Weakly supervised pancreas segmentation based on class activation maps

Mona Schumacher, Andreas Genz, Mattias Heinrich


The quality of the segmentation of organs and pathological tissues has significantly improved in recent years by using deep learning approaches, which are typically trained with fully supervised learning. For these fully supervised learning methods, an immense amount of fully labeled training data is required, which are, especially in medicine, costly to generate. To overcome this issue, weakly supervised training methods are used, because they do not need fully labeled ground truth data. For the localization of objects weakly supervised learning has already become more important. Recently, weakly supervised learning also became increasingly important in the area of segmentation of pathological tissues. However, these currently available approaches still require additional anatomical information. In this paper, we present a weakly supervised segmentation method that does not need ground truth segmentations as input or additional anatomical information. Our method consists of three classification networks in sagittal, axial, and coronal direction that decide whether a slice contains the structure to be segmented. Then, we use the class activation maps of the classification output to generate a combined segmentation. Our network was trained for the challenging task of pancreas segmentation with the publicly available TCIA pancreas dataset and we reached Dice scores for slices of up to 0.86 and an overall Dice score of up to 0.53.

Original languageEnglish
Title of host publicationMedical Imaging 2020: Image Processing
EditorsIvana Išgum, Bennett A. Landman
Number of pages6
Publication date10.03.2020
Article number2549353
ISBN (Print)978-151063393-3
Publication statusPublished - 10.03.2020
EventMedical Imaging 2020: Image-Guided Procedures, Robotic Interventions, and Modeling - Houston, United States
Duration: 16.02.202019.02.2020
Conference number: 159792


Dive into the research topics of 'Weakly supervised pancreas segmentation based on class activation maps'. Together they form a unique fingerprint.

Cite this