Publications

nnDetection: A Self-configuring Method for Medical Object Detection

Baumgartner M, Jäger P, Isensee F, Maier-Hein K - Medical Image Computing and Computer Assisted Intervention – MICCAI 2021 - 2021


nnDetection: A Self-configuring Method for Medical Object Detection Image
License
Apache-2.0, BSD-3, BSD-2

Helmholtz Imaging spinning wheel

Please wait, your data is processed