# nnDetection: A Self-configuring Method for Medical Object Detection

**Authors:** Michael Baumgartner, Paul F. Jaeger, Fabian Isensee, Klaus H. Maier-Hein  
**arXiv:** [2106.00817](https://arxiv.org/abs/2106.00817)  
**Date:** 2021-06-01  
**Summary page:** https://www.summarizepaper.com/en/paper/2106.00817v2/

## Key points

- Medical object detection involves simultaneous localization and categorization of objects in medical images.
- Configuring methods for this task can be cumbersome and iterative, leading to a research bottleneck.
- Researchers have developed nnDetection - a self-configuring method for medical object detection that automates the configuration process.
- nnDetection adapts itself without any manual intervention to arbitrary medical detection problems while achieving results that are on par with or superior to state-of-the-art methods.
- The researchers demonstrate the effectiveness of nnDetection on two public benchmarks – ADAM and LUNA16 – and propose 11 further medical object detection tasks on public datasets for comprehensive method evaluation.
- The development of nnDetection represents a significant advancement in medical object detection research as it streamlines the configuration process while maintaining high levels of accuracy.
- The code for nnDetection is available on GitHub at https://github.com/MIC-DKFZ/nnDetection.

## Summary

Medical object detection is a crucial task in clinical diagnosis that involves simultaneous localization and categorization of objects in medical images. However, the process of configuring methods for this task can be cumbersome and iterative, leading to a research bottleneck. To address this challenge, researchers Michael Baumgartner, Paul F. Jaeger, Fabian Isensee and Klaus H. Maier-Hein have developed nnDetection – a self-configuring method for medical object detection that automates the configuration process. Building on the success of nnU-Net's approach to image segmentation, nnDetection adapts itself without any manual intervention to arbitrary medical detection problems while achieving results that are on par with or superior to state-of-the-art methods. The researchers demonstrate the effectiveness of nnDetection on two public benchmarks – ADAM and LUNA16 – and propose 11 further medical object detection tasks on public datasets for comprehensive method evaluation. The development of nnDetection represents a significant advancement in medical object detection research as it streamlines the configuration process while maintaining high levels of accuracy. The code for nnDetection is available on GitHub at https://github.com/MIC-DKFZ/nnDetection.