Inter Extreme Points Geodesics for End-to-End Weakly Supervised Image Segmentation

Abstract

We introduce $\text{InExtremIS}$, a weakly supervised 3D approach to train a deep image segmentation network using particularly weak train-time annotations: only 6 extreme clicks at the boundary of the objects of interest. Our fully-automatic method is trained end-to-end and does not require any test-time annotations. From the extreme points, 3D bounding boxes are extracted around objects of interest. Then, deep geodesics connecting extreme points are generated to increase the amount of "annotated" voxels within the bounding boxes. Finally, a weakly supervised regularised loss derived from a Conditional Random Field formulation is used to encourage prediction consistency over homogeneous regions. Extensive experiments are performed on a large open dataset for Vestibular Schwannoma segmentation. $\text{InExtremIS}$ obtained competitive performance, approaching full supervision and outperforming significantly other weakly supervised techniques based on bounding boxes. Moreover, given a fixed annotation time budget, $\text{InExtremIS}$ outperforms full supervision. Our code and data are available online.

Early accept at MICCAI 2021 - code available at: https://github.com/ReubenDo/InExtremIS

References in corpus (10)

• Automated Design of Deep Learning Methods for Biomedical Image Segmentation
• Interactive Medical Image Segmentation using Deep Learning with Image-specific Fine-tuning
• TorchIO: A Python library for efficient loading, preprocessing, augmentation and patch-based sampling of medical images in deep learning
• DeepIGeoS: A Deep Interactive Geodesic Framework for Medical Image Segmentation
• Constrained-CNN losses for weakly supervised segmentation
• Manual segmentation versus semi-automated segmentation for quantifying vestibular schwannoma volume on MRI
• Automatic Segmentation of Vestibular Schwannoma from T2-Weighted MRI by Deep Spatial Attention with Hardness-Weighted Loss
• Scribble-based Domain Adaptation via Co-segmentation
• Permutohedral Attention Module for Efficient Non-Local Neural Networks
• Extreme Points Derived Confidence Map as a Cue For Class-Agnostic Segmentation Using Deep Neural Network

Cited by in corpus (2)

• Weakly-Supervised 3D Medical Image Segmentation using Geometric Prior and Contrastive Similarity
• OOD-SEG: Exploiting out-of-distribution detection techniques for learning image segmentation from sparse multi-class positive-only annotations