Sergios Gatidis1
1University Hospital Tübingen, Tubingen, Germany
Segmentation of anatomical structures is a prerequisite for numerous down-stream tasks in medical image analysis. The advance of machine learning methodology over the past decade, specifically the introduction of dedicated deep learning architectures (e.g. the UNet architecture) has enabled automated organ segmentation on 3D medical image data with a previously unknown accuracy. These methods have widespread use for many different clinical and scientific applications such as automated quantification of organ volumes or characterization of tumor lesions.
Despite these impressive algorithmic developments, significant challenges still lie ahead when it comes to reliable clinical deployment of machine learning methods for abdominal organ segmentation. Specifically with respect to MRI data, which can be highly variable and prone to image artifacts, thorough quality control of segmentation results is of high importance.
This talk will provide an overview of current methods and applications of machine learning-based methods for abdominal organ segmentation on MRI data and discuss current challenges and research directions.Acknowledgements
No acknowledgement found.References
Lavdas I, Glocker B, Kamnitsas K, Rueckert D, Mair H, Sandhu A, Taylor SA, Aboagye EO, Rockall AG. Fully automatic, multiorgan segmentation in normal whole body magnetic resonance imaging (MRI), using classification forests (CFs), convolutional neural networks (CNNs), and a multi-atlas (MA) approach. Med Phys. 2017 Oct;44(10):5210-5220.
Ronneberger O, Fischer P, Brox T, editors. U-Net: Convolutional Networks for Biomedical Image Segmentation. Medical Image Computing and Computer-Assisted Intervention – MICCAI 2015; 2015
Isensee F, Jaeger PF, Kohl SAA, Petersen J, Maier-Hein KH. nnU-Net: a self-configuring method for deep learning-based biomedical image segmentation. Nat Methods. 2021 Feb;18(2):203-211.
Kart T, Fischer M, Küstner T, Hepp T, Bamberg F, Winzeck S, Glocker B, Rueckert D, Gatidis S. Deep Learning-Based Automated Abdominal Organ Segmentation in the UK Biobank and German National Cohort Magnetic Resonance Imaging Studies. Invest Radiol. 2021 Jun 1;56(6):401-408.